2023-12-20 08:45:34 +00:00
<?xml version="1.0" encoding="UTF-8"?>
web: Replace lingui.js with lit-localize (#5761)
* \#\# Details
web: replace lingui with lit/localize
\#\# Changes
This rather massive shift replaces the lingui and `t()` syntax with lit-localize, XLIFF, and the `msg()`
syntax used by lit-localize. 90% of this work was mechanized; simple perl scripts found and replaced
all uses of `t()` with the appropriate corresponding syntax for `msg()` and `msg(str())`.
The XLIFF files were auto-generated from the PO files. They have not been audited, and they should be
checked over by professional translators. The actual _strings_ have not been changed, but as this was
a mechanized change there is always the possibility of mis-translation-- not by the translator, but by
the script.
* web: revise lit/localize: fix two installation issues.
* web: revise localization
TL;DR:
- Replaced all of Lingui's `t()` syntax with `msg()` syntax.
- Mechanically (i.e with a script) converted all of the PO files to XLIFF files
- Refactored the localization code to be a bit smarter:
- the function `getBestMatchLocale` takes the locale lists and a requested locale, and returns the
first match of:
- The locale's code exactly matches the requested locale
- The locale code exactly matches the prefix of the requested locale (i.e the "en" part of "en-US")
- the locale code's prefix exactly matches the prefix of the requested locale
This function is passed to lit-locate's `loadLocale()`.
- `activateLocale()` just calls `loadLocale()` now.
- `autodetectLanguage` searches the following, and picks the first that returns a valid locale
object, before passing it to `loadLocale()`:
- The User's settings
- A `?locale=` component found in `window.location.search`
- The `window.navigator.language` field
- English
The `msg()` only runs when it's run. This seems obvious, but it means that you cannot cache
strings at load time; they must be kept inside functions that are re-run so that the `msg()` engine
can look up the strings in the preferred language of the user at that moment.
You can use thunks-of-strings if you really need them that way.
* Including the 'xliff-converter' in case anyone wants to review it.
* The xliff-converter is tagged as 'xliff-converter', but has been
deleted.
\#\# Details
- Resolves #5171
\#\# Changes
\#\#\# New Features
- Adds a "Add an Application" to the LibraryView if there are no applications and the user is an administrator.
\#\#\# Breaking Changes
- Adds breaking change which causes \<issue\>.
\#\# Checklist
- [ ] Local tests pass (`ak test authentik/`)
- [ ] The code has been formatted (`make lint-fix`)
If an API change has been made
- [ ] The API schema has been updated (`make gen-build`)
If changes to the frontend have been made
- [ ] The code has been formatted (`make web`)
- [ ] The translation files have been updated (`make i18n-extract`)
If applicable
- [ ] The documentation has been updated
- [ ] The documentation has been formatted (`make website`)
* web: fix redundant locales for zh suite.
* web: prettier pass for locale update
* web: localization moderization
Changed the names of the lit-localize commands to make it clear they're
part of the localization effort, and not just "build" and "extract".
* update transifex config
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix package lock?
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* use build not compile
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: conversion to lit-localize
The CI produced a list of problems that I hadn't caught earlier,
due to a typo ("localize build" is correct, "localize compile" is
not) I had left in package.json. They were minor and linty, but
it was still wise to fix them.
* web: replace lingui with lit/locale
This commit fixes some minor linting issues that were hidden by a typo in package.json. The
issues were not apparently problematic from a Javascript point of view, but they pointed
to sloppy thinking in the progression of types through the system, so I cleaned them
up and formalized the types from LocaleModule to AkLocale.
* web: replace lingui with lit/localize
One problem that has repeatedly come up is that localize's templates do not produce
JavaScript that conforms with our shop style. I've replaced `build-locale` with
a two-step that builds the locale *and* ensures that it conforms to the shop style
via `prettier` every time.
* web: replace lingui with lit-locale
This commit applies the most recent bundle of translations to the
new lit-locale aspect component. It also revises the algorithm
for *finding* the correct locale, replacing the complex fall-back
with some rather straightforward regular expressions.
In the case of Chinese, the fallback comes at the end of the
selection list, which may not be, er, politically valuable
(since Taiwan and Hong Kong come before, being exceptions that
need to be tested). If we need a different order for presentation,
that'll be a future feature.
* web: replace lingui with lit/locale
Well, that was embarassing.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-06-02 15:08:36 +00:00
<xliff version="1.2" xmlns="urn:oasis:names:tc:xliff:document:1.2">
2023-12-20 08:45:34 +00:00
<file target-language="zh-Hant" source-language="en" original="lit-localize-inputs" datatype="plaintext">
<body>
<trans-unit id="saae1c70e168b45b4">
<source>Admin</source>
2023-06-12 13:41:44 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s364c4f177a2f8322">
<source>Open API drawer</source>
2023-06-12 13:41:44 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s9ba989e69344ff29">
<source>Open Notification drawer</source>
2023-06-12 13:41:44 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sfe629863ba1338c2">
<source>Connection error, reconnecting...</source>
2023-06-12 13:41:44 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s49730f3d5751a433">
<source>Loading...</source>
2023-06-12 13:41:44 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sf1e9d421f35b51e5">
<source>Application</source>
2023-06-12 13:41:44 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s310d8757ce319673">
<source>Logins</source>
2023-06-13 13:41:48 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s3e99ea082ca5ade9">
<source>Failed to fetch</source>
2023-06-13 13:41:48 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s09810653c832e935">
<source>Click to change value</source>
2023-06-20 10:09:13 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sfefce784ec55868f">
<source>Select an object.</source>
2023-06-20 10:09:13 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s04ceadb276bbe149">
<source>Loading options...</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s1cd198d689c66e4b">
<source>API Access</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sf29883ac9ec43085">
<source>App password</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6ac670086eb137c6">
<source>Recovery</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sfe211545fd02f73e">
<source>Verification</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sd73b202ec04eefd9">
<source>Unknown intent</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc8da3cc71de63832">
<source>Login</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb4564c127ab8b921">
<source>Failed login</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s67749057edb2586b">
<source>Logout</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7e537ad68d7c16e1">
<source>User was written to</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa0e0bdd7e244416b">
<source>Suspicious request</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7bda44013984fc48">
<source>Password set</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa1b41e334ad89d94">
<source>Secret was viewed</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s92ca679592a36b35">
<source>Secret was rotated</source>
2023-07-17 15:57:08 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s8a1d9403ca90989b">
<source>Invitation used</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s5f496533610103f2">
<source>Application authorized</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sdc9e222be9612939">
<source>Source linked</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb1c91762ae3a9bee">
<source>Impersonation started</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s9c73bd29b279d26b">
<source>Impersonation ended</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s1cd264012278c047">
<source>Flow execution</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s32f04d33924ce8ad">
<source>Policy execution</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb6d7128df5978cee">
<source>Policy exception</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s77f572257f69a8db">
<source>Property Mapping exception</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s2543cffd6ebb6803">
<source>System task execution</source>
2023-07-21 16:23:51 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="se2f258b996f7279c">
<source>System task exception</source>
2023-07-28 12:25:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s81eff3409d572a21">
<source>General system exception</source>
2023-07-28 12:25:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sf8f49cdbf0036343">
<source>Configuration error</source>
2023-07-28 12:25:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s9c6f61dc47bc4f0a">
<source>Model created</source>
2023-07-28 12:25:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s47a4983a2c6bb749">
<source>Model updated</source>
2023-07-28 12:25:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc9f69360b58706c7">
<source>Model deleted</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa266303caf1bd27f">
<source>Email sent</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6c410fedda2a575f">
<source>Update available</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sf1ec4acb8d744ed9">
<source>Alert</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s9117fb5195e75151">
<source>Notice</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s34be76c6b1eadbef">
<source>Warning</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s02240309358f557c">
<source>Unknown severity</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s858e7ac4b3cf955f">
<source>Static tokens</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sfcfcf85a57eea78a">
<source>TOTP Device</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="se2d65e13768468e0">
<source>Internal</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s84fcddede27b8e2a">
<source>External</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s1a635369edaf4dc3">
<source>Service account</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sff930bf2834e2201">
<source>Service account (internal)</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa50a6326530d8a0d">
<source>Show less</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb2c57b2d347203dd">
<source>Show more</source>
2023-07-31 17:35:09 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6238f519db67980d">
<source>UID</source>
2023-08-17 21:48:05 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sef49aec68fd1dc66">
<source>Name</source>
2023-08-17 21:48:05 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sf9f2c719a04066ec">
<source>App</source>
2023-08-17 21:48:05 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sda796c87fa97ed4d">
<source>Model Name</source>
2023-08-23 11:20:42 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s79e8cc71a5975b04">
<source>Message</source>
2023-08-23 11:20:42 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sbbc53e0e54d7946f">
<source>Subject</source>
2023-08-30 18:10:36 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa6ab5184d6315895">
<source>From</source>
2023-09-02 16:59:17 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s09353907b5c79284">
<source>To</source>
2023-09-02 16:59:17 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s63e03c70f67ebf9c">
<source>Context</source>
2023-09-02 16:59:17 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa48f81f001b893d2">
<source>User</source>
2023-09-02 16:59:17 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s119498d4e4cf59a6">
<source>Affected model:</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa3660d505e7011e0">
<source>Authorized application:</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s95a032ae86881bf5">
<source>Using flow</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="scb5c9a7cc4ccd68d">
<source>Email info:</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s677f1b675fc21bb1">
<source>Secret:</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa6905be242387f36">
<source>Exception</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sd947d57c9a9b7108">
<source>Open issue on GitHub...</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6ab73c998850c5ab">
<source>Expression</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s50ebe627b4bc7d02">
<source>Binding</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s3c6de3f257e0c912">
<source>Request</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s730182ad28374cda">
<source>Object</source>
2023-09-15 11:12:38 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s890e983a7be64da4">
<source>Result</source>
2023-09-25 10:43:24 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sd3a853f63f45dcb0">
<source>Passing</source>
2023-09-25 10:43:24 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sbdeedc1c60306b35">
<source>Messages</source>
2023-09-25 10:43:24 +00:00
</trans-unit>
2023-09-15 11:12:38 +00:00
<trans-unit id="s4ba4473f3d4ec896">
<source>New version available</source>
2023-09-25 10:43:24 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s0a5401d4419f9958">
<source>Using source</source>
2023-09-25 10:43:24 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s14622ee6de586485">
<source>Attempted to log in as <x id="0" equiv-text="${this.event.context.username}"/></source>
2023-09-25 10:43:24 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb07bf992e3d00664">
<source>No additional data available.</source>
2023-09-25 10:43:24 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sf45a0d2f00bcc6ff">
<source>no tabs defined</source>
2023-10-08 22:08:16 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s08df8d0a773a3ea0">
<source>Remove item</source>
2023-10-08 22:08:16 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s04c5a637328c9b67">
<source><x id="0" equiv-text="${this.pages?.startIndex}"/> - <x id="1" equiv-text="${this.pages?.endIndex}"/> of <x id="2" equiv-text="${this.pages?.count}"/></source>
2023-10-08 22:08:16 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6a89bb10338369b4">
<source>Go to previous page</source>
2023-10-12 12:00:45 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7edad99c6b7bfe88">
<source>Go to next page</source>
2023-10-12 13:12:46 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sffa721bb6aa3128d">
<source>Search...</source>
2023-10-12 19:33:45 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb59d68ed12d46377">
<source>Loading</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7bc8c327f1f7c82c">
<source>No objects found.</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sfd44ce578f643145">
<source>Failed to fetch objects.</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7b7163270e57e8b4">
<source>Refresh</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s909e876731a8febb">
<source>Select all rows</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa442044b586ec8bf">
<source>Action</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s02839b01844d6ca8">
<source>Creation Date</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s4d00f1de1c82281b">
<source>Client IP</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s744401846fea6e76">
<source>Brand</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s2152f3482784705f">
<source>Recent events</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc35581d9c1cd67ff">
<source>On behalf of <x id="0" equiv-text="${item.user.on_behalf_of.username}"/></source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="saf63a04c86018698">
<source>-</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s1b448a4ea79d4eef">
<source>No Events found.</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s50911ec1c8aee99a">
<source>No matching events could be found.</source>
2023-10-12 19:33:45 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s113c05ef9996ca4b">
<source>Embedded outpost is not configured correctly.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="seb5ba88f21937c98">
<source>Check outposts.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sd0bc94e11935ee5a">
<source>HTTPS is not detected correctly</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s40bf151b56a64f51">
<source>Server and client are further than 5 seconds apart.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s091d3d07b5b3076f">
<source>OK</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sae486938be80729c">
<source>Everything is ok.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sea91c57b3d3969fe">
<source>System status</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="scefe482c547fb3f3">
<source>Based on <x id="0" equiv-text="${value.versionCurrent}"/></source>
</trans-unit>
<trans-unit id="s68a50b1ee6efee7b">
<source><x id="0" equiv-text="${value.versionLatest}"/> is available!</source>
</trans-unit>
<trans-unit id="s713d147e1761d0f0">
<source>Up-to-date!</source>
</trans-unit>
<trans-unit id="sf4122b220926be97">
<source>Version</source>
</trans-unit>
<trans-unit id="s0a63a8be0b2b422c">
<source>Workers</source>
</trans-unit>
<trans-unit id="s341ab68d4130de20">
<source>No workers connected. Background tasks will not run.</source>
</trans-unit>
<trans-unit id="s2ed8eb02525a920a">
<source><x id="0" equiv-text="${ago}"/> hour(s) ago</source>
</trans-unit>
<trans-unit id="s98327528f00365a7">
<source>Failed to fetch data.</source>
</trans-unit>
<trans-unit id="s1f1c857c0c4250e4">
<source><x id="0" equiv-text="${ago}"/> day(s) ago</source>
</trans-unit>
<trans-unit id="s11bc220e8fa9d797">
<source>Authorizations</source>
</trans-unit>
<trans-unit id="s3ef3c252ada78076">
<source>Failed Logins</source>
</trans-unit>
<trans-unit id="sc2f1e5dd74c1b7df">
<source>Successful Logins</source>
</trans-unit>
<trans-unit id="s0382d73823585617">
<source><x id="0" equiv-text="${this.errorMessage}"/>: <x id="1" equiv-text="${e.toString()}"/></source>
</trans-unit>
<trans-unit id="s2ceb11be2290bb1b">
<source>Cancel</source>
</trans-unit>
<trans-unit id="se085f35c8a9203a1">
<source>LDAP Source</source>
</trans-unit>
<trans-unit id="s477de089b505a6ea">
<source>SCIM Provider</source>
</trans-unit>
<trans-unit id="s8a75e83497a183a2">
<source>Healthy</source>
</trans-unit>
<trans-unit id="sbf9c5c5a8e5efad4">
<source>Failed</source>
</trans-unit>
<trans-unit id="se4cd073c125382af">
<source>Unsynced / N/A</source>
</trans-unit>
<trans-unit id="sfeb82261bcf99edd">
<source>Healthy outposts</source>
</trans-unit>
<trans-unit id="s21b3058faf874368">
<source>Outdated outposts</source>
</trans-unit>
<trans-unit id="s51f92b6fa76656ca">
<source>Unhealthy outposts</source>
</trans-unit>
<trans-unit id="s0a11c2ffb8309d1a">
<source>Not found</source>
</trans-unit>
<trans-unit id="saa0e2675da69651b">
<source>The URL "<x id="0" equiv-text="${this.url}"/>" was not found.</source>
</trans-unit>
<trans-unit id="s58cd9c2fe836d9c6">
<source>Return home</source>
</trans-unit>
<trans-unit id="s41e035c4bb8d15f2">
<source>General system status</source>
</trans-unit>
<trans-unit id="s6dfd15978586d05f">
<source>Welcome, <x id="0" equiv-text="${name}"/>.</source>
</trans-unit>
<trans-unit id="sc381422c585b867f">
<source>Quick actions</source>
</trans-unit>
<trans-unit id="sfd13ca8ebd857c2e">
<source>Create a new application</source>
</trans-unit>
<trans-unit id="s079d388d3cbfa54f">
<source>Check the logs</source>
</trans-unit>
<trans-unit id="sed8d4c3fd5f60e1f">
<source>Explore integrations</source>
</trans-unit>
<trans-unit id="sfffb0d0958bfbc42">
<source>Manage users</source>
</trans-unit>
<trans-unit id="s66313b45b69cfc88">
<source>Check the release notes</source>
</trans-unit>
<trans-unit id="s8763a33c3d46aaf5">
<source>Outpost status</source>
</trans-unit>
<trans-unit id="scc286303aa9c6cb0">
<source>Sync status</source>
</trans-unit>
<trans-unit id="sbdc4a833de9ca502">
<source>Logins and authorizations over the last week (per 8 hours)</source>
</trans-unit>
<trans-unit id="s6e09a19aa3952509">
<source>Apps with most usage</source>
</trans-unit>
<trans-unit id="sda5e1499f93146ad">
<source><x id="0" equiv-text="${ago}"/> days ago</source>
</trans-unit>
<trans-unit id="s51ea3a244c781b1f">
<source>Objects created</source>
</trans-unit>
<trans-unit id="sb4d7bae2440d9781">
<source>User Statistics</source>
</trans-unit>
<trans-unit id="sfbadb77fbc61efb8">
<source>Users created per day in the last month</source>
</trans-unit>
<trans-unit id="s4a34a6be4c68ec87">
<source>Users created</source>
</trans-unit>
<trans-unit id="sb0669da3df95837c">
<source>Logins per day in the last month</source>
</trans-unit>
<trans-unit id="s835da49b4dc83a51">
<source>Failed Logins per day in the last month</source>
</trans-unit>
<trans-unit id="s275c956687e2e656">
<source>Failed logins</source>
</trans-unit>
<trans-unit id="s5f4586bc1e2740e6">
<source>Clear search</source>
</trans-unit>
<trans-unit id="s3b34d9930e33bd46">
<source>System Tasks</source>
</trans-unit>
<trans-unit id="saaa3abe03c7260f9">
<source>Long-running operations which authentik executes in the background.</source>
</trans-unit>
<trans-unit id="s7468e87263dfff7e">
<source>Identifier</source>
</trans-unit>
<trans-unit id="s63d894b1ddb06289">
<source>Description</source>
</trans-unit>
<trans-unit id="sa9b2a245441557dc">
<source>Last run</source>
</trans-unit>
<trans-unit id="sad3e3c8146fc920f">
<source>Status</source>
</trans-unit>
<trans-unit id="s8af61807443f32a4">
<source>Actions</source>
</trans-unit>
<trans-unit id="sbe9a51f29a4a2c5b">
<source>Successful</source>
</trans-unit>
<trans-unit id="s5f343a43e7ea9f91">
<source>Error</source>
</trans-unit>
<trans-unit id="sc592307ea80f16b9">
<source>Unknown</source>
</trans-unit>
<trans-unit id="s92921878e886e36d">
<source>Duration</source>
</trans-unit>
<trans-unit id="se7e1ababbc4868b8">
<source><x id="0" equiv-text="${item.taskDuration.toFixed(2)}"/> seconds</source>
</trans-unit>
<trans-unit id="s14bf17e2a1a2c381">
<source>Restart task</source>
</trans-unit>
<trans-unit id="s5e8250fb85d64c23">
<source>Close</source>
</trans-unit>
<trans-unit id="s382a2aa3984474dd">
<source>Create</source>
</trans-unit>
<trans-unit id="s0fbf6dc6a1966408">
<source>Next</source>
</trans-unit>
<trans-unit id="sc16e00a7a8b2fde2">
<source>Back</source>
</trans-unit>
<trans-unit id="sb3d4f79d9d8b71e5">
<source>Submit</source>
</trans-unit>
<trans-unit id="sd2223afb7d6b100d">
<source>Type</source>
</trans-unit>
<trans-unit id="s398f6ba74ba8943a">
<source>Select providers to add to application</source>
</trans-unit>
<trans-unit id="sf9aee319a006c9b4">
<source>Add</source>
</trans-unit>
<trans-unit id="sa90b7809586c35ce">
<source>Either input a full URL, a relative path, or use 'fa://fa-test' to use the Font Awesome icon "fa-test".</source>
</trans-unit>
<trans-unit id="s0410779cb47de312">
<source>Path template for users created. Use placeholders like `%(slug)s` to insert the source slug.</source>
</trans-unit>
<trans-unit id="s67e20cd8018d7e3c">
<source>Currently set to:</source>
</trans-unit>
<trans-unit id="sd5903cc8de68b3fc">
<source>No form found</source>
</trans-unit>
<trans-unit id="s45935843b1b5b496">
<source>Form didn't return a promise for submitting</source>
</trans-unit>
<trans-unit id="s1a0e95458b44d7f8">
<source>Any policy must match to grant access</source>
</trans-unit>
<trans-unit id="s7fc1ace65486dc25">
<source>All policies must match to grant access</source>
</trans-unit>
<trans-unit id="s58fd2aafa4261c55">
<source>Successfully updated application.</source>
</trans-unit>
<trans-unit id="s9222ca30ae7786e4">
<source>Successfully created application.</source>
</trans-unit>
<trans-unit id="s03907d7a66c6164e">
<source>Application's display Name.</source>
</trans-unit>
<trans-unit id="s91f70424f5d5d23e">
<source>Slug</source>
</trans-unit>
<trans-unit id="s1cffe58249b04669">
<source>Internal application name used in URLs.</source>
</trans-unit>
<trans-unit id="s98b1cb8fb62909ec">
<source>Group</source>
</trans-unit>
<trans-unit id="sdae55084f6cb2662">
<source>Optionally enter a group name. Applications with identical groups are shown grouped together.</source>
</trans-unit>
<trans-unit id="s7f5869b3d14d7cbc">
<source>Provider</source>
</trans-unit>
<trans-unit id="s350a616ff5e145ec">
<source>Select a provider that this application should use.</source>
</trans-unit>
<trans-unit id="sc92ea8fbf9ba06a7">
<source>Backchannel Providers</source>
</trans-unit>
<trans-unit id="s4c6534a118f52fdd">
<source>Select backchannel providers which augment the functionality of the main provider.</source>
</trans-unit>
<trans-unit id="s19409e8712ddd369">
<source>Add provider</source>
</trans-unit>
<trans-unit id="s0639662111324466">
<source>Policy engine mode</source>
</trans-unit>
<trans-unit id="s8be4abc7ca71da6c">
<source>UI settings</source>
</trans-unit>
<trans-unit id="s427f788ff333f45b">
<source>Launch URL</source>
</trans-unit>
<trans-unit id="s992f8d1a776e763c">
<source>If left empty, authentik will try to extract the launch URL based on the selected provider.</source>
</trans-unit>
<trans-unit id="s2348f46ebf436671">
<source>Open in new tab</source>
</trans-unit>
<trans-unit id="s8655c52824caac63">
<source>If checked, the launch URL will open in a new browser tab or window from the user's application library.</source>
</trans-unit>
<trans-unit id="s068d4dd16d9106d0">
<source>Icon</source>
</trans-unit>
<trans-unit id="s80e6d6fe5ad458d3">
<source>Clear icon</source>
</trans-unit>
<trans-unit id="s485c05d34eb00415">
<source>Delete currently set icon.</source>
</trans-unit>
<trans-unit id="s6d3b4d0561ba1cff">
<source>Publisher</source>
</trans-unit>
<trans-unit id="se2b29e6cfe59414c">
<source>UI Settings</source>
</trans-unit>
<trans-unit id="s4498e890d47a8066">
<source>OAuth2/OIDC (Open Authorization/OpenID Connect)</source>
</trans-unit>
<trans-unit id="sc2cedfb22488ccb2">
<source>Modern applications, APIs and Single-page applications.</source>
</trans-unit>
<trans-unit id="s4f2e195d09e2868c">
<source>LDAP (Lightweight Directory Access Protocol)</source>
</trans-unit>
<trans-unit id="sffd5481034a1bd41">
<source>Provide an LDAP interface for applications and users to authenticate against.</source>
</trans-unit>
<trans-unit id="s836148f721d8913b">
<source>Transparent Reverse Proxy</source>
</trans-unit>
<trans-unit id="s945a6b94361ee45b">
<source>For transparent reverse proxies with required authentication</source>
</trans-unit>
<trans-unit id="s7f5bb0c9923315ed">
<source>Forward Auth (Single Application)</source>
</trans-unit>
<trans-unit id="sdc9a6ad1af30572c">
<source>For nginx's auth_request or traefik's forwardAuth</source>
</trans-unit>
<trans-unit id="sf8008d2d6b064b95">
<source>Forward Auth (Domain Level)</source>
</trans-unit>
<trans-unit id="sfc31264ef7ff86ef">
<source>For nginx's auth_request or traefik's forwardAuth per root domain</source>
</trans-unit>
<trans-unit id="sfa8a1ffa9fee07d3">
<source>SAML (Security Assertion Markup Language)</source>
</trans-unit>
<trans-unit id="s40830ec037f34626">
<source>Configure SAML provider manually</source>
</trans-unit>
<trans-unit id="s848a23972e388662">
<source>RADIUS (Remote Authentication Dial-In User Service)</source>
</trans-unit>
<trans-unit id="sea9fc40dfd1d18b1">
<source>Configure RADIUS provider manually</source>
</trans-unit>
<trans-unit id="s3e902999ddf7b50e">
<source>SCIM (System for Cross-domain Identity Management)</source>
</trans-unit>
<trans-unit id="sa1b0052ae095b9b3">
<source>Configure SCIM provider manually</source>
</trans-unit>
<trans-unit id="s15831fa50a116545">
<source>Saving Application...</source>
</trans-unit>
<trans-unit id="s823abdb61543a826">
<source>Authentik was unable to save this application:</source>
</trans-unit>
<trans-unit id="s848288f8c2265aad">
<source>Your application has been saved</source>
</trans-unit>
<trans-unit id="scda8dc24b561e205">
<source>There was an error in the application.</source>
</trans-unit>
<trans-unit id="sdaca9c2c0361ed3a">
<source>Review the application.</source>
</trans-unit>
<trans-unit id="sb50000a8fada5672">
<source>There was an error in the provider.</source>
</trans-unit>
<trans-unit id="s21f95eaf151d4ce3">
<source>Review the provider.</source>
</trans-unit>
<trans-unit id="s9fd39a5cb20b4e61">
<source>There was an error</source>
</trans-unit>
<trans-unit id="s7a6b3453209e1066">
<source>There was an error creating the application, but no error message was sent. Please review the server logs.</source>
</trans-unit>
<trans-unit id="sc25edca57df81461">
<source>Authentication</source>
</trans-unit>
<trans-unit id="s6dfb7283452f78fe">
<source>Authorization</source>
</trans-unit>
<trans-unit id="sddcfc6ab24e3a6ed">
<source>Enrollment</source>
</trans-unit>
<trans-unit id="s1fc9c70610c4c67d">
<source>Invalidation</source>
</trans-unit>
<trans-unit id="sdf22dcf939c27cc7">
<source>Stage Configuration</source>
</trans-unit>
<trans-unit id="s6d5bce4321f57cda">
<source>Unenrollment</source>
</trans-unit>
<trans-unit id="sde2bb5418562c5b2">
<source>Unknown designation</source>
</trans-unit>
<trans-unit id="sb9834316ffd4ae3e">
<source>Stacked</source>
</trans-unit>
<trans-unit id="s12146091b2b539a3">
<source>Content left</source>
</trans-unit>
<trans-unit id="sa800871782eba1ac">
<source>Content right</source>
</trans-unit>
<trans-unit id="sb4e50ca3cffdbc10">
<source>Sidebar left</source>
</trans-unit>
<trans-unit id="s745a55f9abf9f2e5">
<source>Sidebar right</source>
</trans-unit>
<trans-unit id="sb3182a87ded1bc91">
<source>Unknown layout</source>
</trans-unit>
<trans-unit id="s8915e64b8b999bfe">
<source>Cached binding</source>
</trans-unit>
<trans-unit id="s842d690eb3c11762">
<source>Flow is executed and session is cached in memory. Flow is executed when session expires</source>
</trans-unit>
<trans-unit id="s6a66759749bf31ed">
<source>Direct binding</source>
</trans-unit>
<trans-unit id="se0adaf83627104fb">
<source>Always execute the configured bind flow to authenticate the user</source>
</trans-unit>
<trans-unit id="s9065fcccd837a679">
<source>Cached querying</source>
</trans-unit>
<trans-unit id="s30d0d0e6c626a234">
<source>The outpost holds all users and groups in-memory and will refresh every 5 Minutes</source>
</trans-unit>
<trans-unit id="sffc14b8200a9f938">
<source>Direct querying</source>
</trans-unit>
<trans-unit id="sdce4680288083fe3">
<source>Always returns the latest data, but slower than cached querying</source>
</trans-unit>
<trans-unit id="s1889ba2eaeec2f1e">
<source>When enabled, code-based multi-factor authentication can be used by appending a semicolon and the TOTP code to the password. This should only be enabled if all users that will bind to this provider have a TOTP device configured, as otherwise a password may incorrectly be rejected if it contains a semicolon.</source>
</trans-unit>
<trans-unit id="s1c8e9816dcae6d9c">
<source>The start for gidNumbers, this number is added to a number generated from the group.Pk to make sure that the numbers aren't too low for POSIX groups. Default is 4000 to ensure that we don't collide with local groups or users primary groups gidNumber</source>
</trans-unit>
<trans-unit id="s4647b2c92638d6fd">
<source>The certificate for the above configured Base DN. As a fallback, the provider uses a self-signed certificate.</source>
</trans-unit>
<trans-unit id="s2acef4f6ba39bf11">
<source>DNS name for which the above configured certificate should be used. The certificate cannot be detected based on the base DN, as the SSL/TLS negotiation happens before such data is exchanged.</source>
</trans-unit>
<trans-unit id="s60edbcfac8ed1f90">
<source>The start for uidNumbers, this number is added to the user.Pk to make sure that the numbers aren't too low for POSIX users. Default is 2000 to ensure that we don't collide with local users uidNumber</source>
</trans-unit>
<trans-unit id="s1a711c19cda48375">
<source>Configure LDAP Provider</source>
</trans-unit>
<trans-unit id="s67d858051b34c38b">
<source>Method's display Name.</source>
</trans-unit>
<trans-unit id="sff69c1a637f899a6">
<source>Bind flow</source>
</trans-unit>
<trans-unit id="s319040353f479853">
<source>Flow used for users to authenticate.</source>
</trans-unit>
<trans-unit id="sbc80eab557fbf782">
<source>Search group</source>
</trans-unit>
<trans-unit id="se5973e7c8ba0fc71">
<source>Bind mode</source>
</trans-unit>
<trans-unit id="scef3f4ad80abbd22">
<source>Configure how the outpost authenticates requests.</source>
</trans-unit>
<trans-unit id="sbcae51a6f06e53d4">
<source>Search mode</source>
</trans-unit>
<trans-unit id="s8b87df5664de7eb8">
<source>Configure how the outpost queries the core authentik server's users.</source>
</trans-unit>
<trans-unit id="s254d527e3a53dbb7">
<source>Code-based MFA Support</source>
</trans-unit>
<trans-unit id="sfe388f0313f52da2">
<source>Protocol settings</source>
</trans-unit>
<trans-unit id="s55d731be1ef66efe">
<source>Base DN</source>
</trans-unit>
<trans-unit id="s0b15ff11a0049cfd">
<source>LDAP DN under which bind requests and search requests can be made.</source>
</trans-unit>
<trans-unit id="sb157267c85fdff30">
<source>Certificate</source>
</trans-unit>
<trans-unit id="scd247ffad6e04ac0">
<source>TLS Server name</source>
</trans-unit>
<trans-unit id="sac43cb9690260b86">
<source>UID start number</source>
</trans-unit>
<trans-unit id="s5acb607b40356974">
<source>GID start number</source>
</trans-unit>
<trans-unit id="sdfd22a21660f6002">
<source>Successfully updated provider.</source>
</trans-unit>
<trans-unit id="s457c639088c547c5">
<source>Successfully created provider.</source>
</trans-unit>
<trans-unit id="s2236dc563c2dbf76">
<source>(Format: hours=-1;minutes=-2;seconds=-3).</source>
</trans-unit>
<trans-unit id="sbec40ef4e6f139b7">
<source>(Format: hours=1;minutes=2;seconds=3).</source>
</trans-unit>
<trans-unit id="sbb8ad22c83d375b1">
<source>The following keywords are supported:</source>
</trans-unit>
<trans-unit id="s399cc2d67d92e957">
<source>Confidential</source>
</trans-unit>
<trans-unit id="s95f09b229a0a0bb0">
<source>Confidential clients are capable of maintaining the confidentiality of their credentials such as client secrets</source>
</trans-unit>
<trans-unit id="sdd1ff479d04ac140">
<source>Public</source>
</trans-unit>
<trans-unit id="s51c6b8403c2dc5d9">
<source>Public clients are incapable of maintaining the confidentiality and should use methods like PKCE. </source>
</trans-unit>
<trans-unit id="sccc47f82044453f9">
<source>Based on the User's hashed ID</source>
</trans-unit>
<trans-unit id="sbd5be4fb7442a34c">
<source>Based on the User's ID</source>
</trans-unit>
<trans-unit id="sc9cf9ecaf9e5d67e">
<source>Based on the User's UUID</source>
</trans-unit>
<trans-unit id="s4291727352c4f295">
<source>Based on the User's username</source>
</trans-unit>
<trans-unit id="sd62cfc27ad4aa33b">
<source>Based on the User's Email</source>
</trans-unit>
<trans-unit id="s55eb75bedf96be0f">
<source>This is recommended over the UPN mode.</source>
</trans-unit>
<trans-unit id="sf80e9547166117e6">
<source>Based on the User's UPN</source>
</trans-unit>
<trans-unit id="sde949d0ef44572eb">
<source>Requires the user to have a 'upn' attribute set, and falls back to hashed user ID. Use this mode only if you have different UPN and Mail domains.</source>
</trans-unit>
<trans-unit id="sde56783222b527d6">
<source>Each provider has a different issuer, based on the application slug</source>
</trans-unit>
<trans-unit id="s8d32d7b9e8ca60b1">
<source>Same identifier is used for all providers</source>
</trans-unit>
<trans-unit id="s7f9eb9c8bd26e8fd">
<source>Valid redirect URLs after a successful authorization flow. Also specify any origins here for Implicit flows.</source>
</trans-unit>
<trans-unit id="s2a369bc2febb5d55">
<source>If no explicit redirect URIs are specified, the first successfully used redirect URI will be saved.</source>
</trans-unit>
<trans-unit id="sa8384c9c26731f83">
<source>To allow any redirect URI, set this value to ".*". Be aware of the possible security implications this can have.</source>
</trans-unit>
<trans-unit id="sbb3243352661428f">
<source>Authentication flow</source>
</trans-unit>
<trans-unit id="sa72a3bd1e7e89926">
<source>Flow used when a user access this provider and is not authenticated.</source>
</trans-unit>
<trans-unit id="s62f7c59b0606a8d6">
<source>Authorization flow</source>
</trans-unit>
<trans-unit id="sfbaeb0de54fbfdbb">
<source>Flow used when authorizing this provider.</source>
</trans-unit>
<trans-unit id="sc8de93a7dc0d78ba">
<source>Client type</source>
</trans-unit>
<trans-unit id="s4d00e5de1c8213b7">
<source>Client ID</source>
</trans-unit>
<trans-unit id="s03fb3fa232f0434a">
<source>Client Secret</source>
</trans-unit>
<trans-unit id="sde0ad51b14f77cf6">
<source>Redirect URIs/Origins (RegEx)</source>
</trans-unit>
<trans-unit id="s55787f4dfcdce52b">
<source>Signing Key</source>
</trans-unit>
<trans-unit id="sc6c57419ad3a01a8">
<source>Key used to sign the tokens.</source>
</trans-unit>
<trans-unit id="s124f93a61ee772d6">
<source>Advanced protocol settings</source>
</trans-unit>
<trans-unit id="s926e0ecf124fb01a">
<source>Access code validity</source>
</trans-unit>
<trans-unit id="sa578033f134a83b6">
<source>Configure how long access codes are valid for.</source>
</trans-unit>
<trans-unit id="sbea3db12fd799210">
<source>Access Token validity</source>
</trans-unit>
<trans-unit id="s72559845d38bf688">
<source>Configure how long access tokens are valid for.</source>
</trans-unit>
<trans-unit id="s821f6014c1a435b9">
<source>Refresh Token validity</source>
</trans-unit>
<trans-unit id="s00c2db16ea9bc263">
<source>Configure how long refresh tokens are valid for.</source>
</trans-unit>
<trans-unit id="s2e3ef41a0edd8608">
<source>Scopes</source>
</trans-unit>
<trans-unit id="s3a3fae99373ce56b">
<source>Select which scopes can be used by the client. The client still has to specify the scope to access the data.</source>
</trans-unit>
<trans-unit id="sffd2e553143d1b0e">
<source>Hold control/command to select multiple items.</source>
</trans-unit>
<trans-unit id="s26bf2730430efbea">
<source>Subject mode</source>
</trans-unit>
<trans-unit id="s9f23ed1799b4d49a">
<source>Configure what data should be used as unique User Identifier. For most cases, the default should be fine.</source>
</trans-unit>
<trans-unit id="s17d1e337f6c11c1e">
<source>Include claims in id_token</source>
</trans-unit>
<trans-unit id="sbf41e0db12834133">
<source>Include User claims from scopes in the id_token, for applications that don't access the userinfo endpoint.</source>
</trans-unit>
<trans-unit id="s850a58c683682809">
<source>Issuer mode</source>
</trans-unit>
<trans-unit id="s37d9155b9f4cc7bd">
<source>Configure how the issuer field of the ID Token should be filled.</source>
</trans-unit>
<trans-unit id="se2adaf0371ffcd65">
<source>Machine-to-Machine authentication settings</source>
</trans-unit>
<trans-unit id="s33318837e6c54a9b">
<source>Trusted OIDC Sources</source>
</trans-unit>
<trans-unit id="s22e566052f7bec81">
<source>JWTs signed by certificates configured in the selected sources can be used to authenticate to this provider.</source>
</trans-unit>
<trans-unit id="s9368e965b5c292ab">
<source>Configure OAuth2/OpenId Provider</source>
</trans-unit>
<trans-unit id="s072c6d12d3d37501">
<source>HTTP-Basic Username Key</source>
</trans-unit>
<trans-unit id="sb2bb6f93773a4594">
<source>User/Group Attribute used for the user part of the HTTP-Basic Header. If not set, the user's Email address is used.</source>
</trans-unit>
<trans-unit id="s70f6471de355b98c">
<source>HTTP-Basic Password Key</source>
</trans-unit>
<trans-unit id="sf4de1644dcdb53d5">
<source>User/Group Attribute used for the password part of the HTTP-Basic Header.</source>
</trans-unit>
<trans-unit id="sf5cbccdc6254c8dc">
<source>Configure Proxy Provider</source>
</trans-unit>
<trans-unit id="s7c10976de6411844">
<source>Token validity</source>
</trans-unit>
<trans-unit id="s3e87ce98ba3c4d80">
<source>Configure how long tokens are valid for.</source>
</trans-unit>
<trans-unit id="sf6d46bb442b77e91">
<source>AdditionalScopes</source>
</trans-unit>
<trans-unit id="s8f12575f694e85a2">
<source>Additional scope mappings, which are passed to the proxy.</source>
</trans-unit>
<trans-unit id="s93cea6ca1f93349d">
<source>Unauthenticated URLs</source>
</trans-unit>
<trans-unit id="sc4508175bf6b09dd">
<source>Unauthenticated Paths</source>
</trans-unit>
<trans-unit id="sc9fc206433f67588">
<source>Regular expressions for which authentication is not required. Each new line is interpreted as a new expression.</source>
</trans-unit>
<trans-unit id="sd503fabef9691134">
<source>When using proxy or forward auth (single application) mode, the requested URL Path is checked against the regular expressions. When using forward auth (domain mode), the full requested URL including scheme and host is matched against the regular expressions.</source>
</trans-unit>
<trans-unit id="sb488dee0be434f7e">
<source>Authentication settings</source>
</trans-unit>
<trans-unit id="s23cee624c735f266">
<source>Intercept header authentication</source>
</trans-unit>
<trans-unit id="sc007cca5af67eae0">
<source>When enabled, authentik will intercept the Authorization header to authenticate the request.</source>
</trans-unit>
<trans-unit id="s36e630ba56617556">
<source>Send HTTP-Basic Authentication</source>
</trans-unit>
<trans-unit id="s9d5796a4b9b7560e">
<source>Send a custom HTTP-Basic Authentication header based on values from authentik.</source>
</trans-unit>
<trans-unit id="s44c90273f08fb718">
<source>Use this provider with nginx's auth_request or traefik's forwardAuth. Only a single provider is required per root domain. You can't do per-application authorization, but you don't have to create a provider for each application.</source>
</trans-unit>
<trans-unit id="sf55d28d4dff0e41b">
<source>An example setup can look like this:</source>
</trans-unit>
<trans-unit id="sb4a1d1c19438e929">
<source>authentik running on auth.example.com</source>
</trans-unit>
<trans-unit id="s68f935c9ca792016">
<source>app1 running on app1.example.com</source>
</trans-unit>
<trans-unit id="sf813a72d8fadd765">
<source>In this case, you'd set the Authentication URL to auth.example.com and Cookie domain to example.com.</source>
</trans-unit>
<trans-unit id="sa29b5680cfafacc8">
<source>External host</source>
</trans-unit>
<trans-unit id="sa03fe48e892df2d8">
<source>The external URL you'll authenticate at. The authentik core server should be reachable under this URL.</source>
</trans-unit>
<trans-unit id="s7def067ed3ad3ad9">
<source>Cookie domain</source>
</trans-unit>
<trans-unit id="s211b75e868072162">
<source>Set this to the domain you wish the authentication to be valid for. Must be a parent domain of the URL above. If you're running applications as app1.domain.tld, app2.domain.tld, set this to 'domain.tld'.</source>
</trans-unit>
<trans-unit id="s93574c03953f25dd">
<source>This provider will behave like a transparent reverse-proxy, except requests must be authenticated. If your upstream application uses HTTPS, make sure to connect to the outpost using HTTPS as well.</source>
</trans-unit>
<trans-unit id="s764bccb30868bf62">
<source>The external URL you'll access the application at. Include any non-standard port.</source>
</trans-unit>
<trans-unit id="scb317851cbcc6b12">
<source>Internal host</source>
</trans-unit>
<trans-unit id="sf05e384059a0a7c1">
<source>Upstream host that the requests are forwarded to.</source>
</trans-unit>
<trans-unit id="s3d34068a31cab30b">
<source>Internal host SSL Validation</source>
</trans-unit>
<trans-unit id="s4a26798e1c3c37dd">
<source>Validate SSL Certificates of upstream servers.</source>
</trans-unit>
<trans-unit id="h10ef80d434185070">
<source>Use this provider with nginx's <x id="0" equiv-text="<code>"/>auth_request<x id="1" equiv-text="</code>"/> or traefik's
<x id="2" equiv-text="<code>"/>forwardAuth<x id="3" equiv-text="</code>"/>. Each application/domain needs its own provider.
Additionally, on each domain, <x id="4" equiv-text="<code>"/>/outpost.goauthentik.io<x id="5" equiv-text="</code>"/> must be
routed to the outpost (when using a managed outpost, this is done for you).</source>
</trans-unit>
<trans-unit id="s2c8c6f89089b31d4">
<source>Configure Radius Provider</source>
</trans-unit>
<trans-unit id="sb95baab425322600">
<source>Shared secret</source>
</trans-unit>
<trans-unit id="s9e9316a6b0c16231">
<source>Client Networks</source>
</trans-unit>
<trans-unit id="s7f2dcf01f7a8c0b7">
<source>List of CIDRs (comma-seperated) that clients can connect from. A more specific
CIDR will match before a looser one. Clients connecting from a non-specified CIDR
will be dropped.</source>
</trans-unit>
<trans-unit id="sd8f220c999726151">
<source>Redirect</source>
</trans-unit>
<trans-unit id="sb357ea19a722d827">
<source>Post</source>
</trans-unit>
<trans-unit id="sfe906cde5dddc041">
<source>Configure SAML Provider</source>
</trans-unit>
<trans-unit id="s11204eeb1e27ea8f">
<source>ACS URL</source>
</trans-unit>
<trans-unit id="sb7a30abc1dcf6c36">
<source>Issuer</source>
</trans-unit>
<trans-unit id="sf54c562d8a10ce77">
<source>Also known as EntityID.</source>
</trans-unit>
<trans-unit id="s991b750e2d5c4234">
<source>Service Provider Binding</source>
</trans-unit>
<trans-unit id="s4e28e2899e08a5f8">
<source>Determines how authentik sends the response back to the Service Provider.</source>
</trans-unit>
<trans-unit id="sd5a4b41c6c883b03">
<source>Audience</source>
</trans-unit>
<trans-unit id="sc741d9ebe07ad103">
<source>Signing Certificate</source>
</trans-unit>
<trans-unit id="sd6c3ddb62de0e8f7">
<source>Certificate used to sign outgoing Responses going to the Service Provider.</source>
</trans-unit>
<trans-unit id="s5be3b0567172e415">
<source>Verification Certificate</source>
</trans-unit>
<trans-unit id="s7c27e113f90a89e0">
<source>When selected, incoming assertion's Signatures will be validated against this certificate. To allow unsigned Requests, leave on default.</source>
</trans-unit>
<trans-unit id="sce106606ae84d46f">
<source>Property Mappings</source>
</trans-unit>
<trans-unit id="sb3defbacd01ad972">
<source>Property mappings used for user mapping.</source>
</trans-unit>
<trans-unit id="s1a2797874b7fe852">
<source>NameID Property Mapping</source>
</trans-unit>
<trans-unit id="s256b8452664ccae4">
<source>Configure how the NameID value will be created. When left empty, the NameIDPolicy of the incoming request will be respected.</source>
</trans-unit>
<trans-unit id="s9f91cc8bcfabb40f">
<source>Assertion valid not before</source>
</trans-unit>
<trans-unit id="s733f83ff9d50da30">
<source>Configure the maximum allowed time drift for an assertion.</source>
</trans-unit>
<trans-unit id="s2af5754090898640">
<source>Assertion valid not on or after</source>
</trans-unit>
<trans-unit id="s43c1f927936f0a02">
<source>Assertion not valid on or after current time + this value.</source>
</trans-unit>
<trans-unit id="sad8550b8731518d8">
<source>Session valid not on or after</source>
</trans-unit>
<trans-unit id="s0dd00fbaba08748a">
<source>Session not valid on or after current time + this value.</source>
</trans-unit>
<trans-unit id="s2a0f60e74b478804">
<source>Digest algorithm</source>
</trans-unit>
<trans-unit id="s693d975d38ff0214">
<source>Signature algorithm</source>
</trans-unit>
<trans-unit id="s7ccce0ec8d228db6">
<source>Configure SCIM Provider</source>
</trans-unit>
<trans-unit id="s61eacb19db252f5e">
<source>URL</source>
</trans-unit>
<trans-unit id="sb21f33b039c86322">
<source>SCIM base url, usually ends in /v2.</source>
</trans-unit>
<trans-unit id="se68398e3c2c760b2">
<source>Token</source>
</trans-unit>
<trans-unit id="s33ed903c210a6209">
<source>Token to authenticate with. Currently only bearer authentication is supported.</source>
</trans-unit>
<trans-unit id="sfc8bb104e2c05af8">
<source>User filtering</source>
</trans-unit>
<trans-unit id="sc0d0890fbd46ef62">
<source>Exclude service accounts</source>
</trans-unit>
<trans-unit id="s23ab136ad85f0ad2">
<source>Only sync users within the selected group.</source>
</trans-unit>
<trans-unit id="sfdedc3b0b2b7ce3d">
<source>Attribute mapping</source>
</trans-unit>
<trans-unit id="saf794c74c9ea731e">
<source>User Property Mappings</source>
</trans-unit>
<trans-unit id="s7cb9aa9ee1783f00">
<source>Group Property Mappings</source>
</trans-unit>
<trans-unit id="sd7728d2b6e1d25e9">
<source>Property mappings used for group creation.</source>
</trans-unit>
<trans-unit id="s4bd386db7302bb22">
<source>Create With Wizard</source>
</trans-unit>
<trans-unit id="s0c9670f429e74283">
<source>New application</source>
</trans-unit>
<trans-unit id="s71c5d51d5a357dbd">
<source>Don't show this message again.</source>
</trans-unit>
<trans-unit id="s070fdfb03034ca9b">
<source>One hint, 'New Application Wizard', is currently hidden</source>
</trans-unit>
<trans-unit id="s04b7f8d6aaef3756">
<source>Users in the selected group can do search queries. If no group is selected, no LDAP Searches are allowed.</source>
</trans-unit>
<trans-unit id="sb8dd788adf7b907b">
<source>Proxy</source>
</trans-unit>
<trans-unit id="s7489f76224f8120d">
<source>Forward auth (single application)</source>
</trans-unit>
<trans-unit id="s25d0cd75377daf75">
<source>Forward auth (domain level)</source>
</trans-unit>
<trans-unit id="sff0ac1ace2d90709">
<source>Use this provider with nginx's auth_request or traefik's forwardAuth. Each application/domain needs its own provider. Additionally, on each domain, /outpost.goauthentik.io must be routed to the outpost (when using a managed outpost, this is done for you).</source>
</trans-unit>
<trans-unit id="s31d15c6f16951464">
<source>Authentication URL</source>
</trans-unit>
<trans-unit id="s2345170f7e272668">
<source>Unknown proxy mode</source>
</trans-unit>
<trans-unit id="sd539548ca4c71619">
<source>Additional scopes</source>
</trans-unit>
<trans-unit id="se6d950402810c34f">
<source>Property mappings</source>
</trans-unit>
<trans-unit id="scb58b8a60cad8762">
<source>Default relay state</source>
</trans-unit>
<trans-unit id="s6827a456c9dfc6ee">
<source>When using IDP-initiated logins, the relay state will be set to this value.</source>
</trans-unit>
<trans-unit id="sd1a5560fde6f2271">
<source>Successfully imported provider.</source>
</trans-unit>
<trans-unit id="s252a52330d32b900">
<source>Metadata</source>
</trans-unit>
<trans-unit id="s7181a5504472e856">
<source>Apply changes</source>
</trans-unit>
<trans-unit id="sad59707375956ad2">
<source>Finish</source>
</trans-unit>
<trans-unit id="s74475586afc1fb0f">
<source>Select type</source>
</trans-unit>
<trans-unit id="s0b3bf19b31dd6bac">
<source>Try the new application wizard</source>
</trans-unit>
<trans-unit id="sa18e1c6e0e6f16cc">
<source>The new application wizard greatly simplifies the steps required to create applications and providers.</source>
</trans-unit>
<trans-unit id="s01ef54f5d7c6ed47">
<source>Try it now</source>
</trans-unit>
<trans-unit id="s58d1eb482059da12">
<source>New provider</source>
</trans-unit>
<trans-unit id="sa661ea7d7a50f2e9">
<source>Create a new provider.</source>
</trans-unit>
<trans-unit id="s5d6af4c100ad321b">
<source>Create <x id="0" equiv-text="${type.name}"/></source>
</trans-unit>
<trans-unit id="s019555b5a442aa00">
<source>Property mappings used to user mapping.</source>
</trans-unit>
<trans-unit id="sa319e3bf44c85963">
<source>Property mappings used to group creation.</source>
</trans-unit>
<trans-unit id="se09ab93d69f7f45b">
<source>Not used by any other object.</source>
</trans-unit>
<trans-unit id="s10922bd0ac765562">
<source>object will be DELETED</source>
</trans-unit>
<trans-unit id="sf3981f36525b0dbd">
<source>connection will be deleted</source>
</trans-unit>
<trans-unit id="s93cf77a59db53395">
<source>reference will be reset to default value</source>
</trans-unit>
<trans-unit id="s3e211d29fa10f843">
<source>reference will be set to an empty value</source>
</trans-unit>
<trans-unit id="s55fa598b754cc3cc">
<source><x id="0" equiv-text="${ub.name}"/> (<x id="1" equiv-text="${consequence}"/>)</source>
</trans-unit>
<trans-unit id="s09240e07b5b8d640">
<source>ID</source>
</trans-unit>
<trans-unit id="se33b158a1ec02a09">
<source>Successfully deleted <x id="0" equiv-text="${this.objects.length} ${this.objectLabel}"/></source>
</trans-unit>
<trans-unit id="sf6eb148db23d19de">
<source>Failed to delete <x id="0" equiv-text="${this.objectLabel}"/>: <x id="1" equiv-text="${e.toString()}"/></source>
</trans-unit>
<trans-unit id="s039b6434e8a75560">
<source>Delete <x id="0" equiv-text="${this.objectLabel}"/></source>
</trans-unit>
<trans-unit id="s5819a49638f6d7cb">
<source>Are you sure you want to delete <x id="0" equiv-text="${this.objects.length} ${this.objectLabel}"/>?</source>
</trans-unit>
<trans-unit id="sdc673e73b5c13aea">
<source>Delete</source>
</trans-unit>
<trans-unit id="sb0b86b8ca6ab13bd">
<source>Providers</source>
</trans-unit>
<trans-unit id="s3ffa320128991a45">
<source>Provide support for protocols like SAML and OAuth to assigned applications.</source>
</trans-unit>
<trans-unit id="s10929ca568ae10bc">
<source>Provider(s)</source>
</trans-unit>
<trans-unit id="sb2b3b281954752c4">
<source>Assigned to application </source>
</trans-unit>
<trans-unit id="sa6c0ba4910c7ad7f">
<source>Assigned to application (backchannel) </source>
</trans-unit>
<trans-unit id="s97f5e0c138eae172">
<source>Warning: Provider not assigned to any application.</source>
</trans-unit>
<trans-unit id="s8b0432eecbd8b034">
<source>Update</source>
</trans-unit>
<trans-unit id="sc9175cb129fdc306">
<source>Update <x id="0" equiv-text="${item.verboseName}"/></source>
</trans-unit>
<trans-unit id="s64ef2a6c2dd1d3d1">
<source>Edit</source>
</trans-unit>
<trans-unit id="sa8c45b6b92a8ba1f">
<source>Create Application</source>
</trans-unit>
<trans-unit id="sf485014051ad0cf7">
<source>Successfully assigned permission.</source>
</trans-unit>
<trans-unit id="sca7fed2bef53cb99">
<source>Role</source>
</trans-unit>
<trans-unit id="sc92c1a54034e21cc">
<source>Assign</source>
</trans-unit>
<trans-unit id="scd84d10ee9137070">
<source>Assign permission to role</source>
</trans-unit>
<trans-unit id="s5ee6f1b84e9ebc69">
<source>Assign to new role</source>
</trans-unit>
<trans-unit id="sb923723d27df40ba">
<source>Permission(s)</source>
</trans-unit>
<trans-unit id="sc5fb00b25c7f5a02">
<source>Permission</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
<trans-unit id="s4afb26a8fae257e9">
<source>Directly assigned</source>
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sd8051c26e155f043">
<source>Assign permission to user</source>
</trans-unit>
<trans-unit id="sf79f8681e5ffaee2">
<source>Assign to new user</source>
</trans-unit>
<trans-unit id="s9fb28be12e2c6317">
<source>Superuser</source>
</trans-unit>
<trans-unit id="sc615309d10a9228c">
<source>RBAC is in preview.</source>
</trans-unit>
<trans-unit id="sa1db89262360550b">
<source>Send us feedback!</source>
</trans-unit>
<trans-unit id="saabeb4cab074b0b9">
<source>User Object Permissions</source>
</trans-unit>
<trans-unit id="s8489d5559dda260c">
<source>Role Object Permissions</source>
</trans-unit>
<trans-unit id="s3d197283cb019b5a">
<source>Overview</source>
</trans-unit>
<trans-unit id="s6c3daaac4eed12f9">
<source>Changelog</source>
</trans-unit>
<trans-unit id="se8dca0132c66ae03">
<source>Permissions</source>
</trans-unit>
<trans-unit id="s05e395ff60af047b">
<source>Warning: Provider is not used by any Outpost.</source>
</trans-unit>
<trans-unit id="sccbfc4dec0c8d80c">
<source>Assigned to application</source>
</trans-unit>
<trans-unit id="s2d46e3a9ee8e0e7e">
<source>Update LDAP Provider</source>
</trans-unit>
<trans-unit id="saf24e253b3b006d4">
<source>How to connect</source>
</trans-unit>
<trans-unit id="s02b3fade1795d03f">
<source>Connect to the LDAP Server on port 389:</source>
</trans-unit>
<trans-unit id="sa00cf67b54c44c71">
<source>Check the IP of the Kubernetes service, or</source>
</trans-unit>
<trans-unit id="s28f270859c5f4d51">
<source>The Host IP of the docker host</source>
</trans-unit>
<trans-unit id="sb7794c2910b1a9ec">
<source>Bind DN</source>
</trans-unit>
<trans-unit id="s5694f9421c428227">
<source>Bind Password</source>
</trans-unit>
<trans-unit id="s086e1bbe7c97ea16">
<source>Search base</source>
</trans-unit>
<trans-unit id="s417b90913e05bc17">
<source>Preview</source>
</trans-unit>
<trans-unit id="s17f3eaf3b07ece26">
<source>Warning: Provider is not used by an Application.</source>
</trans-unit>
<trans-unit id="s56806e9f63efa298">
<source>Redirect URIs</source>
</trans-unit>
<trans-unit id="sdbc08adee233f180">
<source>Update OAuth2 Provider</source>
</trans-unit>
<trans-unit id="s9d96eb5ca93e6473">
<source>OpenID Configuration URL</source>
</trans-unit>
<trans-unit id="s74f809a69e030351">
<source>OpenID Configuration Issuer</source>
</trans-unit>
<trans-unit id="s028be8989873f001">
<source>Authorize URL</source>
</trans-unit>
<trans-unit id="sebda1d54a3f9f967">
<source>Token URL</source>
</trans-unit>
<trans-unit id="s2fc3eb68c7ced3af">
<source>Userinfo URL</source>
</trans-unit>
<trans-unit id="s145483489b87a622">
<source>Logout URL</source>
</trans-unit>
<trans-unit id="s59f5eda30a904b75">
<source>JWKS URL</source>
</trans-unit>
<trans-unit id="s453b0c150a7ca58e">
<source>Example JWT payload (for currently authenticated user)</source>
</trans-unit>
<trans-unit id="scb489a1a173ac3f0">
<source>Yes</source>
</trans-unit>
<trans-unit id="s09205907b5b56cda">
<source>No</source>
</trans-unit>
<trans-unit id="sc6e8a34361c7c272">
<source>Forward auth (domain-level)</source>
</trans-unit>
<trans-unit id="s6df42b3072a2d7e9">
<source>Nginx (Ingress)</source>
</trans-unit>
<trans-unit id="s8e01a852c1db8d29">
<source>Nginx (Proxy Manager)</source>
</trans-unit>
<trans-unit id="sabebdc7fa6a5bddb">
<source>Nginx (standalone)</source>
</trans-unit>
<trans-unit id="s5d9f93f1fe1c19d3">
<source>Traefik (Ingress)</source>
</trans-unit>
<trans-unit id="se2b62f7e9017965e">
<source>Traefik (Compose)</source>
</trans-unit>
<trans-unit id="s4c4c504a48c3b7bd">
<source>Traefik (Standalone)</source>
</trans-unit>
<trans-unit id="s7ba9677d069e5f02">
<source>Caddy (Standalone)</source>
</trans-unit>
<trans-unit id="s4a1e774ab25aa232">
<source>Internal Host</source>
</trans-unit>
<trans-unit id="sc9c3578cce3cf7a8">
<source>External Host</source>
</trans-unit>
<trans-unit id="s7a141f1b61074fbe">
<source>Basic-Auth</source>
</trans-unit>
<trans-unit id="s37cbecaec58e2192">
<source>Mode</source>
</trans-unit>
<trans-unit id="s4e474b9e2e737dd1">
<source>Update Proxy Provider</source>
</trans-unit>
<trans-unit id="s37eb2f1b6e3c19c2">
<source>Protocol Settings</source>
</trans-unit>
<trans-unit id="s5116b89f7db1fbec">
<source>Allowed Redirect URIs</source>
</trans-unit>
<trans-unit id="saeff3596e1ac31b6">
<source>Setup</source>
</trans-unit>
<trans-unit id="s1b783856ab4aaaf3">
<source>No additional setup is required.</source>
</trans-unit>
<trans-unit id="s09b671b120443043">
<source>Update Radius Provider</source>
</trans-unit>
<trans-unit id="sd3386a2ef42e80b9">
<source>Download</source>
</trans-unit>
<trans-unit id="sf417c13d7a0f7995">
<source>Copy download URL</source>
</trans-unit>
<trans-unit id="sc1cfce89ebcf1bf9">
<source>Download signing certificate</source>
</trans-unit>
<trans-unit id="s2c0de3d35a7bc784">
<source>Related objects</source>
</trans-unit>
<trans-unit id="s803b0621006085be">
<source>Update SAML Provider</source>
</trans-unit>
<trans-unit id="s44b1f042790cd1a2">
<source>SAML Configuration</source>
</trans-unit>
<trans-unit id="sba999428083abce3">
<source>EntityID/Issuer</source>
</trans-unit>
<trans-unit id="scd2984ee5552643a">
<source>SSO URL (Post)</source>
</trans-unit>
<trans-unit id="saa79b47f60c66458">
<source>SSO URL (Redirect)</source>
</trans-unit>
<trans-unit id="s2da51a6287118ba8">
<source>SSO URL (IdP-initiated Login)</source>
</trans-unit>
<trans-unit id="s0a57e911e457302b">
<source>SLO URL (Post)</source>
</trans-unit>
<trans-unit id="s1e7308bb1ca323e1">
<source>SLO URL (Redirect)</source>
</trans-unit>
<trans-unit id="sd2c58d7c6dddc515">
<source>SAML Metadata</source>
</trans-unit>
<trans-unit id="s382b702673776873">
<source>Example SAML attributes</source>
</trans-unit>
<trans-unit id="sea3bfc143ced73db">
<source>NameID attribute</source>
</trans-unit>
<trans-unit id="s2f0f6691de0b0388">
<source>Warning: Provider is not assigned to an application as backchannel provider.</source>
</trans-unit>
<trans-unit id="sc6c575c5ff64cdb1">
<source>Update SCIM Provider</source>
</trans-unit>
<trans-unit id="s7da38af36522ff6a">
<source>Sync not run yet.</source>
</trans-unit>
<trans-unit id="sbecf8dc03c978d15">
<source>Run sync again</source>
</trans-unit>
<trans-unit id="sb564f81eb057342e">
<source>Application Icon</source>
</trans-unit>
<trans-unit id="s6ba50bb0842ba1e2">
<source>Applications</source>
</trans-unit>
<trans-unit id="s61bd841e66966325">
<source>External applications that use authentik as an identity provider via protocols like OAuth2 and SAML. All applications are shown here, even ones you cannot access.</source>
</trans-unit>
<trans-unit id="s96b2fefc550e4b1c">
<source>Provider Type</source>
</trans-unit>
<trans-unit id="sd20f6cd02c90867f">
<source>Application(s)</source>
</trans-unit>
<trans-unit id="sa347e31efbb60be2">
<source>Update Application</source>
</trans-unit>
<trans-unit id="s1f7698c061c208c9">
<source>Open</source>
</trans-unit>
<trans-unit id="sd9b556a84ae25690">
<source>Successfully sent test-request.</source>
</trans-unit>
<trans-unit id="s5deac600e329de1b">
<source>Log messages</source>
</trans-unit>
<trans-unit id="s3feea7b49673bef2">
<source>No log messages.</source>
</trans-unit>
<trans-unit id="sa45a194b58837e4f">
<source>Active</source>
</trans-unit>
<trans-unit id="s58c867aac77b9158">
<source>Last login</source>
</trans-unit>
<trans-unit id="s3e3bb9e7cb1de4fd">
<source>Select users to add</source>
</trans-unit>
<trans-unit id="s75d5ff5dd8d3c6d2">
<source>Successfully updated group.</source>
</trans-unit>
<trans-unit id="s3079ca1184e77573">
<source>Successfully created group.</source>
</trans-unit>
<trans-unit id="s63cb05541b294335">
<source>Is superuser</source>
</trans-unit>
<trans-unit id="s29315e374008d0c5">
<source>Users added to this group will be superusers.</source>
</trans-unit>
<trans-unit id="s4eb514ebcb80553d">
<source>Parent</source>
</trans-unit>
<trans-unit id="s6b2beba7ab637e9e">
<source>Roles</source>
</trans-unit>
<trans-unit id="s96d2bb4be3f5e8aa">
<source>Select roles to grant this groups' users' permissions from the selected roles.</source>
</trans-unit>
<trans-unit id="s16b9446e3a70e1f4">
<source>Attributes</source>
</trans-unit>
<trans-unit id="sec97cdaf7af8648b">
<source>Set custom attributes using YAML or JSON.</source>
</trans-unit>
<trans-unit id="s1e36813d3504ed48">
<source>Successfully updated binding.</source>
</trans-unit>
<trans-unit id="s1bf56ee106e9e711">
<source>Successfully created binding.</source>
</trans-unit>
<trans-unit id="s042baf59902a711f">
<source>Policy</source>
</trans-unit>
<trans-unit id="s5f5bf4ef2bd93c04">
<source>Group mappings can only be checked if a user is already logged in when trying to access this source.</source>
</trans-unit>
<trans-unit id="s6c607d74bdfe9f36">
<source>User mappings can only be checked if a user is already logged in when trying to access this source.</source>
</trans-unit>
<trans-unit id="s965c503c3e42fdfe">
<source>Enabled</source>
</trans-unit>
<trans-unit id="s6b85380416964890">
<source>Negate result</source>
</trans-unit>
<trans-unit id="s3bfa0258999fb629">
<source>Negates the outcome of the binding. Messages are unaffected.</source>
</trans-unit>
<trans-unit id="s2ba5f4d8f3bd7c57">
<source>Order</source>
</trans-unit>
<trans-unit id="se1e040b55319a0e8">
<source>Timeout</source>
</trans-unit>
<trans-unit id="s6b1ed7507f26cb4a">
<source>Failure result</source>
</trans-unit>
<trans-unit id="s2e422519ed38f7d8">
<source>Pass</source>
</trans-unit>
<trans-unit id="s81a45c4fd11e8e1a">
<source>Don't pass</source>
</trans-unit>
<trans-unit id="s95b73e0f4e47eb9a">
<source>Result used when policy execution fails.</source>
</trans-unit>
<trans-unit id="s29ec5e7889f4787f">
<source>Successfully updated policy.</source>
</trans-unit>
<trans-unit id="sfc400b2d71e49d28">
<source>Successfully created policy.</source>
</trans-unit>
<trans-unit id="safc0e0656d572f4e">
<source>A policy used for testing. Always returns the same result as specified below after waiting a random duration.</source>
</trans-unit>
<trans-unit id="s9ffa1ac03ce6fd20">
<source>Execution logging</source>
</trans-unit>
<trans-unit id="saf31b3c610036ed6">
<source>When this option is enabled, all executions of this policy will be logged. By default, only execution errors are logged.</source>
</trans-unit>
<trans-unit id="sa879d5ce584875cf">
<source>Policy-specific settings</source>
</trans-unit>
<trans-unit id="s838418d1a0815157">
<source>Pass policy?</source>
</trans-unit>
<trans-unit id="sd8c5339b82b71507">
<source>Wait (min)</source>
</trans-unit>
<trans-unit id="sda4e78c19f5b6f35">
<source>The policy takes a random time to execute. This controls the minimum time it will take.</source>
</trans-unit>
<trans-unit id="s1d30ff9ba938e68d">
<source>Wait (max)</source>
</trans-unit>
<trans-unit id="s303b5e552246e613">
<source>Matches an event against a set of criteria. If any of the configured values match, the policy passes.</source>
</trans-unit>
<trans-unit id="s890810efbe103cbc">
<source>Match created events with this action type. When left empty, all action types will be matched.</source>
</trans-unit>
<trans-unit id="sfab527528ea64618">
<source>Matches Event's Client IP (strict matching, for network matching use an Expression Policy.</source>
</trans-unit>
<trans-unit id="s5a15a8f39c699273">
<source>Match events created by selected application. When left empty, all applications are matched.</source>
</trans-unit>
<trans-unit id="s5da52af9b083c29a">
<source>Model</source>
</trans-unit>
<trans-unit id="s3ba9b8aeb686d9f7">
<source>Match events created by selected model. When left empty, all models are matched.</source>
</trans-unit>
<trans-unit id="s5a13f4bbe004503f">
<source>Checks if the request's user's password has been changed in the last x days, and denys based on settings.</source>
</trans-unit>
<trans-unit id="sfad8af8ce38104a3">
<source>Maximum age (in days)</source>
</trans-unit>
<trans-unit id="s9307f3dbb07a73b5">
<source>Only fail the policy, don't invalidate user's password</source>
</trans-unit>
<trans-unit id="scea1f16238093e35">
<source>Executes the python snippet to determine whether to allow or deny a request.</source>
</trans-unit>
<trans-unit id="sabd1bc9fb7da71e7">
<source>Expression using Python.</source>
</trans-unit>
<trans-unit id="s8d08843f397d9e81">
<source>See documentation for a list of all variables.</source>
</trans-unit>
<trans-unit id="se2cc93bd2647baec">
<source>Static rules</source>
</trans-unit>
<trans-unit id="sc96dd9d2e7b05fc5">
<source>Minimum length</source>
</trans-unit>
<trans-unit id="s33d48fb745f4d4ae">
<source>Minimum amount of Uppercase Characters</source>
</trans-unit>
<trans-unit id="s883b544e2b4aa3b5">
<source>Minimum amount of Lowercase Characters</source>
</trans-unit>
<trans-unit id="s43be3ce2439ffe9c">
<source>Minimum amount of Digits</source>
</trans-unit>
<trans-unit id="sb3651834cca86735">
<source>Minimum amount of Symbols Characters</source>
</trans-unit>
<trans-unit id="sc2f116c0ea77d58a">
<source>Error message</source>
</trans-unit>
<trans-unit id="s21d0e290c51a8ef9">
<source>Symbol charset</source>
</trans-unit>
<trans-unit id="s545d99afa61e4095">
<source>Characters which are considered as symbols.</source>
</trans-unit>
<trans-unit id="s1293ad87acc7a609">
<source>HaveIBeenPwned settings</source>
</trans-unit>
<trans-unit id="sdf4e1c6a2f072600">
<source>Allowed count</source>
</trans-unit>
<trans-unit id="scd8062ff5e1326d8">
<source>Allow up to N occurrences in the HIBP database.</source>
</trans-unit>
<trans-unit id="s3fd219b045193507">
<source>zxcvbn settings</source>
</trans-unit>
<trans-unit id="s28d84abfbaf555ea">
<source>Score threshold</source>
</trans-unit>
<trans-unit id="s7b3148ffba9f4527">
<source>If the password's score is less than or equal this value, the policy will fail.</source>
</trans-unit>
<trans-unit id="sd94e99af8b41ff54">
<source>0: Too guessable: risky password. (guesses &lt; 10^3)</source>
</trans-unit>
<trans-unit id="sc926385d1a624c3a">
<source>1: Very guessable: protection from throttled online attacks. (guesses &lt; 10^6)</source>
</trans-unit>
<trans-unit id="s8aae61c41319602c">
<source>2: Somewhat guessable: protection from unthrottled online attacks. (guesses &lt; 10^8)</source>
</trans-unit>
<trans-unit id="sc1f4b57e722a89d6">
<source>3: Safely unguessable: moderate protection from offline slow-hash scenario. (guesses &lt; 10^10)</source>
</trans-unit>
<trans-unit id="sd47f3d3c9741343d">
<source>4: Very unguessable: strong protection from offline slow-hash scenario. (guesses &gt;= 10^10)</source>
</trans-unit>
<trans-unit id="sd6cd7ce2310a73a4">
<source>Checks the value from the policy request against several rules, mostly used to ensure password strength.</source>
</trans-unit>
<trans-unit id="s2a957e843960b604">
<source>Password field</source>
</trans-unit>
<trans-unit id="se8a81c75b6e30a33">
<source>Field key to check, field keys defined in Prompt stages are available.</source>
</trans-unit>
<trans-unit id="s2f8c4cf12350a36c">
<source>Check static rules</source>
</trans-unit>
<trans-unit id="sd75a9a71309fb387">
<source>Check haveibeenpwned.com</source>
</trans-unit>
<trans-unit id="se5cb18408df3284e">
<source>For more info see:</source>
</trans-unit>
<trans-unit id="scef7abb8456b06d6">
<source>Check zxcvbn</source>
</trans-unit>
<trans-unit id="sdfdb58cd232b363d">
<source>Password strength estimator created by Dropbox, see:</source>
</trans-unit>
<trans-unit id="s40b034801fcb843b">
<source>Allows/denys requests based on the users and/or the IPs reputation.</source>
</trans-unit>
<trans-unit id="scf4afecb0f1e69b2">
<source>Invalid login attempts will decrease the score for the client's IP, and the
username they are attempting to login as, by one.</source>
</trans-unit>
<trans-unit id="s8323a9af28e10502">
<source>The policy passes when the reputation score is below the threshold, and
doesn't pass when either or both of the selected options are equal or above the threshold.</source>
</trans-unit>
<trans-unit id="s1828fbfc2c56379c">
<source>Check IP</source>
</trans-unit>
<trans-unit id="s4751df77cfd8a5f9">
<source>Check Username</source>
</trans-unit>
<trans-unit id="se19cc57dd8675498">
<source>Threshold</source>
</trans-unit>
<trans-unit id="sdbccb39a658f0e45">
<source>New policy</source>
</trans-unit>
<trans-unit id="sf693300708a40d2c">
<source>Create a new policy.</source>
</trans-unit>
<trans-unit id="s5b1fb0d4c0daeba8">
<source>Create Binding</source>
</trans-unit>
<trans-unit id="s9f5a5f23312798f0">
<source>Members</source>
</trans-unit>
<trans-unit id="s7eb3d239e0b491ab">
<source>Select groups to add user to</source>
</trans-unit>
<trans-unit id="sec5cdfa358f9dbf7">
<source>Warning: Adding the user to the selected group(s) will give them superuser permissions.</source>
</trans-unit>
<trans-unit id="scab2900019953050">
<source>Successfully updated user.</source>
</trans-unit>
<trans-unit id="s3d2a8b86a4f5a810">
<source>Successfully created user and added to group <x id="0" equiv-text="${this.group.name}"/></source>
</trans-unit>
<trans-unit id="s9c3c272944dcfca3">
<source>Successfully created user.</source>
</trans-unit>
<trans-unit id="s03f42eea72154959">
<source>Username</source>
</trans-unit>
<trans-unit id="s5a802e46a033c8af">
<source>User's primary identifier. 150 characters or fewer.</source>
</trans-unit>
<trans-unit id="sd34be0d0fcb39971">
<source>User's display name.</source>
</trans-unit>
<trans-unit id="s9f9492d30a96b9c6">
<source>User type</source>
</trans-unit>
<trans-unit id="scef2eb6a2bfe3110">
<source>Internal users might be users such as company employees, which will get access to the full Enterprise feature set.</source>
</trans-unit>
<trans-unit id="sf66389b04fcc219c">
<source>External users might be external consultants or B2C customers. These users don't get access to enterprise features.</source>
</trans-unit>
<trans-unit id="s77e8668a27dbc402">
<source>Service accounts should be used for machine-to-machine authentication or other automations.</source>
</trans-unit>
<trans-unit id="sd1f44f1a8bc20e67">
<source>Email</source>
</trans-unit>
<trans-unit id="sbe3b416a356f1c91">
<source>Is active</source>
</trans-unit>
<trans-unit id="s35fac2e5677d55cd">
<source>Designates whether this user should be treated as active. Unselect this instead of deleting accounts.</source>
</trans-unit>
<trans-unit id="s2e532e19ed477a56">
<source>Path</source>
</trans-unit>
<trans-unit id="s67560d7e37d984c3">
<source>Policy / User / Group</source>
</trans-unit>
<trans-unit id="s030ac0829bb50a49">
<source>Policy <x id="0" equiv-text="${item.policyObj?.name}"/></source>
</trans-unit>
<trans-unit id="s2a64d2dca3da9b0e">
<source>Group <x id="0" equiv-text="${item.groupObj?.name}"/></source>
</trans-unit>
<trans-unit id="se5dc026819a32ff8">
<source>User <x id="0" equiv-text="${item.userObj?.name}"/></source>
</trans-unit>
<trans-unit id="s50c312bea93b6925">
<source>Edit Policy</source>
</trans-unit>
<trans-unit id="s0b55a57f473ab8af">
<source>Update Group</source>
</trans-unit>
<trans-unit id="s494e1ed913d9351a">
<source>Edit Group</source>
</trans-unit>
<trans-unit id="sad130c2d925fb7bf">
<source>Update User</source>
</trans-unit>
<trans-unit id="s5cd31f4a88adf180">
<source>Edit User</source>
</trans-unit>
<trans-unit id="se291dfd2a59d7842">
<source>Policy binding(s)</source>
</trans-unit>
<trans-unit id="s7e87ab366c199345">
<source>Update Binding</source>
</trans-unit>
<trans-unit id="s40b80eb4cc1f0e0c">
<source>Edit Binding</source>
</trans-unit>
<trans-unit id="sbad5b96fb855ef36">
<source>No Policies bound.</source>
</trans-unit>
<trans-unit id="sc15d60377cc8aaac">
<source>No policies are currently bound to this object.</source>
</trans-unit>
<trans-unit id="s911a27022aba349f">
<source>Create and bind Policy</source>
</trans-unit>
<trans-unit id="sddb040c47daae56b">
<source>Bind existing policy</source>
</trans-unit>
<trans-unit id="saa855c61e0403fe6">
<source>Warning: Application is not used by any Outpost.</source>
</trans-unit>
<trans-unit id="sb6cbd4f92ebaf5d8">
<source>Related</source>
</trans-unit>
<trans-unit id="sd71081c23d1cd38b">
<source>Check access</source>
</trans-unit>
<trans-unit id="s42cbd8dca939a9c7">
<source>Check</source>
</trans-unit>
<trans-unit id="sf22f7f8a9309b4ed">
<source>Check Application access</source>
</trans-unit>
<trans-unit id="s2474e7fb1aec9f05">
<source>Test</source>
</trans-unit>
<trans-unit id="s512957aa09384646">
<source>Launch</source>
</trans-unit>
<trans-unit id="sed02f831e653deb3">
<source>Logins over the last week (per 8 hours)</source>
</trans-unit>
<trans-unit id="s2b1bc31276c4c477">
<source>Policy / Group / User Bindings</source>
</trans-unit>
<trans-unit id="s473f0143efa3f706">
<source>These policies control which users can access this application.</source>
</trans-unit>
<trans-unit id="s24875d5475e82526">
<source>Successfully updated source.</source>
</trans-unit>
<trans-unit id="s60d891ed3ee9ebc5">
<source>Successfully created source.</source>
</trans-unit>
<trans-unit id="s8af7239354f7e7b6">
<source>Sync users</source>
</trans-unit>
<trans-unit id="sd80b0b8aeae3abe3">
<source>User password writeback</source>
</trans-unit>
<trans-unit id="s2b952e9dc99cbded">
<source>Login password is synced from LDAP into authentik automatically. Enable this option only to write password changes in authentik back to LDAP.</source>
</trans-unit>
<trans-unit id="saf7ce4165a1025f6">
<source>Sync groups</source>
</trans-unit>
<trans-unit id="s2035f889f576bca6">
<source>Connection settings</source>
</trans-unit>
<trans-unit id="s0a72e65aef45b1e8">
<source>Server URI</source>
</trans-unit>
<trans-unit id="sa599dbe5776897ad">
<source>Specify multiple server URIs by separating them with a comma.</source>
</trans-unit>
<trans-unit id="se36b55dfcf5dc80b">
<source>Enable StartTLS</source>
</trans-unit>
<trans-unit id="s33683c3b1dbaf264">
<source>To use SSL instead, use 'ldaps://' and disable this option.</source>
</trans-unit>
<trans-unit id="s1024166475850a65">
<source>Use Server URI for SNI verification</source>
</trans-unit>
<trans-unit id="se65beb94fffc3c4b">
<source>Required for servers using TLS 1.3+</source>
</trans-unit>
<trans-unit id="s2221fef80f4753a2">
<source>TLS Verification Certificate</source>
</trans-unit>
<trans-unit id="sb8c13bd58191cea2">
<source>When connecting to an LDAP Server with TLS, certificates are not checked by default. Specify a keypair to validate the remote certificate.</source>
</trans-unit>
<trans-unit id="s000ee3e634868b3c">
<source>TLS Client authentication certificate</source>
</trans-unit>
<trans-unit id="s5506b35a1bceb141">
<source>Client certificate keypair to authenticate against the LDAP Server's Certificate.</source>
</trans-unit>
<trans-unit id="sb7684e2910a33a1f">
<source>Bind CN</source>
</trans-unit>
<trans-unit id="s3de6db803012016a">
<source>LDAP Attribute mapping</source>
</trans-unit>
<trans-unit id="s7c05ee41d634aa45">
<source>Property mappings used to user creation.</source>
</trans-unit>
<trans-unit id="s94333971a07803b9">
<source>Additional settings</source>
</trans-unit>
<trans-unit id="sd14a19a19d507f9e">
<source>Parent group for all the groups imported from LDAP.</source>
</trans-unit>
<trans-unit id="sfbc59ff17a73503d">
<source>User path</source>
</trans-unit>
<trans-unit id="sd18170637295bace">
<source>Addition User DN</source>
</trans-unit>
<trans-unit id="s9ae089fd248e72db">
<source>Additional user DN, prepended to the Base DN.</source>
</trans-unit>
<trans-unit id="s5944355d69db1fb8">
<source>Addition Group DN</source>
</trans-unit>
<trans-unit id="sfae9f4ea5749a36b">
<source>Additional group DN, prepended to the Base DN.</source>
</trans-unit>
<trans-unit id="s66ffc06300964849">
<source>User object filter</source>
</trans-unit>
<trans-unit id="s1c2a173db0e1ec61">
<source>Consider Objects matching this filter to be Users.</source>
</trans-unit>
<trans-unit id="s2ec94a7c7f5bcd1b">
<source>Group object filter</source>
</trans-unit>
<trans-unit id="saf5eb7596b3a355b">
<source>Consider Objects matching this filter to be Groups.</source>
</trans-unit>
<trans-unit id="sf325a4adba4d6278">
<source>Group membership field</source>
</trans-unit>
<trans-unit id="s76768bebabb7d543">
<source>Field which contains members of a group. Note that if using the "memberUid" field, the value is assumed to contain a relative distinguished name. e.g. 'memberUid=some-user' instead of 'memberUid=cn=some-user,ou=groups,...'</source>
</trans-unit>
<trans-unit id="s026555347e589f0e">
<source>Object uniqueness field</source>
</trans-unit>
<trans-unit id="s24211f319e5b7e98">
<source>Field which contains a unique Identifier.</source>
</trans-unit>
<trans-unit id="s900b0d85b872d134">
<source>Link users on unique identifier</source>
</trans-unit>
<trans-unit id="s6c70a73265e14521">
<source>Link to a user with identical email address. Can have security implications when a source doesn't validate email addresses</source>
</trans-unit>
<trans-unit id="s995535e7af30d754">
<source>Use the user's email address, but deny enrollment when the email address already exists</source>
</trans-unit>
<trans-unit id="s542ecb4130f6cea5">
<source>Link to a user with identical username. Can have security implications when a username is used with another source</source>
</trans-unit>
<trans-unit id="s2a1debf34e5aeba4">
<source>Use the user's username, but deny enrollment when the username already exists</source>
</trans-unit>
<trans-unit id="s81ce0d54727f42d2">
<source>Unknown user matching mode</source>
</trans-unit>
<trans-unit id="sd04376c4216c921f">
<source>URL settings</source>
</trans-unit>
<trans-unit id="s872d0e88ab34ed83">
<source>Authorization URL</source>
</trans-unit>
<trans-unit id="see3ff55262fd6500">
<source>URL the user is redirect to to consent the authorization.</source>
</trans-unit>
<trans-unit id="sb932dead79567c7b">
<source>Access token URL</source>
</trans-unit>
<trans-unit id="s88b8a2892635a2fc">
<source>URL used by authentik to retrieve tokens.</source>
</trans-unit>
<trans-unit id="s69bd313dd12fc2f3">
<source>Profile URL</source>
</trans-unit>
<trans-unit id="sa8d83cd8023e8e4d">
<source>URL used by authentik to get user information.</source>
</trans-unit>
<trans-unit id="sc7707b3ba3a2a7ca">
<source>Request token URL</source>
</trans-unit>
<trans-unit id="s3926da5b20cdf3b6">
<source>URL used to request the initial token. This URL is only required for OAuth 1.</source>
</trans-unit>
<trans-unit id="s199b55513a739f43">
<source>OIDC Well-known URL</source>
</trans-unit>
<trans-unit id="s8b149b30b5b523ef">
<source>OIDC well-known configuration URL. Can be used to automatically configure the URLs above.</source>
</trans-unit>
<trans-unit id="s9db2c836ade1339c">
<source>OIDC JWKS URL</source>
</trans-unit>
<trans-unit id="s4b2a1b657c160f5b">
<source>JSON Web Key URL. Keys from the URL will be used to validate JWTs from this source.</source>
</trans-unit>
<trans-unit id="s2df0b65125600de9">
<source>OIDC JWKS</source>
</trans-unit>
<trans-unit id="s02de8d9e8583b480">
<source>Raw JWKS data.</source>
</trans-unit>
<trans-unit id="s81a87652ade099e4">
<source>User matching mode</source>
</trans-unit>
<trans-unit id="se8987bdfb35e46b2">
<source>Consumer key</source>
</trans-unit>
<trans-unit id="sb35c08e3a541188f">
<source>Also known as Client ID.</source>
</trans-unit>
<trans-unit id="sabaf0061f7e41b0b">
<source>Consumer secret</source>
</trans-unit>
<trans-unit id="sd46fd9b647cfea10">
<source>Also known as Client Secret.</source>
</trans-unit>
<trans-unit id="sa61966cd83b4924c">
<source>Additional scopes to be passed to the OAuth Provider, separated by space. To replace existing scopes, prefix with *.</source>
</trans-unit>
<trans-unit id="s1cc0e66dbd2b5502">
<source>Flow settings</source>
</trans-unit>
<trans-unit id="sfe6977a3aea3ee6e">
<source>Flow to use when authenticating existing users.</source>
</trans-unit>
<trans-unit id="s2801a48ceac691b3">
<source>Enrollment flow</source>
</trans-unit>
<trans-unit id="s5d0a14d29ebad561">
<source>Flow to use when enrolling new users.</source>
</trans-unit>
<trans-unit id="s91f389c796720a81">
<source>Load servers</source>
</trans-unit>
<trans-unit id="s24f405197ede5ebb">
<source>Re-authenticate with plex</source>
</trans-unit>
<trans-unit id="sc297b2e13c28ecf9">
<source>Allow friends to authenticate via Plex, even if you don't share any servers</source>
</trans-unit>
<trans-unit id="sfee91e08b8b47477">
<source>Allowed servers</source>
</trans-unit>
<trans-unit id="s216eb300543edd91">
<source>Select which server a user has to be a member of to be allowed to authenticate.</source>
</trans-unit>
<trans-unit id="s31d7f3ba04d306a5">
<source>SSO URL</source>
</trans-unit>
<trans-unit id="s1d9d6c5b424fdc1f">
<source>URL that the initial Login request is sent to.</source>
</trans-unit>
<trans-unit id="sd94db2b8c85d10a6">
<source>SLO URL</source>
</trans-unit>
<trans-unit id="sc764ddf60b5149de">
<source>Optional URL if the IDP supports Single-Logout.</source>
</trans-unit>
<trans-unit id="se7430794fa89005a">
<source>Also known as Entity ID. Defaults the Metadata URL.</source>
</trans-unit>
<trans-unit id="s5615bb595ad6ded6">
<source>Binding Type</source>
</trans-unit>
<trans-unit id="sa2e4d6830226d3ec">
<source>Redirect binding</source>
</trans-unit>
<trans-unit id="s6f96a78d81ef277c">
<source>Post-auto binding</source>
</trans-unit>
<trans-unit id="sc2c70fd56f5d0b48">
<source>Post binding but the request is automatically sent and the user doesn't have to confirm.</source>
</trans-unit>
<trans-unit id="s968c90258dcf7562">
<source>Post binding</source>
</trans-unit>
<trans-unit id="se10bbf4cf861c81b">
<source>Signing keypair</source>
</trans-unit>
<trans-unit id="s838ed611b533b19e">
<source>Keypair which is used to sign outgoing requests. Leave empty to disable signing.</source>
</trans-unit>
<trans-unit id="s39c8c0bf4d927c9f">
<source> Allow IDP-initiated logins</source>
</trans-unit>
<trans-unit id="s65d507f1513c2f03">
<source>Allows authentication flows initiated by the IdP. This can be a security risk, as no validation of the request ID is done.</source>
</trans-unit>
<trans-unit id="s297a2075bd7e40db">
<source>NameID Policy</source>
</trans-unit>
<trans-unit id="s004e9a2c90f23900">
<source>Persistent</source>
</trans-unit>
<trans-unit id="s38887b94b3320533">
<source>Email address</source>
</trans-unit>
<trans-unit id="s2d34c87f67f66c6a">
<source>Windows</source>
</trans-unit>
<trans-unit id="s1665454e31e14941">
<source>X509 Subject</source>
</trans-unit>
<trans-unit id="s0c3ac7f9383a8cfd">
<source>Transient</source>
</trans-unit>
<trans-unit id="s20a0ce62823bfa97">
<source>Delete temporary users after</source>
</trans-unit>
<trans-unit id="s3198c384c2f68b08">
<source>Time offset when temporary users should be deleted. This only applies if your IDP uses the NameID Format 'transient', and the user doesn't log out manually.</source>
</trans-unit>
<trans-unit id="sb32e9c1faa0b8673">
<source>Pre-authentication flow</source>
</trans-unit>
<trans-unit id="sa3c1f6ac5e63a70f">
<source>Flow used before authentication.</source>
</trans-unit>
<trans-unit id="se12969ade44cd2b6">
<source>New source</source>
</trans-unit>
<trans-unit id="s19b09f4fc72175d1">
<source>Create a new source.</source>
</trans-unit>
<trans-unit id="sb1a4e9b288e2f005">
<source>Federation and Social login</source>
</trans-unit>
<trans-unit id="s6152026c364ad974">
<source>Sources of identities, which can either be synced into authentik's database, or can be used by users to authenticate and enroll themselves.</source>
</trans-unit>
<trans-unit id="s0a0ca63b967f1630">
<source>Source(s)</source>
</trans-unit>
<trans-unit id="s66722bc2ea775e05">
<source>Disabled</source>
</trans-unit>
<trans-unit id="s4ff2c202b4e5bdc5">
<source>Built-in</source>
</trans-unit>
<trans-unit id="s4476e9c50cfd13f4">
<source>Global status</source>
</trans-unit>
<trans-unit id="sd21a971eea208533">
<source>Vendor</source>
</trans-unit>
<trans-unit id="sadadfe9dfa06d7dd">
<source>No sync status.</source>
</trans-unit>
<trans-unit id="s2b1c81130a65a55b">
<source>Sync currently running.</source>
</trans-unit>
<trans-unit id="s31a2d43bc1cf1790">
<source>Not synced yet.</source>
</trans-unit>
<trans-unit id="s388ee787bbf2271b">
<source>Task finished with warnings</source>
</trans-unit>
<trans-unit id="s949826fad0fe0909">
<source>Task finished with errors</source>
</trans-unit>
<trans-unit id="sbedb77365a066648">
<source>Last sync: <x id="0" equiv-text="${task.taskFinishTimestamp.toLocaleString()}"/></source>
</trans-unit>
<trans-unit id="s52b500138a2d2b8a">
<source>Update LDAP Source</source>
</trans-unit>
<trans-unit id="sf36170f71cea38c2">
<source>Connectivity</source>
</trans-unit>
<trans-unit id="sf3fec8353106ac2f">
<source>OAuth Source <x id="0" equiv-text="${this.source.name}"/></source>
</trans-unit>
<trans-unit id="se09d055771f3a11d">
<source>Generic OpenID Connect</source>
</trans-unit>
<trans-unit id="s5c1dc164c89ac13e">
<source>Unknown provider type</source>
</trans-unit>
<trans-unit id="s355b21b89ce5d9c5">
<source>Details</source>
</trans-unit>
<trans-unit id="s01088b6625d2443b">
<source>Callback URL</source>
</trans-unit>
<trans-unit id="sb6d5146d5efb3058">
<source>Access Key</source>
</trans-unit>
<trans-unit id="s065604a41e9d1584">
<source>Update OAuth Source</source>
</trans-unit>
<trans-unit id="s7b576aa71acb36a6">
<source>Diagram</source>
</trans-unit>
<trans-unit id="s587ba266269297ab">
<source>Policy Bindings</source>
</trans-unit>
<trans-unit id="s2feae323f46479f8">
<source>These bindings control which users can access this source.
You can only use policies here as access is checked before the user is authenticated.</source>
</trans-unit>
<trans-unit id="se17fcb1f159ee382">
<source>Update Plex Source</source>
</trans-unit>
<trans-unit id="saa10777250a6deca">
<source>Update SAML Source</source>
</trans-unit>
<trans-unit id="s643d8f2e5e5e930d">
<source>Successfully updated mapping.</source>
</trans-unit>
<trans-unit id="sffeef5b119d8625c">
<source>Successfully created mapping.</source>
</trans-unit>
<trans-unit id="s1c33d22492029aba">
<source>Object field</source>
</trans-unit>
<trans-unit id="s06df3c3b6a503da8">
<source>Field of the user object this value is written to.</source>
</trans-unit>
<trans-unit id="sd39c5e998efecf93">
<source>SAML Attribute Name</source>
</trans-unit>
<trans-unit id="scf2790cf3ad89283">
<source>Attribute name used for SAML Assertions. Can be a URN OID, a schema reference, or a any other string. If this property mapping is used for NameID Property, this field is discarded.</source>
</trans-unit>
<trans-unit id="sab6d24c5ec8dc361">
<source>Friendly Name</source>
</trans-unit>
<trans-unit id="s9f8aac89fe318acc">
<source>Optionally set the 'FriendlyName' value of the Assertion attribute.</source>
</trans-unit>
<trans-unit id="s851c108679653d2a">
<source>Scope name</source>
</trans-unit>
<trans-unit id="s23fd4411419fca06">
<source>Scope which the client can specify to access these properties.</source>
</trans-unit>
<trans-unit id="s7754f0e34f27fb6e">
<source>Description shown to the user when consenting. If left empty, the user won't be informed.</source>
</trans-unit>
<trans-unit id="sb6c3bf5489d7556e">
<source>Example context data</source>
</trans-unit>
<trans-unit id="s4a697f0b36c4fe83">
<source>Active Directory User</source>
</trans-unit>
<trans-unit id="s9277b90db38e1983">
<source>Active Directory Group</source>
</trans-unit>
<trans-unit id="sc2e03590269d5a10">
<source>New property mapping</source>
</trans-unit>
<trans-unit id="s713e8666ed70f8b3">
<source>Create a new property mapping.</source>
</trans-unit>
<trans-unit id="sb37880a2a7288ef0">
<source>Update Permissions</source>
</trans-unit>
<trans-unit id="s271a7e04ff9865b1">
<source>Control how authentik exposes and interprets information.</source>
</trans-unit>
<trans-unit id="s59dc0eda07f9e2b6">
<source>Property Mapping(s)</source>
</trans-unit>
<trans-unit id="sa57c393736e2732c">
<source>Test Property Mapping</source>
</trans-unit>
<trans-unit id="sc39fb3ff3753d5ab">
<source>Hide managed mappings</source>
</trans-unit>
<trans-unit id="s476ffc07e6d66f18">
<source>Successfully updated token.</source>
</trans-unit>
<trans-unit id="s93c1e5fbe8184895">
<source>Successfully created token.</source>
</trans-unit>
<trans-unit id="s4165cd175bc4c0c4">
<source>Expires on</source>
</trans-unit>
<trans-unit id="s5fc4269c2addee61">
<source>Unique identifier the token is referenced by.</source>
</trans-unit>
<trans-unit id="sb8bc2b8376c96a6b">
<source>Intent</source>
</trans-unit>
<trans-unit id="sbd34d118bcb1aaf2">
<source>API Token</source>
</trans-unit>
<trans-unit id="se31d92bea7f3a186">
<source>Used to access the API programmatically</source>
</trans-unit>
<trans-unit id="sfd586951c75eb291">
<source>App password.</source>
</trans-unit>
<trans-unit id="s59bf194136d0d13a">
<source>Used to login using a flow executor</source>
</trans-unit>
<trans-unit id="s1b14062c44e5ef45">
<source>Expiring</source>
</trans-unit>
<trans-unit id="safcc54b2aedb1a17">
<source>If this is selected, the token will expire. Upon expiration, the token will be rotated.</source>
</trans-unit>
<trans-unit id="sdc5690be4a342985">
<source>The token has been copied to your clipboard</source>
</trans-unit>
<trans-unit id="s7f3edfee24690c9f">
<source>The token was displayed because authentik does not have permission to write to the clipboard</source>
</trans-unit>
<trans-unit id="s78fd8c03f8c967f3">
<source>Tokens</source>
</trans-unit>
<trans-unit id="sdcc7b2c109ce9775">
<source>Tokens are used throughout authentik for Email validation stages, Recovery keys and API access.</source>
</trans-unit>
<trans-unit id="sf71dba2c30283a54">
<source>Expires?</source>
</trans-unit>
<trans-unit id="sc7be80a7f8ec597e">
<source>Expiry date</source>
</trans-unit>
<trans-unit id="s71dcd9cf808449aa">
<source>Token(s)</source>
</trans-unit>
<trans-unit id="sb15e8daacf26bdfc">
<source>Create Token</source>
</trans-unit>
<trans-unit id="s8d7ecd944ebe834b">
<source>Token is managed by authentik.</source>
</trans-unit>
<trans-unit id="sd1288ca57e221cf9">
<source>Update Token</source>
</trans-unit>
<trans-unit id="se9c07cf256774d81">
<source>Editing is disabled for managed tokens</source>
</trans-unit>
<trans-unit id="scc3487e74c5a3e89">
<source>Copy token</source>
</trans-unit>
<trans-unit id="sab21e1f62676b56c">
<source>Successfully updated brand.</source>
</trans-unit>
<trans-unit id="sa43e43fd3a23e22d">
<source>Successfully created brand.</source>
</trans-unit>
<trans-unit id="s41706a202b6c40f1">
<source>Domain</source>
</trans-unit>
<trans-unit id="se74ce42d41e392ba">
<source>Matching is done based on domain suffix, so if you enter domain.tld, foo.domain.tld will still match.</source>
</trans-unit>
<trans-unit id="s11326fd2590f4e5e">
<source>Default</source>
</trans-unit>
<trans-unit id="s41b3f9b4c98aabd9">
<source>Use this brand for each domain that doesn't have a dedicated brand.</source>
</trans-unit>
<trans-unit id="sc19838ca8c135c1b">
<source>Branding settings</source>
</trans-unit>
<trans-unit id="s99f110d27e30b289">
<source>Title</source>
</trans-unit>
<trans-unit id="sab6bad52985c6676">
<source>Branding shown in page title and several other places.</source>
</trans-unit>
<trans-unit id="s4f1af2b48a5e249a">
<source>Logo</source>
</trans-unit>
<trans-unit id="sd6b8b4156f7df696">
<source>Icon shown in sidebar/header and flow executor.</source>
</trans-unit>
<trans-unit id="s3626433940124897">
<source>Favicon</source>
</trans-unit>
<trans-unit id="se99efc0873031976">
<source>Icon shown in the browser tab.</source>
</trans-unit>
<trans-unit id="s10356fd921037fbf">
<source>Default flows</source>
</trans-unit>
<trans-unit id="sd216b08bafb297ee">
<source>Flow used to authenticate users. If left empty, the first applicable flow sorted by the slug is used.</source>
</trans-unit>
<trans-unit id="s35e6e60e83a8c003">
<source>Invalidation flow</source>
</trans-unit>
<trans-unit id="s7989db5f4819af89">
<source>Flow used to logout. If left empty, the first applicable flow sorted by the slug is used.</source>
</trans-unit>
<trans-unit id="sfeb779d4ccbc5a0e">
<source>Recovery flow</source>
</trans-unit>
<trans-unit id="s1c2fd8097e14a608">
<source>Recovery flow. If left empty, the first applicable flow sorted by the slug is used.</source>
</trans-unit>
<trans-unit id="s836aa192b30c21da">
<source>Unenrollment flow</source>
</trans-unit>
<trans-unit id="s081d3c4b47a6ff83">
<source>If set, users are able to unenroll themselves using this flow. If no flow is set, option is not shown.</source>
</trans-unit>
<trans-unit id="secbfd13bdae95a59">
<source>User settings flow</source>
</trans-unit>
<trans-unit id="s523160b433311521">
<source>If set, users are able to configure details of their profile.</source>
</trans-unit>
<trans-unit id="s134177568525dbc8">
<source>Device code flow</source>
</trans-unit>
<trans-unit id="s7b298427bdea81ae">
<source>If set, the OAuth Device Code profile can be used, and the selected flow will be used to enter the code.</source>
</trans-unit>
<trans-unit id="s7f4e4054fbe132e1">
<source>Other global settings</source>
</trans-unit>
<trans-unit id="sbadde673052efc02">
<source>Web Certificate</source>
</trans-unit>
<trans-unit id="s17260b71484b307f">
<source>Set custom attributes using YAML or JSON. Any attributes set here will be inherited by users, if the request is handled by this brand.</source>
</trans-unit>
<trans-unit id="s79fc990a2b58f27f">
<source>Brands</source>
</trans-unit>
<trans-unit id="s164be9a7537b99f6">
<source>Configure visual settings and defaults for different domains.</source>
</trans-unit>
<trans-unit id="s4802636d55022ed3">
<source>Default?</source>
</trans-unit>
<trans-unit id="s02774bc46a167346">
<source>Brand(s)</source>
</trans-unit>
<trans-unit id="s801bf3d03f4a3ff1">
<source>Update Brand</source>
</trans-unit>
<trans-unit id="s5c3efec5330e0000">
<source>Create Brand</source>
</trans-unit>
<trans-unit id="s8cb7bb82e96d5d77">
<source>Policies</source>
</trans-unit>
<trans-unit id="sec1808532fe107b9">
<source>Allow users to use Applications based on properties, enforce Password Criteria and selectively apply Stages.</source>
</trans-unit>
<trans-unit id="se16ac750b81fa93d">
<source>Assigned to <x id="0" equiv-text="${item.boundTo}"/> object(s).</source>
</trans-unit>
<trans-unit id="s5a48d5171e1a1522">
<source>Warning: Policy is not assigned.</source>
</trans-unit>
<trans-unit id="s544142ce35050751">
<source>Test Policy</source>
</trans-unit>
<trans-unit id="s00c8354318addfa0">
<source>Policy / Policies</source>
</trans-unit>
<trans-unit id="s76da2c978dcc5ef4">
<source>Successfully cleared policy cache</source>
</trans-unit>
<trans-unit id="sa717841a602fe7d8">
<source>Failed to delete policy cache</source>
</trans-unit>
<trans-unit id="s3ed5607ad78d4224">
<source>Clear cache</source>
</trans-unit>
<trans-unit id="s1b07757762cda372">
<source>Clear Policy cache</source>
</trans-unit>
<trans-unit id="s15b46b78edebb20a">
<source>Are you sure you want to clear the policy cache? This will cause all policies to be re-evaluated on their next usage.</source>
</trans-unit>
<trans-unit id="s62ddcbaaa91d120d">
<source>Reputation scores</source>
</trans-unit>
<trans-unit id="sd080b2370aa82967">
<source>Reputation for IP and user identifiers. Scores are decreased for each failed login and increased for each successful login.</source>
</trans-unit>
<trans-unit id="s09242207b5b8f83c">
<source>IP</source>
</trans-unit>
<trans-unit id="s7d684b6257284e55">
<source>Score</source>
</trans-unit>
<trans-unit id="s10d2dbc4613397f0">
<source>Updated</source>
</trans-unit>
<trans-unit id="sa33d061d2ade20aa">
<source>Reputation</source>
</trans-unit>
<trans-unit id="s9f26843287bb592d">
<source>Groups</source>
</trans-unit>
<trans-unit id="s4dcb9288f7e9e4d7">
<source>Group users together and give them permissions based on the membership.</source>
</trans-unit>
<trans-unit id="s62f93cfcb45d5a06">
<source>Superuser privileges?</source>
</trans-unit>
<trans-unit id="s9fdda7ea4642306c">
<source>Group(s)</source>
</trans-unit>
<trans-unit id="s416a540b16275f2e">
<source>Create Group</source>
</trans-unit>
<trans-unit id="s7c5774fad9d050ce">
<source>Create group</source>
</trans-unit>
<trans-unit id="s2a12e0b5527ff99a">
<source>Enabling this toggle will create a group named after the user, with the user as member.</source>
</trans-unit>
<trans-unit id="s6b6e6eb037aef7da">
<source>Use the username and password below to authenticate. The password can be retrieved later on the Tokens page.</source>
</trans-unit>
<trans-unit id="sf6e1665c7022a1f8">
<source>Password</source>
</trans-unit>
<trans-unit id="sbb57cd8a3ed12915">
<source>Valid for 360 days, after which the password will automatically rotate. You can copy the password from the Token List.</source>
</trans-unit>
<trans-unit id="s4414164d120de61a">
<source>The following objects use <x id="0" equiv-text="${objName}"/></source>
</trans-unit>
<trans-unit id="s92e241c9f3c101a2">
<source>connecting object will be deleted</source>
</trans-unit>
<trans-unit id="se6a13beff646557b">
<source>Successfully updated <x id="0" equiv-text="${this.objectLabel} ${this.obj?.name}"/></source>
</trans-unit>
<trans-unit id="s14401ff4a0cba208">
<source>Failed to update <x id="0" equiv-text="${this.objectLabel}"/>: <x id="1" equiv-text="${e.toString()}"/></source>
</trans-unit>
<trans-unit id="sa95a538bfbb86111">
<source>Are you sure you want to update <x id="0" equiv-text="${this.objectLabel}"/> "<x id="1" equiv-text="${this.obj?.name}"/>"?</source>
</trans-unit>
<trans-unit id="sc92d7cfb6ee1fec6">
<source>Successfully updated password.</source>
</trans-unit>
<trans-unit id="se5498954255620b4">
<source>Successfully sent email.</source>
</trans-unit>
<trans-unit id="s44ea4e9a81ce730d">
<source>Email stage</source>
</trans-unit>
<trans-unit id="sdb53ccdd6174e6e3">
<source>Successfully added user(s).</source>
</trans-unit>
<trans-unit id="s306a35df5d0d38bb">
<source>Users to add</source>
</trans-unit>
<trans-unit id="s424f57afae0caac4">
<source>Add users</source>
</trans-unit>
<trans-unit id="s7d499be3b781a3ca">
<source>User(s)</source>
</trans-unit>
<trans-unit id="s7220fcf4fec4e0df">
<source>Remove Users(s)</source>
</trans-unit>
<trans-unit id="s5d7748b1d2363478">
<source>Are you sure you want to remove the selected users from the group <x id="0" equiv-text="${this.targetGroup?.name}"/>?</source>
</trans-unit>
<trans-unit id="sea4f08110bb8f15d">
<source>Remove</source>
</trans-unit>
<trans-unit id="sf466142da6a65052">
<source>Impersonate</source>
</trans-unit>
<trans-unit id="s58888ef1ee9b5bb8">
<source>User status</source>
</trans-unit>
<trans-unit id="s4409ada9c5c2a7f8">
<source>Inactive</source>
</trans-unit>
<trans-unit id="s7ec7036b249f4f22">
<source>Regular user</source>
</trans-unit>
<trans-unit id="sf9e61f4f8e90f0f1">
<source>Change status</source>
</trans-unit>
<trans-unit id="sf56998949bdf6b33">
<source>Deactivate</source>
</trans-unit>
<trans-unit id="s27976e94b05c6970">
<source>Activate</source>
</trans-unit>
<trans-unit id="s3794c596ee7964ad">
<source>Update password</source>
</trans-unit>
<trans-unit id="sce8d867ca5f35304">
<source>Set password</source>
</trans-unit>
<trans-unit id="s0ae3395d8f48e624">
<source>Successfully generated recovery link</source>
</trans-unit>
<trans-unit id="s8ca0dbaec5d48563">
<source>No recovery flow is configured.</source>
</trans-unit>
<trans-unit id="sb69119c9f0547bed">
<source>Copy recovery link</source>
</trans-unit>
<trans-unit id="s7fa236d26b798301">
<source>Send link</source>
</trans-unit>
<trans-unit id="sa9dbe2fb284e26fe">
<source>Send recovery link to user</source>
</trans-unit>
<trans-unit id="s03fd2c252ad7972a">
<source>Email recovery link</source>
</trans-unit>
<trans-unit id="sd7fa99e4d82b374a">
<source>Recovery link cannot be emailed, user has no email address saved.</source>
</trans-unit>
<trans-unit id="sa9d13ce9e83aac17">
<source>To let a user directly reset a their password, configure a recovery flow on the currently active brand.</source>
</trans-unit>
<trans-unit id="s720594461542943f">
<source>Add User</source>
</trans-unit>
<trans-unit id="s4c41f3f4c23e8eaa">
<source>Warning: This group is configured with superuser access. Added users will have superuser access.</source>
</trans-unit>
<trans-unit id="scee721983b1c28d0">
<source>Add existing user</source>
</trans-unit>
<trans-unit id="sd600334ec2c39b74">
<source>Create user</source>
</trans-unit>
<trans-unit id="s53ad3455d9523b54">
<source>Create User</source>
</trans-unit>
<trans-unit id="s824e0943a7104668">
<source>This user will be added to the group "<x id="0" equiv-text="${this.targetGroup.name}"/>".</source>
</trans-unit>
<trans-unit id="s06c163334767a381">
<source>Create Service account</source>
</trans-unit>
<trans-unit id="sc744f3691efe310d">
<source>Hide service-accounts</source>
</trans-unit>
<trans-unit id="secdb4b4c4e66aa38">
<source>Group Info</source>
</trans-unit>
<trans-unit id="s005053d82b712e0a">
<source>Notes</source>
</trans-unit>
<trans-unit id="s634448e4942cf452">
<source>Edit the notes attribute of this group to add notes here.</source>
</trans-unit>
<trans-unit id="s586d6bd2eca2da93">
<source>Users</source>
</trans-unit>
<trans-unit id="s2da4aa7a9abeb653">
<source>Pseudolocale (for testing)</source>
</trans-unit>
<trans-unit id="s4caed5b7a7e5d89b">
<source>English</source>
</trans-unit>
<trans-unit id="sf1868dc19e3917bb">
<source>Spanish</source>
</trans-unit>
<trans-unit id="s63e71d20d1eaca93">
<source>German</source>
</trans-unit>
<trans-unit id="s75a27f43413e02c5">
<source>French</source>
</trans-unit>
<trans-unit id="s03f49e598ffb11cc">
<source>Polish</source>
</trans-unit>
<trans-unit id="s9d2d00982edafabb">
<source>Turkish</source>
</trans-unit>
<trans-unit id="se3e6af2ce24d80e8">
<source>Chinese (traditional)</source>
</trans-unit>
<trans-unit id="s4660da32fb311ac0">
<source>Taiwanese Mandarin</source>
</trans-unit>
<trans-unit id="s354e0a9f146d2869">
<source>Chinese (simplified)</source>
</trans-unit>
<trans-unit id="s7df5b92a3f93544f">
<source>Warning: The current user count has exceeded the configured licenses.</source>
</trans-unit>
<trans-unit id="s0141f42936495787">
<source>Click here for more info.</source>
</trans-unit>
<trans-unit id="s32a3efa23718e713">
<source>API Requests</source>
</trans-unit>
<trans-unit id="sddb3b0176f437721">
<source>Open API Browser</source>
</trans-unit>
<trans-unit id="s3b3c333481944862">
<source>Show details</source>
</trans-unit>
<trans-unit id="s5be3c6d61cd9182f">
<source>Notifications</source>
</trans-unit>
<trans-unit id="sa3438c7bb4e9cce8">
<source><x id="0" equiv-text="${this.unread}"/> unread</source>
</trans-unit>
<trans-unit id="s6e6e737601f44b2c">
<source>Successfully cleared notifications</source>
</trans-unit>
<trans-unit id="s8cda828dac449ea5">
<source>Clear all</source>
</trans-unit>
<trans-unit id="s7031e6928c44cedd">
<source>User interface</source>
</trans-unit>
<trans-unit id="s8849ece8c65e3a18">
<source>Dashboards</source>
</trans-unit>
<trans-unit id="s5e169e1bac20b4a6">
<source>Outposts</source>
</trans-unit>
<trans-unit id="sc265a3e29e1206e4">
<source>Events</source>
</trans-unit>
<trans-unit id="s4f1ad6b48a5df506">
<source>Logs</source>
</trans-unit>
<trans-unit id="s98c3bdf4fd5cdf65">
<source>Notification Rules</source>
</trans-unit>
<trans-unit id="s819509c33a7534ac">
<source>Notification Transports</source>
</trans-unit>
<trans-unit id="s1823625e6f831d73">
<source>Customisation</source>
</trans-unit>
<trans-unit id="saab79cd956ee56a9">
<source>Blueprints</source>
</trans-unit>
<trans-unit id="s1a65ee08832fbfe2">
<source>Flows and Stages</source>
</trans-unit>
<trans-unit id="sb69a4b0acd0895f2">
<source>Flows</source>
</trans-unit>
<trans-unit id="s41b105819b67ee7a">
<source>Stages</source>
</trans-unit>
<trans-unit id="s2be6121210e2a2f8">
<source>Prompts</source>
</trans-unit>
<trans-unit id="sc0829ee663ced008">
<source>Directory</source>
</trans-unit>
<trans-unit id="sc6b4ebd37b7a91c7">
<source>Tokens and App passwords</source>
</trans-unit>
<trans-unit id="s8226f48cb1a80997">
<source>Invitations</source>
</trans-unit>
<trans-unit id="sa81e2cdaf6921adc">
<source>System</source>
</trans-unit>
<trans-unit id="s5515a897ae98bed9">
<source>Certificates</source>
</trans-unit>
<trans-unit id="s6b79e73ca77148a0">
<source>Outpost Integrations</source>
</trans-unit>
<trans-unit id="s9d8b8aa2b404c2c8">
<source>Settings</source>
</trans-unit>
<trans-unit id="s4207178ba0b99418">
<source>A newer version of the frontend is available.</source>
</trans-unit>
<trans-unit id="s96b3cddf33e1c853">
<source>You're currently impersonating <x id="0" equiv-text="${this.impersonation}"/>. Click to stop.</source>
</trans-unit>
<trans-unit id="s7be2df39f727faa2">
<source>Enterprise</source>
</trans-unit>
<trans-unit id="s2e109263b73c12d5">
<source>Licenses</source>
</trans-unit>
<trans-unit id="sca7cfe2bef51b2a5">
<source>Root</source>
</trans-unit>
<trans-unit id="saf6097bfa25205b8">
<source>A copy of this recovery link has been placed in your clipboard</source>
</trans-unit>
<trans-unit id="s6709b81e1ed4e39f">
<source>The current brand must have a recovery flow configured to use a recovery link</source>
</trans-unit>
<trans-unit id="s3616cc78631f5893">
<source>Warning: You're about to delete the user you're logged in as (<x id="0" equiv-text="${shouldShowWarning.username}"/>). Proceed at your own risk.</source>
</trans-unit>
<trans-unit id="s510c7add9e24c306">
<source>Hide deactivated user</source>
</trans-unit>
<trans-unit id="s0924f51b028233a3">
<source><No name set></source>
</trans-unit>
<trans-unit id="s895514dda9cb9c94">
<source>Create recovery link</source>
</trans-unit>
<trans-unit id="s94055b4eb957dc8f">
<source>User folders</source>
</trans-unit>
<trans-unit id="sa982875b258fea07">
<source>Successfully added user to group(s).</source>
</trans-unit>
<trans-unit id="s1bd5920d8adf2bd5">
<source>Groups to add</source>
</trans-unit>
<trans-unit id="sd9f67fbf3f86efcf">
<source>Add group</source>
</trans-unit>
<trans-unit id="s5f71fa3c53828e30">
<source>Remove from Group(s)</source>
</trans-unit>
<trans-unit id="sb4c9ed2a487b238f">
<source>Are you sure you want to remove user <x id="0" equiv-text="${this.targetUser?.username}"/> from the following groups?</source>
</trans-unit>
<trans-unit id="s964f6725aeb7662f">
<source>Add Group</source>
</trans-unit>
<trans-unit id="s65ca2f256ea09c11">
<source>Add to existing group</source>
</trans-unit>
<trans-unit id="s505fbbdcbc6aa921">
<source>Add new group</source>
</trans-unit>
<trans-unit id="s506beb486fa41241">
<source>Application authorizations</source>
</trans-unit>
<trans-unit id="s78ab26da7f067de8">
<source>Select permissions to grant</source>
</trans-unit>
<trans-unit id="sdeb90bfd8a80b86b">
<source>Permissions to add</source>
</trans-unit>
<trans-unit id="s36247910d67421e1">
<source>Select permissions</source>
</trans-unit>
<trans-unit id="s67e136af8fc1107b">
<source>Assign permission</source>
</trans-unit>
<trans-unit id="s1455753daa00f1bc">
<source>User doesn't have view permission so description cannot be retrieved.</source>
</trans-unit>
<trans-unit id="s7301a7069b7bc83e">
<source>Revoked?</source>
</trans-unit>
<trans-unit id="sd924045605feea63">
<source>Expires</source>
</trans-unit>
<trans-unit id="s1c8916418c334935">
<source>ID Token</source>
</trans-unit>
<trans-unit id="s90760e5e02e95dfe">
<source>Refresh Tokens(s)</source>
</trans-unit>
<trans-unit id="s1b88fa3df4423292">
<source>Last IP</source>
</trans-unit>
<trans-unit id="se63f9d833700af49">
<source>Session(s)</source>
</trans-unit>
<trans-unit id="sf679b7a62808287e">
<source>Expiry</source>
</trans-unit>
<trans-unit id="sde1907073fd96017">
<source>(Current session)</source>
</trans-unit>
<trans-unit id="s76881c01b6a3a8c7">
<source>Consent(s)</source>
</trans-unit>
<trans-unit id="sf05c700a1250824e">
<source>Confirmed</source>
</trans-unit>
<trans-unit id="sa0b01f479f40c52d">
<source>Device(s)</source>
</trans-unit>
<trans-unit id="s64a33dcdaf90af26">
<source>User Info</source>
</trans-unit>
<trans-unit id="s028d385389b5aac0">
<source>Lock the user out of this system</source>
</trans-unit>
<trans-unit id="sd2122c514f0778b5">
<source>Allow the user to log in and use this system</source>
</trans-unit>
<trans-unit id="s43fe853bf219a9b8">
<source>Temporarily assume the identity of this user</source>
</trans-unit>
<trans-unit id="se28b5f3fcadaeeb1">
<source>Enter a new password for this user</source>
</trans-unit>
<trans-unit id="s6f5bb31e2733ecd5">
<source>Create a link for this user to reset their password</source>
</trans-unit>
<trans-unit id="se5c795faf2c07514">
<source>Create Recovery Link</source>
</trans-unit>
<trans-unit id="sc44bae5cde0083fa">
<source>Actions over the last week (per 8 hours)</source>
</trans-unit>
<trans-unit id="sb57dbcda1929c642">
<source>Edit the notes attribute of this user to add notes here.</source>
</trans-unit>
<trans-unit id="s5c18cae48b93138c">
<source>Sessions</source>
</trans-unit>
<trans-unit id="s27586544c447d9e3">
<source>User events</source>
</trans-unit>
<trans-unit id="s4d31797d81e9cea3">
<source>Explicit Consent</source>
</trans-unit>
<trans-unit id="sb6770fa90be6d8b3">
<source>OAuth Refresh Tokens</source>
</trans-unit>
<trans-unit id="s28b3de1561da72b3">
<source>MFA Authenticators</source>
</trans-unit>
<trans-unit id="sa3a3e09b88ed9791">
<source>Assigned permissions</source>
</trans-unit>
<trans-unit id="s9cc631505c17b028">
<source>Assigned global permissions</source>
</trans-unit>
<trans-unit id="s8f85a0e678846080">
<source>Assigned object permissions</source>
</trans-unit>
<trans-unit id="s9103a949a3963aa9">
<source>Successfully updated role.</source>
</trans-unit>
<trans-unit id="sdf87c5661b31359e">
<source>Successfully created role.</source>
</trans-unit>
<trans-unit id="s3484b1e6d0b5335f">
<source>Manage roles which grant permissions to objects within authentik.</source>
</trans-unit>
<trans-unit id="s259de999919316db">
<source>Role(s)</source>
</trans-unit>
<trans-unit id="s2ffad156e8332f04">
<source>Update Role</source>
</trans-unit>
<trans-unit id="sc5f923729564fbf3">
<source>Create Role</source>
</trans-unit>
<trans-unit id="s14bfa8fd1bec8889">
<source>Role doesn't have view permission so description cannot be retrieved.</source>
</trans-unit>
<trans-unit id="s7e796fe83982863f">
<source>Role <x id="0" equiv-text="${this._role?.name || ""}"/></source>
</trans-unit>
<trans-unit id="s526e2c66bd51ff5f">
<source>Role Info</source>
</trans-unit>
<trans-unit id="s7a322c89298dd27c">
<source>Successfully updated invitation.</source>
</trans-unit>
<trans-unit id="sc554339ffc7b04e7">
<source>Successfully created invitation.</source>
</trans-unit>
<trans-unit id="sfcebd18506f1e535">
<source>Flow</source>
</trans-unit>
<trans-unit id="sa84a7fd11ba85e88">
<source>When selected, the invite will only be usable with the flow. By default the invite is accepted on all flows with invitation stages.</source>
</trans-unit>
<trans-unit id="sd18b18f91b804c3f">
<source>Custom attributes</source>
</trans-unit>
<trans-unit id="s7520286c8419a266">
<source>Optional data which is loaded into the flow's 'prompt_data' context variable. YAML or JSON.</source>
</trans-unit>
<trans-unit id="sb8795b799c70776a">
<source>Single use</source>
</trans-unit>
<trans-unit id="sf232d42142eacc23">
<source>When enabled, the invitation will be deleted after usage.</source>
</trans-unit>
<trans-unit id="sa4a8086275475714">
<source>Select an enrollment flow</source>
</trans-unit>
<trans-unit id="s839cb09cb2193da9">
<source>Link to use the invitation.</source>
</trans-unit>
<trans-unit id="s57448f10eb973100">
<source>Create Invitation Links to enroll Users, and optionally force specific attributes of their account.</source>
</trans-unit>
<trans-unit id="s4aee34a672e5cfc0">
<source>Created by</source>
</trans-unit>
<trans-unit id="sd5ba2d61ee4796fe">
<source>Invitation(s)</source>
</trans-unit>
<trans-unit id="s96dcf7ec8342c335">
<source>Invitation not limited to any flow, and can be used with any enrollment flow.</source>
</trans-unit>
<trans-unit id="s1b42b49e7b392013">
<source>Update Invitation</source>
</trans-unit>
<trans-unit id="s38c72e1cf120b8d8">
<source>Create Invitation</source>
</trans-unit>
<trans-unit id="s802826db4e2c852e">
<source>Warning: No invitation stage is bound to any flow. Invitations will not work as expected.</source>
</trans-unit>
<trans-unit id="s2f995efbb1e46b18">
<source>Auto-detect (based on your browser)</source>
</trans-unit>
<trans-unit id="s296fbffaaa7c910a">
<source>Required.</source>
</trans-unit>
<trans-unit id="s81ecf2d4386b8e84">
<source>Continue</source>
</trans-unit>
<trans-unit id="s8b2b2a43fcf688a3">
<source>Successfully updated prompt.</source>
</trans-unit>
<trans-unit id="s5572ac4d2208f5ec">
<source>Successfully created prompt.</source>
</trans-unit>
<trans-unit id="s54e7a23a95d99649">
<source>Text: Simple Text input</source>
</trans-unit>
<trans-unit id="s63e54b86e2a2cc43">
<source>Text Area: Multiline text input</source>
</trans-unit>
<trans-unit id="s12de1c06a1e18cc5">
<source>Text (read-only): Simple Text input, but cannot be edited.</source>
</trans-unit>
<trans-unit id="s4e5646b23e41231f">
<source>Text Area (read-only): Multiline text input, but cannot be edited.</source>
</trans-unit>
<trans-unit id="s1e4c3de6e12cd87b">
<source>Username: Same as Text input, but checks for and prevents duplicate usernames.</source>
</trans-unit>
<trans-unit id="s5462c7f56ed65e6c">
<source>Email: Text field with Email type.</source>
</trans-unit>
<trans-unit id="s1c5574968b29ab1c">
<source>Password: Masked input, multiple inputs of this type on the same prompt need to be identical.</source>
</trans-unit>
<trans-unit id="sbbb97b1c63507dc0">
<source>Number</source>
</trans-unit>
<trans-unit id="sdae649fae731e838">
<source>Checkbox</source>
</trans-unit>
<trans-unit id="s34edeb18f887161d">
<source>Radio Button Group (fixed choice)</source>
</trans-unit>
<trans-unit id="s57730b6870e8916c">
<source>Dropdown (fixed choice)</source>
</trans-unit>
<trans-unit id="sac8252732f2edb19">
<source>Date</source>
</trans-unit>
<trans-unit id="s45960273852a61b2">
<source>Date Time</source>
</trans-unit>
<trans-unit id="sd1f81284eeb7b503">
<source>File</source>
</trans-unit>
<trans-unit id="s21e3c227cc2c5873">
<source>Separator: Static Separator Line</source>
</trans-unit>
<trans-unit id="s706af57c1af42c6d">
<source>Hidden: Hidden field, can be used to insert data into form.</source>
</trans-unit>
<trans-unit id="s40e2c72dae905a50">
<source>Static: Static value, displayed as-is.</source>
</trans-unit>
<trans-unit id="sdd4bd4224c4e943d">
<source>authentik: Locale: Displays a list of locales authentik supports.</source>
</trans-unit>
<trans-unit id="saf84e7732a9e1336">
<source>Preview errors</source>
</trans-unit>
<trans-unit id="sb71ace8e9b35c749">
<source>Data preview</source>
</trans-unit>
<trans-unit id="s4d53f4b7ff33bedd">
<source>Unique name of this field, used for selecting fields in prompt stages.</source>
</trans-unit>
<trans-unit id="s3b58f8d2155ae90c">
<source>Field Key</source>
</trans-unit>
<trans-unit id="s2b088ba65eb69b7e">
<source>Name of the form field, also used to store the value.</source>
</trans-unit>
<trans-unit id="s662fcb3761ad9df7">
<source>When used in conjunction with a User Write stage, use attributes.foo to write attributes.</source>
</trans-unit>
<trans-unit id="s5590dbf7e425789d">
<source>Label</source>
</trans-unit>
<trans-unit id="s0c135eba6017d94f">
<source>Label shown next to/above the prompt.</source>
</trans-unit>
<trans-unit id="sae5d87e99fe081e0">
<source>Required</source>
</trans-unit>
<trans-unit id="s37dbfe2133b74d2d">
<source>Interpret placeholder as expression</source>
</trans-unit>
<trans-unit id="s4a953e6234cb4808">
<source>When checked, the placeholder will be evaluated in the same way a property mapping is.
If the evaluation fails, the placeholder itself is returned.</source>
</trans-unit>
<trans-unit id="sf90be97cb08f3d5a">
<source>Placeholder</source>
</trans-unit>
<trans-unit id="sf76ead4c4708dd06">
<source>Optionally provide a short hint that describes the expected input value.
When creating a fixed choice field, enable interpreting as expression and return a
list to return multiple choices.</source>
</trans-unit>
<trans-unit id="saa7ba2057bd524a1">
<source>Interpret initial value as expression</source>
</trans-unit>
<trans-unit id="sd60415c7666859f0">
<source>When checked, the initial value will be evaluated in the same way a property mapping is.
If the evaluation fails, the initial value itself is returned.</source>
</trans-unit>
<trans-unit id="sa9c7044d9fd1f3e6">
<source>Initial value</source>
</trans-unit>
<trans-unit id="seab35681cbf36755">
<source>Optionally pre-fill the input with an initial value.
When creating a fixed choice field, enable interpreting as expression and
return a list to return multiple default choices.</source>
</trans-unit>
<trans-unit id="s72c1c17a9bdc76ad">
<source>Help text</source>
</trans-unit>
<trans-unit id="s584d1c38ad20d560">
<source>Any HTML can be used.</source>
</trans-unit>
<trans-unit id="s42fc6f4b64eff5d9">
<source>Single Prompts that can be used for Prompt Stages.</source>
</trans-unit>
<trans-unit id="s42a1ebe17efda727">
<source>Field</source>
</trans-unit>
<trans-unit id="sec7443a45fd141e5">
<source>Prompt(s)</source>
</trans-unit>
<trans-unit id="scc733ba98740038a">
<source>Update Prompt</source>
</trans-unit>
<trans-unit id="s61b6f3e6bc59c6dd">
<source>Create Prompt</source>
</trans-unit>
<trans-unit id="sff5bb7742c2896c8">
<source>Target</source>
</trans-unit>
<trans-unit id="sae5da213b7f896ed">
<source>Stage</source>
</trans-unit>
<trans-unit id="s0a61796c1956d32c">
<source>Evaluate when flow is planned</source>
</trans-unit>
<trans-unit id="sf533f13321fee530">
<source>Evaluate policies during the Flow planning process.</source>
</trans-unit>
<trans-unit id="s6336fa345e96dde9">
<source>Evaluate when stage is run</source>
</trans-unit>
<trans-unit id="sff3b708e23bb96b2">
<source>Evaluate policies before the Stage is present to the user.</source>
</trans-unit>
<trans-unit id="s0dc46deb8f181baf">
<source>Invalid response behavior</source>
</trans-unit>
<trans-unit id="seb0805249661d15b">
<source>Returns the error message and a similar challenge to the executor</source>
</trans-unit>
<trans-unit id="sd891d8463d0ebace">
<source>Restarts the flow from the beginning</source>
</trans-unit>
<trans-unit id="s6b9a1dd402750a8a">
<source>Restarts the flow from the beginning, while keeping the flow context</source>
</trans-unit>
<trans-unit id="sbc88fb27a4c3b894">
<source>Configure how the flow executor should handle an invalid response to a challenge given by this bound stage.</source>
</trans-unit>
<trans-unit id="s916b32ac64ea2b05">
<source>Successfully updated stage.</source>
</trans-unit>
<trans-unit id="s14c8f36e180d6bbc">
<source>Successfully created stage.</source>
</trans-unit>
<trans-unit id="sf22a28f83cc45fcc">
<source>Stage used to configure a duo-based authenticator. This stage should be used for configuration flows.</source>
</trans-unit>
<trans-unit id="s5adafce329aaa853">
<source>Authenticator type name</source>
</trans-unit>
<trans-unit id="s23e6a57201fba25e">
<source>Display name of this authenticator, used by users when they enroll an authenticator.</source>
</trans-unit>
<trans-unit id="s276d751eb7a186cc">
<source>API Hostname</source>
</trans-unit>
<trans-unit id="s5b6b6e2cb884d59f">
<source>Duo Auth API</source>
</trans-unit>
<trans-unit id="s240ff02ce3a53dee">
<source>Integration key</source>
</trans-unit>
<trans-unit id="s56fd9ed596c724fa">
<source>Secret key</source>
</trans-unit>
<trans-unit id="s88870d7e499e848b">
<source>Duo Admin API (optional)</source>
</trans-unit>
<trans-unit id="s7f13f4a2d0370cf6">
<source>When using a Duo MFA, Access or Beyond plan, an Admin API application can be created.
This will allow authentik to import devices automatically.</source>
</trans-unit>
<trans-unit id="s9a34d1520e320465">
<source>Stage-specific settings</source>
</trans-unit>
<trans-unit id="s0dfc6838c9d07677">
<source>Configuration flow</source>
</trans-unit>
<trans-unit id="sebf44d2471b608ad">
<source>Flow used by an authenticated user to configure this Stage. If empty, user will not be able to configure this stage.</source>
</trans-unit>
<trans-unit id="s3baf512851453712">
<source>Twilio Account SID</source>
</trans-unit>
<trans-unit id="sa738ce390bc24875">
<source>Get this value from https://console.twilio.com</source>
</trans-unit>
<trans-unit id="sa7b56a80ab1801f0">
<source>Twilio Auth Token</source>
</trans-unit>
<trans-unit id="sfe99a8caa70232ab">
<source>Authentication Type</source>
</trans-unit>
<trans-unit id="safd0363143a46a91">
<source>Basic Auth</source>
</trans-unit>
<trans-unit id="sd06b47084fec0ec5">
<source>Bearer Token</source>
</trans-unit>
<trans-unit id="sb1751a1411d6874f">
<source>External API URL</source>
</trans-unit>
<trans-unit id="sbdc1176ff9f93da2">
<source>This is the full endpoint to send POST requests to.</source>
</trans-unit>
<trans-unit id="s51da4de00984fe51">
<source>API Auth Username</source>
</trans-unit>
<trans-unit id="s293ab4331c1dd387">
<source>This is the username to be used with basic auth or the token when used with bearer token</source>
</trans-unit>
<trans-unit id="s634d041fd954ab20">
<source>API Auth password</source>
</trans-unit>
<trans-unit id="sb635ad3c2e357d3c">
<source>This is the password to be used with basic auth</source>
</trans-unit>
<trans-unit id="sa92398dba8b12d85">
<source>Mapping</source>
</trans-unit>
<trans-unit id="s38162f615710c7b4">
<source>Modify the payload sent to the custom provider.</source>
</trans-unit>
<trans-unit id="s5e830ae7688d1219">
<source>Stage used to configure an SMS-based TOTP authenticator.</source>
</trans-unit>
<trans-unit id="s0d5d05bf3d122ced">
<source>Twilio</source>
</trans-unit>
<trans-unit id="sc3c74f5273df459a">
<source>Generic</source>
</trans-unit>
<trans-unit id="sbbb2180b6aed196e">
<source>From number</source>
</trans-unit>
<trans-unit id="sc647dcb91f6958dd">
<source>Number the SMS will be sent from.</source>
</trans-unit>
<trans-unit id="s0ae0072614320ae2">
<source>Hash phone number</source>
</trans-unit>
<trans-unit id="s9ca3310e1999fd5b">
<source>If enabled, only a hash of the phone number will be saved. This can be done for data-protection reasons. Devices created from a stage with this enabled cannot be used with the authenticator validation stage.</source>
</trans-unit>
<trans-unit id="s128e7f5f34bfa155">
<source>Stage used to configure a static authenticator (i.e. static tokens). This stage should be used for configuration flows.</source>
</trans-unit>
<trans-unit id="sabf67834e35dede5">
<source>Token count</source>
</trans-unit>
<trans-unit id="s7caa8f7edb920909">
<source>The number of tokens generated whenever this stage is used. Every token generated per stage execution will be attached to a single static device.</source>
</trans-unit>
<trans-unit id="s4aacc4e0277c1042">
<source>Token length</source>
</trans-unit>
<trans-unit id="s6931695c4f563bc4">
<source>The length of the individual generated tokens. Can be increased to improve security.</source>
</trans-unit>
<trans-unit id="sc5a4711395ffb043">
<source>Stage used to configure a TOTP authenticator (i.e. Authy/Google Authenticator).</source>
</trans-unit>
<trans-unit id="s9d8ad4b85287131f">
<source>Digits</source>
</trans-unit>
<trans-unit id="sc04e92d753742189">
<source>6 digits, widely compatible</source>
</trans-unit>
<trans-unit id="sdc70195469e83e3f">
<source>8 digits, not compatible with apps like Google Authenticator</source>
</trans-unit>
<trans-unit id="s0e15f678445dfc45">
<source>Stage used to validate any authenticator. This stage should be used during authentication or authorization flows.</source>
</trans-unit>
<trans-unit id="s73c13e5a6f5e38a3">
<source>Device classes</source>
</trans-unit>
<trans-unit id="s97d1b0070f50c07f">
<source>Static Tokens</source>
</trans-unit>
<trans-unit id="sb8168ae309c66abc">
<source>TOTP Authenticators</source>
</trans-unit>
<trans-unit id="sde47e4d8b9b21b59">
<source>WebAuthn Authenticators</source>
</trans-unit>
<trans-unit id="s8da88a8a5750bce1">
<source>Duo Authenticators</source>
</trans-unit>
<trans-unit id="s4d182bae8a578010">
<source>SMS-based Authenticators</source>
</trans-unit>
<trans-unit id="sd8d9451f86502d1a">
<source>Device classes which can be used to authenticate.</source>
</trans-unit>
<trans-unit id="se2e9f5a32c93e5f7">
<source>Last validation threshold</source>
</trans-unit>
<trans-unit id="s951281efc92b03fc">
<source>If any of the devices user of the types selected above have been used within this duration, this stage will be skipped.</source>
</trans-unit>
<trans-unit id="s681074b6c1f19c08">
<source>Not configured action</source>
</trans-unit>
<trans-unit id="sa2c29dc5ed47b26d">
<source>Force the user to configure an authenticator</source>
</trans-unit>
<trans-unit id="sa30c58514a3dc0fb">
<source>Deny the user access</source>
</trans-unit>
<trans-unit id="s1e0de9c4f66dc371">
<source>WebAuthn User verification</source>
</trans-unit>
<trans-unit id="sdb7b2173869822bc">
<source>User verification must occur.</source>
</trans-unit>
<trans-unit id="s7683363cdf78cf31">
<source>User verification is preferred if available, but not required.</source>
</trans-unit>
<trans-unit id="scb43f5faeb6a7ca9">
<source>User verification should not occur.</source>
</trans-unit>
<trans-unit id="scae166352a31032c">
<source>Configuration stages</source>
</trans-unit>
<trans-unit id="s6941a67f0038ba4c">
<source>Stages used to configure Authenticator when user doesn't have any compatible devices. After this configuration Stage passes, the user is not prompted again.</source>
</trans-unit>
<trans-unit id="s7e5af9c6ba6f5cc6">
<source>When multiple stages are selected, the user can choose which one they want to enroll.</source>
</trans-unit>
<trans-unit id="s24bce955914b1f0a">
<source>Stage used to configure a WebAuthn authenticator (i.e. Yubikey, FaceID/Windows Hello).</source>
</trans-unit>
<trans-unit id="s34b23ebbac9f6ab9">
<source>User verification</source>
</trans-unit>
<trans-unit id="s6a3cf855140b9511">
<source>Required: User verification must occur.</source>
</trans-unit>
<trans-unit id="sc498a3b05cfe2b08">
<source>Preferred: User verification is preferred if available, but not required.</source>
</trans-unit>
<trans-unit id="s9d2239d2b0402795">
<source>Discouraged: User verification should not occur.</source>
</trans-unit>
<trans-unit id="s9ea472b555374771">
<source>Resident key requirement</source>
</trans-unit>
<trans-unit id="s428b7859907f6db2">
<source>Required: The authenticator MUST create a dedicated credential. If it cannot, the RP is prepared for an error to occur</source>
</trans-unit>
<trans-unit id="s33e3766d4a02b042">
<source>Preferred: The authenticator can create and store a dedicated credential, but if it doesn't that's alright too</source>
</trans-unit>
<trans-unit id="sfb852dd507c25c24">
<source>Discouraged: The authenticator should not create a dedicated credential</source>
</trans-unit>
<trans-unit id="s5fbaeb14f42815e5">
<source>Authenticator Attachment</source>
</trans-unit>
<trans-unit id="s502d2473587032e1">
<source>No preference is sent</source>
</trans-unit>
<trans-unit id="s60cc554fde2676cb">
<source>A non-removable authenticator, like TouchID or Windows Hello</source>
</trans-unit>
<trans-unit id="sdf1d8edef27236f0">
<source>A "roaming" authenticator, like a YubiKey</source>
</trans-unit>
<trans-unit id="sfffba7b23d8fb40c">
<source>This stage checks the user's current session against the Google reCaptcha (or compatible) service.</source>
</trans-unit>
<trans-unit id="sfd1af96798dd8a5f">
<source>Public Key</source>
</trans-unit>
<trans-unit id="sf339673f0f76a8bd">
<source>Public key, acquired from https://www.google.com/recaptcha/intro/v3.html.</source>
</trans-unit>
<trans-unit id="s83d0f62ad1731a03">
<source>Private Key</source>
</trans-unit>
<trans-unit id="s892d2731a6f22e59">
<source>Private key, acquired from https://www.google.com/recaptcha/intro/v3.html.</source>
</trans-unit>
<trans-unit id="scb6620fcd5bff04c">
<source>Advanced settings</source>
</trans-unit>
<trans-unit id="s39e436de1dc4df4f">
<source>JS URL</source>
</trans-unit>
<trans-unit id="s170b705c55ecb2ae">
<source>URL to fetch JavaScript from, defaults to recaptcha. Can be replaced with any compatible alternative.</source>
</trans-unit>
<trans-unit id="s275021658614ce9e">
<source>API URL</source>
</trans-unit>
<trans-unit id="sc8a79fddea3ab4a9">
<source>URL used to validate captcha response, defaults to recaptcha. Can be replaced with any compatible alternative.</source>
</trans-unit>
<trans-unit id="s1cd617e7bbe278d0">
<source>Prompt for the user's consent. The consent can either be permanent or expire in a defined amount of time.</source>
</trans-unit>
<trans-unit id="s26513c9dd154f041">
<source>Always require consent</source>
</trans-unit>
<trans-unit id="s8ce8bdc9cc9c8604">
<source>Consent given last indefinitely</source>
</trans-unit>
<trans-unit id="sb986f15fa9b17805">
<source>Consent expires.</source>
</trans-unit>
<trans-unit id="s6f328f2d8382d998">
<source>Consent expires in</source>
</trans-unit>
<trans-unit id="se0c660020d9cf5b7">
<source>Offset after which consent expires.</source>
</trans-unit>
<trans-unit id="s7f68101a50f526ee">
<source>Statically deny the flow. To use this stage effectively, disable *Evaluate when flow is planned* on the respective binding.</source>
</trans-unit>
<trans-unit id="s1cc306d8e28c4464">
<source>Deny message</source>
</trans-unit>
<trans-unit id="s6985c401e1100122">
<source>Message shown when this stage is run.</source>
</trans-unit>
<trans-unit id="s22b10ed263b96194">
<source>Dummy stage used for testing. Shows a simple continue button and always passes.</source>
</trans-unit>
<trans-unit id="sdb861d9906f18ac2">
<source>Throw error?</source>
</trans-unit>
<trans-unit id="s31ebc5431d677f5d">
<source>SMTP Host</source>
</trans-unit>
<trans-unit id="s289fce7e694b98ac">
<source>SMTP Port</source>
</trans-unit>
<trans-unit id="se4a9da0295597e73">
<source>SMTP Username</source>
</trans-unit>
<trans-unit id="s593db2c00d6516a2">
<source>SMTP Password</source>
</trans-unit>
<trans-unit id="s0d4268408182491d">
<source>Use TLS</source>
</trans-unit>
<trans-unit id="s480c6c40a248f7d2">
<source>Use SSL</source>
</trans-unit>
<trans-unit id="sc1feadd25659c94d">
<source>From address</source>
</trans-unit>
<trans-unit id="sa248e1021d2c27b5">
<source>Verify the user's email address by sending them a one-time-link. Can also be used for recovery to verify the user's authenticity.</source>
</trans-unit>
<trans-unit id="s87b7e3bc944c728c">
<source>Activate pending user on success</source>
</trans-unit>
<trans-unit id="s9e9c8d99f4c26baf">
<source>When a user returns from the email successfully, their account will be activated.</source>
</trans-unit>
<trans-unit id="s618d4e53f455c834">
<source>Use global settings</source>
</trans-unit>
<trans-unit id="sae1e1a59d22609c4">
<source>When enabled, global Email connection settings will be used and connection settings below will be ignored.</source>
</trans-unit>
<trans-unit id="sb1fe947f9ad27b9d">
<source>Token expiry</source>
</trans-unit>
<trans-unit id="s1c6ba8d100453392">
<source>Time in minutes the token sent is valid.</source>
</trans-unit>
<trans-unit id="se47baf2fd16b9d2b">
<source>Template</source>
</trans-unit>
<trans-unit id="s4af8a3ce5a600855">
<source>Let the user identify themselves with their username or Email address.</source>
</trans-unit>
<trans-unit id="s592ab7d2bc1b8973">
<source>User fields</source>
</trans-unit>
<trans-unit id="s61e48919db20538a">
<source>UPN</source>
</trans-unit>
<trans-unit id="s4cdae7635e757555">
<source>Fields a user can identify themselves with. If no fields are selected, the user will only be able to use sources.</source>
</trans-unit>
<trans-unit id="s3380d7cbcebe50f6">
<source>Password stage</source>
</trans-unit>
<trans-unit id="s08c91cb1a2cd3d97">
<source>When selected, a password field is shown on the same page instead of a separate page. This prevents username enumeration attacks.</source>
</trans-unit>
<trans-unit id="sd97d8d0906e6cc47">
<source>Case insensitive matching</source>
</trans-unit>
<trans-unit id="s8aaad223e954f9ca">
<source>When enabled, user fields are matched regardless of their casing.</source>
</trans-unit>
<trans-unit id="s62e7f6ed7d9cb3ca">
<source>Pretend user exists</source>
</trans-unit>
<trans-unit id="s52bdc80690a9a8dc">
<source>When enabled, the stage will always accept the given user identifier and continue.</source>
</trans-unit>
<trans-unit id="sbab723b98dcfe23f">
<source>Show matched user</source>
</trans-unit>
<trans-unit id="se50a08ab71bb96ed">
<source>When a valid username/email has been entered, and this option is enabled, the user's username and avatar will be shown. Otherwise, the text that the user entered will be shown.</source>
</trans-unit>
<trans-unit id="s0295ce5d6f635d75">
<source>Source settings</source>
</trans-unit>
<trans-unit id="s91e3a47599412f51">
<source>Sources</source>
</trans-unit>
<trans-unit id="s17a679298216aca9">
<source>Select sources should be shown for users to authenticate with. This only affects web-based sources, not LDAP.</source>
</trans-unit>
<trans-unit id="sa41aee3ae04c9216">
<source>Show sources' labels</source>
</trans-unit>
<trans-unit id="s54cd35e6224ba65d">
<source>By default, only icons are shown for sources. Enable this to show their full names.</source>
</trans-unit>
<trans-unit id="s9ee20003cb116abf">
<source>Passwordless flow</source>
</trans-unit>
<trans-unit id="s0c8c4d2bb0a9162a">
<source>Optional passwordless flow, which is linked at the bottom of the page. When configured, users can use this flow to authenticate with a WebAuthn authenticator, without entering any details.</source>
</trans-unit>
<trans-unit id="s01a3a7f48ee4edaf">
<source>Optional enrollment flow, which is linked at the bottom of the page.</source>
</trans-unit>
<trans-unit id="s82188c9542510212">
<source>Optional recovery flow, which is linked at the bottom of the page.</source>
</trans-unit>
<trans-unit id="s3e59b8b2debf0209">
<source>This stage can be included in enrollment flows to accept invitations.</source>
</trans-unit>
<trans-unit id="s79ad406777feab1f">
<source>Continue flow without invitation</source>
</trans-unit>
<trans-unit id="s61ccefd661ac2296">
<source>If this flag is set, this Stage will jump to the next Stage when no Invitation is given. By default this Stage will cancel the Flow when no invitation is given.</source>
</trans-unit>
<trans-unit id="sdc30bddeda2f0225">
<source>Validate the user's password against the selected backend(s).</source>
</trans-unit>
<trans-unit id="sb8d4f44a1d5b9a14">
<source>Backends</source>
</trans-unit>
<trans-unit id="sba42248f3f27955c">
<source>User database + standard password</source>
</trans-unit>
<trans-unit id="s3330adb3f0922f7b">
<source>User database + app passwords</source>
</trans-unit>
<trans-unit id="sc10db51c9bb77d5c">
<source>User database + LDAP password</source>
</trans-unit>
<trans-unit id="sd35ae4be63df1f9f">
<source>Selection of backends to test the password against.</source>
</trans-unit>
<trans-unit id="s482ae78809a6822b">
<source>Flow used by an authenticated user to configure their password. If empty, user will not be able to configure change their password.</source>
</trans-unit>
<trans-unit id="s77994108c886b965">
<source>Failed attempts before cancel</source>
</trans-unit>
<trans-unit id="sa9020b93c3bd7235">
<source>How many attempts a user has before the flow is canceled. To lock the user out, use a reputation policy and a user_write stage.</source>
</trans-unit>
<trans-unit id="s5170f9ef331949c0">
<source>Show arbitrary input fields to the user, for example during enrollment. Data is saved in the flow context under the 'prompt_data' variable.</source>
</trans-unit>
<trans-unit id="s36cb242ac90353bc">
<source>Fields</source>
</trans-unit>
<trans-unit id="s2d5f69929bb7221d">
<source><x id="0" equiv-text="${prompt.name}"/> ("<x id="1" equiv-text="${prompt.fieldKey}"/>", of type <x id="2" equiv-text="${prompt.type}"/>)</source>
</trans-unit>
<trans-unit id="s3b7b519444181264">
<source>Validation Policies</source>
</trans-unit>
<trans-unit id="s59691290a232c687">
<source>Selected policies are executed when the stage is submitted to validate the data.</source>
</trans-unit>
<trans-unit id="sbf4ef82e04772a4e">
<source>Delete the currently pending user. CAUTION, this stage does not ask for confirmation. Use a consent stage to ensure the user is aware of their actions.</source>
</trans-unit>
<trans-unit id="s8cc920e6a8430a0d">
<source>Log the currently pending user in.</source>
</trans-unit>
<trans-unit id="sb85ffe141d7c229d">
<source>Session duration</source>
</trans-unit>
<trans-unit id="sece294cd51a85745">
<source>Determines how long a session lasts. Default of 0 seconds means that the sessions lasts until the browser is closed.</source>
</trans-unit>
<trans-unit id="sf7949fbbab2eb566">
<source>Different browsers handle session cookies differently, and might not remove them even when the browser is closed.</source>
</trans-unit>
<trans-unit id="s53bbc3ae4b5fa1d0">
<source>See here.</source>
</trans-unit>
<trans-unit id="s2512334108f06a5a">
<source>Stay signed in offset</source>
</trans-unit>
<trans-unit id="s1608b2f94fa0dbd4">
<source>If set to a duration above 0, the user will have the option to choose to "stay signed in", which will extend their session by the time specified here.</source>
</trans-unit>
<trans-unit id="s542a71bb8f41e057">
<source>Terminate other sessions</source>
</trans-unit>
<trans-unit id="sa920231366378c90">
<source>When enabled, all previous sessions of the user will be terminated.</source>
</trans-unit>
<trans-unit id="sfee06600c15082a9">
<source>Remove the user from the current session.</source>
</trans-unit>
<trans-unit id="s927398c400970760">
<source>Write any data from the flow's context's 'prompt_data' to the currently pending user. If no user
is pending, a new user is created, and data is written to them.</source>
</trans-unit>
<trans-unit id="sb379d861cbed0b47">
<source>Never create users</source>
</trans-unit>
<trans-unit id="s81d673755a86a4f0">
<source>When no user is present in the flow context, the stage will fail.</source>
</trans-unit>
<trans-unit id="s9940e3f073fbdbd4">
<source>Create users when required</source>
</trans-unit>
<trans-unit id="s5414356cc10e80fe">
<source>When no user is present in the the flow context, a new user is created.</source>
</trans-unit>
<trans-unit id="s57337099d96ce6d2">
<source>Always create new users</source>
</trans-unit>
<trans-unit id="se80dd66f23b4fc39">
<source>Create a new user even if a user is in the flow context.</source>
</trans-unit>
<trans-unit id="sed3512fe4560c7f4">
<source>Create users as inactive</source>
</trans-unit>
<trans-unit id="s9193ef1a39a6c872">
<source>Mark newly created users as inactive.</source>
</trans-unit>
<trans-unit id="s89d1847b5e4ad225">
<source>User path template</source>
</trans-unit>
<trans-unit id="s32babfed740fd3c1">
<source>User type used for newly created users.</source>
</trans-unit>
<trans-unit id="s18269e3889d6fa54">
<source>Path new users will be created under. If left blank, the default path will be used.</source>
</trans-unit>
<trans-unit id="sc1cb0eef9ed94e6a">
<source>Newly created users are added to this group, if a group is selected.</source>
</trans-unit>
<trans-unit id="sd8417b41ca27bc8f">
<source>New stage</source>
</trans-unit>
<trans-unit id="s293801033f9fc0d0">
<source>Create a new stage.</source>
</trans-unit>
<trans-unit id="s71633a67e0d7c0e4">
<source>Successfully imported device.</source>
</trans-unit>
<trans-unit id="s7d61705dfb120d7b">
<source>The user in authentik this device will be assigned to.</source>
</trans-unit>
<trans-unit id="s5eaf1d304e03ed4b">
<source>Duo User ID</source>
</trans-unit>
<trans-unit id="s003847d8bc01c676">
<source>The user ID in Duo, can be found in the URL after clicking on a user.</source>
</trans-unit>
<trans-unit id="sbbc806ea3987c781">
<source>Automatic import</source>
</trans-unit>
<trans-unit id="s77299a9d3dd932cd">
<source>Successfully imported <x id="0" equiv-text="${res.count}"/> devices.</source>
</trans-unit>
<trans-unit id="s6a615f6165ef01c9">
<source>Start automatic import</source>
</trans-unit>
<trans-unit id="s9f83d7768aea548a">
<source>Or manually import</source>
</trans-unit>
<trans-unit id="sddc8efe94cb8c210">
<source>Stages are single steps of a Flow that a user is guided through. A stage can only be executed from within a flow.</source>
</trans-unit>
<trans-unit id="s0eaf755fa88c8d97">
<source>Stage(s)</source>
</trans-unit>
<trans-unit id="s3914cb410fca44d4">
<source>Import</source>
</trans-unit>
<trans-unit id="s8a67b33a0d70d322">
<source>Import Duo device</source>
</trans-unit>
<trans-unit id="s254a9a23dc1635df">
<source>Import devices</source>
</trans-unit>
<trans-unit id="s48cf8fd56b1237ed">
<source>Successfully updated flow.</source>
</trans-unit>
<trans-unit id="sc3e0c240b159fbce">
<source>Successfully created flow.</source>
</trans-unit>
<trans-unit id="s925936f647ae52cc">
<source>Shown as the Title in Flow pages.</source>
</trans-unit>
<trans-unit id="s50719dda8f90abf4">
<source>Visible in the URL.</source>
</trans-unit>
<trans-unit id="s0f4c6540c30bd8b4">
<source>Designation</source>
</trans-unit>
<trans-unit id="sb25d9afe10941425">
<source>Decides what this Flow is used for. For example, the Authentication flow is redirect to when an un-authenticated user visits authentik.</source>
</trans-unit>
<trans-unit id="sb36e4c05244278c1">
<source>No requirement</source>
</trans-unit>
<trans-unit id="s7b105164d209f670">
<source>Require authentication</source>
</trans-unit>
<trans-unit id="s239c2a351cde6d39">
<source>Require no authentication.</source>
</trans-unit>
<trans-unit id="s98beadfeeb3acb66">
<source>Require superuser.</source>
</trans-unit>
<trans-unit id="s047a5f0211fedc72">
<source>Require Outpost (flow can only be executed from an outpost).</source>
</trans-unit>
<trans-unit id="sfad9279cc42c6b61">
<source>Required authentication level for this flow.</source>
</trans-unit>
<trans-unit id="sb56674c9ea4f0588">
<source>Behavior settings</source>
</trans-unit>
<trans-unit id="sb6d7d58cb0a1544e">
<source>Compatibility mode</source>
</trans-unit>
<trans-unit id="s14ace18ccf4fb86d">
<source>Increases compatibility with password managers and mobile devices.</source>
</trans-unit>
<trans-unit id="scfbc2f1396ee8550">
<source>Denied action</source>
</trans-unit>
<trans-unit id="sff38031cf061e3ae">
<source>Will follow the ?next parameter if set, otherwise show a message</source>
</trans-unit>
<trans-unit id="s936bf4342b182ad4">
<source>Will either follow the ?next parameter or redirect to the default interface</source>
</trans-unit>
<trans-unit id="s22b0e8c5277dd5a9">
<source>Will notify the user the flow isn't applicable</source>
</trans-unit>
<trans-unit id="s2eeca5cfc99ef19b">
<source>Decides the response when a policy denies access to this flow for a user.</source>
</trans-unit>
<trans-unit id="sbaf20067de176c90">
<source>Appearance settings</source>
</trans-unit>
<trans-unit id="s2e4818861000b13f">
<source>Layout</source>
</trans-unit>
<trans-unit id="s1efbfc3937d565bd">
<source>Background</source>
</trans-unit>
<trans-unit id="s374abf1a54d87b67">
<source>Background shown during execution.</source>
</trans-unit>
<trans-unit id="s3ebf4c166058afce">
<source>Clear background</source>
</trans-unit>
<trans-unit id="sb24755ea94bef31d">
<source>Delete currently set background image.</source>
</trans-unit>
<trans-unit id="sb904f23f17b60c3a">
<source>Successfully imported flow.</source>
</trans-unit>
<trans-unit id="s344c4a2a48997e18">
<source>.yaml files, which can be found on goauthentik.io and can be exported by authentik.</source>
</trans-unit>
<trans-unit id="sc816360d6f5a1eeb">
<source>Flows describe a chain of Stages to authenticate, enroll or recover a user. Stages are chosen based on policies applied to them.</source>
</trans-unit>
<trans-unit id="s6f857299d5db1ecf">
<source>Flow(s)</source>
</trans-unit>
<trans-unit id="s9e830cbc0b42a514">
<source>Update Flow</source>
</trans-unit>
<trans-unit id="sc4fdeccf14be5378">
<source>Execute</source>
</trans-unit>
<trans-unit id="s293aa6a6446fb153">
<source>Export</source>
</trans-unit>
<trans-unit id="s2f1bcfcc5cae94c3">
<source>Create Flow</source>
</trans-unit>
<trans-unit id="s832282d415294df4">
<source>Import Flow</source>
</trans-unit>
<trans-unit id="s098237f7ccb4dc4a">
<source>Successfully cleared flow cache</source>
</trans-unit>
<trans-unit id="s59572c1be31a812e">
<source>Failed to delete flow cache</source>
</trans-unit>
<trans-unit id="sa2b727168b090d34">
<source>Clear Flow cache</source>
</trans-unit>
<trans-unit id="sf12d588a76ba7e51">
<source>Are you sure you want to clear the flow cache?
This will cause all flows to be re-evaluated on their next usage.</source>
</trans-unit>
<trans-unit id="sbe47a5bdeec19ab0">
<source>Stage binding(s)</source>
</trans-unit>
<trans-unit id="sfa88f413e287bb0f">
<source>Stage type</source>
</trans-unit>
<trans-unit id="s04440099d97c0bef">
<source>Edit Stage</source>
</trans-unit>
<trans-unit id="s980270d0fab7ecb3">
<source>Update Stage binding</source>
</trans-unit>
<trans-unit id="sfe938c1585e0bf68">
<source>These bindings control if this stage will be applied to the flow.</source>
</trans-unit>
<trans-unit id="sfac6f995c7670559">
<source>No Stages bound</source>
</trans-unit>
<trans-unit id="s955c1fec1c6fb970">
<source>No stages are currently bound to this flow.</source>
</trans-unit>
<trans-unit id="s9a393a04eaf1eb0e">
<source>Create Stage binding</source>
</trans-unit>
<trans-unit id="s207e8b106806d7e4">
<source>Bind stage</source>
</trans-unit>
<trans-unit id="s6f367f5604d5056d">
<source>Create and bind Stage</source>
</trans-unit>
<trans-unit id="scc2e420c54dc8089">
<source>Bind existing stage</source>
</trans-unit>
<trans-unit id="s30d1f50f476c3f48">
<source>Flow Overview</source>
</trans-unit>
<trans-unit id="s01794c0ee3629c1b">
<source>Flow Info</source>
</trans-unit>
<trans-unit id="s77099d752f1ab773">
<source>Related actions</source>
</trans-unit>
<trans-unit id="sd07866d9f38b2c50">
<source>Execute flow</source>
</trans-unit>
<trans-unit id="s9ff3121d30f88d52">
<source>Normal</source>
</trans-unit>
<trans-unit id="s6e4c997a101b6abf">
<source>with current user</source>
</trans-unit>
<trans-unit id="s8ecdbff1a7329b64">
<source>with inspector</source>
</trans-unit>
<trans-unit id="s3576aead3e68c5c9">
<source>Export flow</source>
</trans-unit>
<trans-unit id="se2c3cbf2ed1403f1">
<source>Stage Bindings</source>
</trans-unit>
<trans-unit id="s78c08391ffbfb8c0">
<source>These bindings control which users can access this flow.</source>
</trans-unit>
<trans-unit id="s7513372fe60f6387">
<source>Event volume</source>
</trans-unit>
<trans-unit id="sc1a1ff47c058bb09">
<source>Event Log</source>
</trans-unit>
<trans-unit id="s65d67612999165e9">
<source>Event <x id="0" equiv-text="${this.event.pk}"/></source>
</trans-unit>
<trans-unit id="s455de2f740b073fd">
<source>Event info</source>
</trans-unit>
<trans-unit id="sb41b2cfbbc52565b">
<source>Created</source>
</trans-unit>
<trans-unit id="s037bc6d25a03c3c8">
<source>Successfully updated transport.</source>
</trans-unit>
<trans-unit id="s1575a15cee001915">
<source>Successfully created transport.</source>
</trans-unit>
<trans-unit id="s4acf840bc792c3ae">
<source>Local (notifications will be created within authentik)</source>
</trans-unit>
<trans-unit id="sede0abbf2b612812">
<source>Webhook (generic)</source>
</trans-unit>
<trans-unit id="s76f5dca6404a1210">
<source>Webhook (Slack/Discord)</source>
</trans-unit>
<trans-unit id="s6873bdbfa24615fb">
<source>Webhook URL</source>
</trans-unit>
<trans-unit id="s25ec2846f6b88214">
<source>Webhook Mapping</source>
</trans-unit>
<trans-unit id="sca2879d96f58a39c">
<source>Send once</source>
</trans-unit>
<trans-unit id="s2430e000b7cfefd0">
<source>Only send notification once, for example when sending a webhook into a chat channel.</source>
</trans-unit>
<trans-unit id="s57072ffb92b6c9c8">
<source>Define how notifications are sent to users, like Email or Webhook.</source>
</trans-unit>
<trans-unit id="s624256f8a4bb4c89">
<source>Notification transport(s)</source>
</trans-unit>
<trans-unit id="sac1332e6f421526e">
<source>Update Notification Transport</source>
</trans-unit>
<trans-unit id="s6b5002c605b39d6d">
<source>Create Notification Transport</source>
</trans-unit>
<trans-unit id="s0a39e4f61ccafacb">
<source>Successfully updated rule.</source>
</trans-unit>
<trans-unit id="s72e102414fec81a4">
<source>Successfully created rule.</source>
</trans-unit>
<trans-unit id="sa55ee64c5c51df0f">
<source>Select the group of users which the alerts are sent to. If no group is selected the rule is disabled.</source>
</trans-unit>
<trans-unit id="sffa171e11d4ae513">
<source>Transports</source>
</trans-unit>
<trans-unit id="s7b18721be331241e">
<source>Select which transports should be used to notify the user. If none are selected, the notification will only be shown in the authentik UI.</source>
</trans-unit>
<trans-unit id="scd0cfe87af6f2ff2">
<source>Severity</source>
</trans-unit>
<trans-unit id="s107bf77afb93c9b8">
<source>Send notifications whenever a specific Event is created and matched by policies.</source>
</trans-unit>
<trans-unit id="sf3f9a0feaf083207">
<source>Sent to group</source>
</trans-unit>
<trans-unit id="sc92ed9d5e01d3f24">
<source>Notification rule(s)</source>
</trans-unit>
<trans-unit id="s5140d157642d7362">
<source>None (rule disabled)</source>
</trans-unit>
<trans-unit id="sd1146418b344f81f">
<source>Update Notification Rule</source>
</trans-unit>
<trans-unit id="sbbc1de43ab6c1f76">
<source>Create Notification Rule</source>
</trans-unit>
<trans-unit id="s5795b310ab271d20">
<source>These bindings control upon which events this rule triggers.
Bindings to groups/users are checked against the user of the event.</source>
</trans-unit>
<trans-unit id="s90c3b62194fe8508">
<source>Outpost Deployment Info</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s35f9df7668d5fa79">
<source>View deployment documentation</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sad09c62cb4ebae68">
<source>Click to copy token</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s0e03fe2dc5b9164b">
<source>If your authentik Instance is using a self-signed certificate, set this value.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc21032b0d37882a0">
<source>If your authentik_host setting does not match the URL you want to login with, add this setting.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6f270e1668c036e9">
<source>Successfully updated outpost.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s79aed8154d7c472c">
<source>Successfully created outpost.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc3259eb55cf91e8c">
<source>LDAP</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s8afc8c5aafb392d3">
<source>Radius</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s03970aa76a09982d">
<source>Integration</source>
</trans-unit>
<trans-unit id="s9c29565c5ae1cc92">
<source>Selecting an integration enables the management of the outpost by authentik.</source>
</trans-unit>
<trans-unit id="s554ce268e9727e79">
<source>You can only select providers that match the type of the outpost.</source>
</trans-unit>
<trans-unit id="sf9b1c0661a02d9f9">
<source>Configuration</source>
</trans-unit>
<trans-unit id="s3abecf1e778c9625">
<source>See more here:</source>
</trans-unit>
<trans-unit id="s74cb3d66f6a668e1">
<source>Documentation</source>
</trans-unit>
<trans-unit id="saa8939ac88a76f98">
<source>Last seen</source>
</trans-unit>
<trans-unit id="s1ac2653a6492b435">
<source><x id="0" equiv-text="${this.outpostHealth.version}"/>, should be <x id="1" equiv-text="${this.outpostHealth.versionShould}"/></source>
</trans-unit>
<trans-unit id="s1e176e35c828318c">
<source>Hostname</source>
</trans-unit>
<trans-unit id="s322e34cfcba47155">
<source>Not available</source>
</trans-unit>
<trans-unit id="s02b632a9ac24a824">
<source>Last seen: <x id="0" equiv-text="${this.outpostHealth.lastSeen?.toLocaleTimeString()}"/></source>
</trans-unit>
<trans-unit id="sa43153d53ae65063">
<source>Unknown type</source>
</trans-unit>
<trans-unit id="s8802553bc57617ee">
<source>Outposts are deployments of authentik components to support different environments and protocols, like reverse proxies.</source>
</trans-unit>
<trans-unit id="s84d7d6ebbedcb586">
<source>Health and Version</source>
</trans-unit>
<trans-unit id="s9bf48a89367282cd">
<source>Warning: authentik Domain is not configured, authentication will not work.</source>
</trans-unit>
<trans-unit id="sbf5f4c5ba679e847">
<source>Logging in via <x id="0" equiv-text="${item.config.authentik_host}"/>.</source>
</trans-unit>
<trans-unit id="s59b6028f19d15cda">
<source>No integration active</source>
</trans-unit>
<trans-unit id="s9bd59e0ea70a3e4a">
<source>Update Outpost</source>
</trans-unit>
<trans-unit id="sc8f286ac783c385d">
<source>View Deployment Info</source>
</trans-unit>
<trans-unit id="s9ee92717d7f63247">
<source>Detailed health (one instance per column, data is cached so may be out of date)</source>
</trans-unit>
<trans-unit id="s1d49ec5030447643">
<source>Outpost(s)</source>
</trans-unit>
<trans-unit id="s1a2f8f4b3861583b">
<source>Create Outpost</source>
</trans-unit>
<trans-unit id="sdc1ef94016f0d855">
<source>Successfully updated integration.</source>
</trans-unit>
<trans-unit id="sc2a1a40a1b4b0170">
<source>Successfully created integration.</source>
</trans-unit>
<trans-unit id="se9b1fec72ffd8f48">
<source>Local</source>
</trans-unit>
<trans-unit id="sc1231049879b8d33">
<source>If enabled, use the local connection. Required Docker socket/Kubernetes Integration.</source>
</trans-unit>
<trans-unit id="s13de04774ff0f210">
<source>Docker URL</source>
</trans-unit>
<trans-unit id="sa7fcf026bd25f231">
<source>Can be in the format of 'unix://' when connecting to a local docker daemon, using 'ssh://' to connect via SSH, or 'https://:2376' when connecting to a remote system.</source>
</trans-unit>
<trans-unit id="saf1d289e3137c2ea">
<source>CA which the endpoint's Certificate is verified against. Can be left empty for no validation.</source>
</trans-unit>
<trans-unit id="s0f2e070d38cd36df">
<source>TLS Authentication Certificate/SSH Keypair</source>
</trans-unit>
<trans-unit id="s2f58bb9905d2b76f">
<source>Certificate/Key used for authentication. Can be left empty for no authentication.</source>
</trans-unit>
<trans-unit id="s8b33660e2ed7212c">
<source>When connecting via SSH, this keypair is used for authentication.</source>
</trans-unit>
<trans-unit id="sa668bd79645c3e06">
<source>Kubeconfig</source>
</trans-unit>
<trans-unit id="sa85cfb884c17d85d">
<source>Verify Kubernetes API SSL Certificate</source>
</trans-unit>
<trans-unit id="se78364ee913ae2bd">
<source>New outpost integration</source>
</trans-unit>
<trans-unit id="s68d69ad0271c8ef6">
<source>Create a new outpost integration.</source>
</trans-unit>
<trans-unit id="sae239213b7c70376">
<source>State</source>
</trans-unit>
<trans-unit id="sb96629f50f2e7fab">
<source>Unhealthy</source>
</trans-unit>
<trans-unit id="sa8e255492bb6ae0d">
<source>Outpost integration(s)</source>
</trans-unit>
<trans-unit id="s9d18948d25c68d66">
<source>Successfully generated certificate-key pair.</source>
</trans-unit>
<trans-unit id="sd4ac926e4ebb1cd7">
<source>Common Name</source>
</trans-unit>
<trans-unit id="s592425143c4f5834">
<source>Subject-alt name</source>
</trans-unit>
<trans-unit id="se9d0f12f95b14095">
<source>Optional, comma-separated SubjectAlt Names.</source>
</trans-unit>
<trans-unit id="s7609ee54e8a7b05a">
<source>Validity days</source>
</trans-unit>
<trans-unit id="s4c24b2baa377e870">
<source>Successfully updated certificate-key pair.</source>
</trans-unit>
<trans-unit id="s122f308b5f198ba7">
<source>Successfully created certificate-key pair.</source>
</trans-unit>
<trans-unit id="s08a8716c214a0efb">
<source>PEM-encoded Certificate data.</source>
</trans-unit>
<trans-unit id="s6e612e5a6a359bbb">
<source>Optional Private Key. If this is set, you can use this keypair for encryption.</source>
</trans-unit>
<trans-unit id="s27ac7a47b390e3cb">
<source>Certificate-Key Pairs</source>
</trans-unit>
<trans-unit id="sb72ebab438cb2983">
<source>Import certificates of external providers or create certificates to sign requests with.</source>
</trans-unit>
<trans-unit id="s4b5af7736aedd6c1">
<source>Private key available?</source>
</trans-unit>
<trans-unit id="s1d6e16d86961c782">
<source>Certificate-Key Pair(s)</source>
</trans-unit>
<trans-unit id="sc1ce2f758935ff48">
<source>Managed by authentik</source>
</trans-unit>
<trans-unit id="sf53a78d889b6c775">
<source>Managed by authentik (Discovered)</source>
</trans-unit>
<trans-unit id="sef50d248448e0df1">
<source>Yes (<x id="0" equiv-text="${item.privateKeyType?.toUpperCase()}"/>)</source>
</trans-unit>
<trans-unit id="s33aa05f435c29753">
<source>Update Certificate-Key Pair</source>
</trans-unit>
<trans-unit id="seffdf887fed7f668">
<source>Certificate Fingerprint (SHA1)</source>
</trans-unit>
<trans-unit id="sdd6b8b56a811080e">
<source>Certificate Fingerprint (SHA256)</source>
</trans-unit>
<trans-unit id="s2a2d3e7c379e9518">
<source>Certificate Subject</source>
</trans-unit>
<trans-unit id="s351246c52548086a">
<source>Download Certificate</source>
</trans-unit>
<trans-unit id="s47bd537a3bcebf19">
<source>Download Private key</source>
</trans-unit>
<trans-unit id="s3a5fec3d73ac9edc">
<source>Create Certificate-Key Pair</source>
</trans-unit>
<trans-unit id="s45cb501abd43ba52">
<source>Generate</source>
</trans-unit>
<trans-unit id="sf9bddaf910f4eea5">
<source>Generate Certificate-Key Pair</source>
</trans-unit>
<trans-unit id="s634e2fd82c397576">
<source>Successfully updated settings.</source>
</trans-unit>
<trans-unit id="sb8e4edaea6f1d935">
<source>Avatars</source>
</trans-unit>
<trans-unit id="s945856050217c828">
<source>Configure how authentik should show avatars for users. The following values can be set:</source>
</trans-unit>
<trans-unit id="sf4ef4c8ce713f775">
<source>Disables per-user avatars and just shows a 1x1 pixel transparent picture</source>
</trans-unit>
<trans-unit id="s5446842a7e4a963b">
<source>Uses gravatar with the user's email address</source>
</trans-unit>
<trans-unit id="s35363b9e1cc2abd3">
<source>Generated avatars based on the user's name</source>
</trans-unit>
<trans-unit id="s48110ca292cad513">
<source>Any URL: If you want to use images hosted on another server, you can set any URL. Additionally, these placeholders can be used:</source>
</trans-unit>
<trans-unit id="sbe1dfda044bdc93b">
<source>The user's username</source>
</trans-unit>
<trans-unit id="s653f257c9c2d4dc5">
<source>The email address, md5 hashed</source>
</trans-unit>
<trans-unit id="s9c9183cd80916b4f">
<source>The user's UPN, if set (otherwise an empty string)</source>
</trans-unit>
<trans-unit id="h4963ed14d7e239a9">
<source>An attribute path like
<x id="0" equiv-text="<code>"/>attributes.something.avatar<x id="1" equiv-text="</code>"/>, which can be used in
combination with the file field to allow users to upload custom
avatars for themselves.</source>
</trans-unit>
<trans-unit id="s4c80c34a67a6f1c9">
<source>Multiple values can be set, comma-separated, and authentik will fallback to the next mode when no avatar could be found.</source>
</trans-unit>
<trans-unit id="h2fafcc3ebafea2f8">
<source>For example, setting this to <x id="0" equiv-text="<code>"/>gravatar,initials<x id="1" equiv-text="</code>"/> will
attempt to get an avatar from Gravatar, and if the user has not
configured on there, it will fallback to a generated avatar.</source>
</trans-unit>
<trans-unit id="sea247dfb18f696d2">
<source>Default user change name</source>
</trans-unit>
<trans-unit id="s078ffec0257621c0">
<source>Enable the ability for users to change their name.</source>
</trans-unit>
<trans-unit id="s26ad11b21253152b">
<source>Default user change email</source>
</trans-unit>
<trans-unit id="s5fc6c14d106f40d3">
<source>Enable the ability for users to change their email.</source>
</trans-unit>
<trans-unit id="s23564aa839755965">
<source>Default user change username</source>
</trans-unit>
<trans-unit id="s6d816a95ca43a99d">
<source>Enable the ability for users to change their username.</source>
</trans-unit>
<trans-unit id="s84c5a011acd608c9">
<source>Event retention</source>
</trans-unit>
<trans-unit id="s2536ac8d32d2e63f">
<source>Duration after which events will be deleted from the database.</source>
</trans-unit>
<trans-unit id="s7b1fba26d245cb1c">
<source>When using an external logging solution for archiving, this can be set to "minutes=5".</source>
</trans-unit>
<trans-unit id="s44536d20bb5c8257">
<source>This setting only affects new Events, as the expiration is saved per-event.</source>
</trans-unit>
<trans-unit id="s57b52b60ed5e2bc7">
<source>Footer links</source>
</trans-unit>
<trans-unit id="s7349802b2f7f99c2">
<source>This option configures the footer links on the flow executor pages. It must be a valid JSON list and can be used as follows:</source>
</trans-unit>
<trans-unit id="s166b59f3cc5d8ec3">
<source>GDPR compliance</source>
</trans-unit>
<trans-unit id="sb8b23770f899e5bb">
<source>When enabled, all the events caused by a user will be deleted upon the user's deletion.</source>
</trans-unit>
<trans-unit id="s29501761df0fe837">
<source>Impersonation</source>
</trans-unit>
<trans-unit id="s8f503553d8432487">
<source>Globally enable/disable impersonation.</source>
</trans-unit>
<trans-unit id="see1eb81c1f734079">
<source>System settings</source>
</trans-unit>
<trans-unit id="s33f85f24c0f5f008">
<source>Save</source>
</trans-unit>
<trans-unit id="see2bcbc11bb91960">
<source>Successfully updated instance.</source>
</trans-unit>
<trans-unit id="s9e51d6de369f320b">
<source>Successfully created instance.</source>
</trans-unit>
<trans-unit id="s92e91071c6a45eb4">
<source>Disabled blueprints are never applied.</source>
</trans-unit>
<trans-unit id="sf63c89c0604c288f">
<source>Local path</source>
</trans-unit>
<trans-unit id="sd6422f7004036cdd">
<source>OCI Registry</source>
</trans-unit>
<trans-unit id="scbb7d3154da629f3">
<source>OCI URL, in the format of oci://registry.domain.tld/path/to/manifest.</source>
</trans-unit>
<trans-unit id="s0195c0df7294228a">
<source>See more about OCI support here:</source>
</trans-unit>
<trans-unit id="sfae395b94a5a0040">
<source>Blueprint</source>
</trans-unit>
<trans-unit id="s7e1342d37124b65b">
<source>Configure the blueprint context, used for templating.</source>
</trans-unit>
<trans-unit id="s6ec8c9d11310300a">
<source>Orphaned</source>
</trans-unit>
<trans-unit id="s6835db03209b4f94">
<source>Automate and template configuration within authentik.</source>
</trans-unit>
<trans-unit id="s23de62f931f7d754">
<source>Last applied</source>
</trans-unit>
<trans-unit id="s2708cac1f4942708">
<source>Blueprint(s)</source>
</trans-unit>
<trans-unit id="s880b8b70b22f9977">
<source>Update Blueprint</source>
</trans-unit>
<trans-unit id="sb8f855b49234b81b">
<source>Apply</source>
</trans-unit>
<trans-unit id="sef3d102324bf8561">
<source>Create Blueprint Instance</source>
</trans-unit>
<trans-unit id="s0e427111d750cc02">
<source>Successfully updated license.</source>
</trans-unit>
<trans-unit id="s06ae64e621f302eb">
<source>Successfully created license.</source>
</trans-unit>
<trans-unit id="s2905c425adae99bd">
<source>Install ID</source>
</trans-unit>
<trans-unit id="sb18ec434a8a3aafb">
<source>License key</source>
</trans-unit>
<trans-unit id="s9ce7cc01fb9b5b53">
<source>Manage enterprise licenses</source>
</trans-unit>
<trans-unit id="sf9ebf11ac2645820">
<source>No licenses found.</source>
</trans-unit>
<trans-unit id="sd49099e9522635f4">
<source>License(s)</source>
</trans-unit>
<trans-unit id="s3be1d90ffa46b7f1">
<source>Enterprise is in preview.</source>
</trans-unit>
<trans-unit id="s4015746f55a8d89f">
<source>Get a license</source>
</trans-unit>
<trans-unit id="sb2cbd06f8e25b47e">
<source>Go to Customer Portal</source>
</trans-unit>
<trans-unit id="sf58825457d61c429">
<source>Forecast internal users</source>
</trans-unit>
<trans-unit id="sde9a3f41977ec1f8">
<source>Estimated user count one year from now based on <x id="0" equiv-text="${this.forecast?.internalUsers}"/> current internal users and <x id="1" equiv-text="${this.forecast?.forecastedInternalUsers}"/> forecasted internal users.</source>
</trans-unit>
<trans-unit id="s4557b6b9da258643">
<source>Forecast external users</source>
</trans-unit>
<trans-unit id="sf52479d6daa0a4a8">
<source>Estimated user count one year from now based on <x id="0" equiv-text="${this.forecast?.externalUsers}"/> current external users and <x id="1" equiv-text="${this.forecast?.forecastedExternalUsers}"/> forecasted external users.</source>
</trans-unit>
<trans-unit id="sd22bd01bdf28c548">
<source>Cumulative license expiry</source>
</trans-unit>
<trans-unit id="s0dd031b58ed4017c">
<source>Internal: <x id="0" equiv-text="${item.internalUsers}"/></source>
</trans-unit>
<trans-unit id="s57b07e524f8f5c2a">
<source>External: <x id="0" equiv-text="${item.externalUsers}"/></source>
</trans-unit>
<trans-unit id="sdeb6cee42435dd07">
<source>Update License</source>
</trans-unit>
<trans-unit id="s6196153c4b0c1ea0">
<source>Install</source>
</trans-unit>
<trans-unit id="s0285b4bd69130fa3">
<source>Install License</source>
</trans-unit>
<trans-unit id="s67ac11d47f1ce794">
<source>WebAuthn requires this page to be accessed via HTTPS.</source>
2023-10-08 22:08:16 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="se9e9e1d6799b86a5">
<source>WebAuthn not supported by browser.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s09f0c100d0ad2fec">
<source>Open Wizard</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sf2ef885f7d0a101d">
<source>Demo Wizard</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s77505ee5d2e45e53">
<source>Run the demo wizard</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sab85321d3b0840b7">
<source>API request failed</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sd766cdc29b25ff95">
<source>Authenticating with Apple...</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s2c8189544e3ea679">
<source>Retry</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s420d2cdedcaf8cd0">
<source>Authenticating with Plex...</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s2ddbebcb8a49b005">
<source>Waiting for authentication...</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb15fe7b9d09bb419">
<source>If no Plex popup opens, click the button below.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sbc625b4c669b9ce8">
<source>Open login</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa3599457b9418bc5">
<source>User's avatar</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s9bd9ba84819493d4">
<source>Something went wrong! Please try again later.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s4090dd0c0e45988b">
<source>Request ID</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s4d7fe7be1c49896c">
<source>You may close this page now.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sf8c76d5fb408de7b">
<source>You're about to be redirect to the following URL.</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s197420b40df164f8">
<source>Follow redirect</source>
2023-10-16 15:31:50 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa11e92683c5860c7">
<source>Request has been denied.</source>
web: Application wizard v2 with tests (#7004)
* A lot of comments about forms.
* Adding comments to the wizard.
* Broke out the text input into a single renderer. Still works as required.
* web: Legibility in the ApplicationForm.
This is a pretty good result. By using the LightDOM setting, this
provides the existing Authentik form manager with access to the
ak-form-horizontal-element components without having to do any
cross-border magic. It's not ideal, and it shows up just how badly
we've got patternfly splattered everywhere, but the actual results
are remarkable. The patterns for text, switch, radio, textarea,
file, and even select are smaller and easier here.
I'm still noodling on what an unspread search-select element would
look like. It's just dependency injection, so it ought to be as
straightforward as that.
* web: Marking down the start of the 'components' library.
* web: Baby steps
I become frustrated with my inability to make any progress on this project, so I decided to reach
for a tool that I consider highly reliable but also incredibly time-consuming and boring: test
driven development.
In this case, I wrote a story about how I wanted to see the first page rendered: just put the HTML
tag, completely unadorned, that will handle the first page of the wizard. Then, add an event handler
that will send the updated content to some parent object, since what we really want is to
orchestrate the state of the user's input with a centralized location. Then, rather than fiddling
with the attributes and properties of the various pages, I wanted them to be able to "look up" the
values they want, much as we'd expect a standalone form to be able to pull its values from the
server, so I added a context object that receives the update event and incorporates the new
knowledge about the state of the process into itself.
The result is surprisingly satisfying: the first page renders cleanly, displays the content that we
want, and as we fiddle with, we can *watch in real time* as the results of the context are updated
and retransmitted to all receiving objects. And the sending object gets the results so it
re-renders, but it ends up looking the same as it was before the render.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* Now, it's starting to look like a complete package. The LDAP method is working, but there is a bug:
the radio is sending the wrong value !?!?!?. Track that down, dammit. The search wrappers now resend
their events as standard `input` events, and that actually seems to work well; the browser is
decorating it with the right target, with the right `name` attribute, and since we have good
definitions of the `value` as a string (the real value of any search object is its UUID4), that
works quite well. Added search wrappers for CoreGroup and CryptoCertificate (CertificateKeyPairs),
and the latter has flags for "use the first one if it's the only one" and "allow the display of
keyless certificates."
Not sure why `state()` is blocking the transmission of typing information from the typed element
to the context handler, but it's a bug in the typechecker, and it's not a problem so far.
* web: tracked down that weirld bug with the radio.
Because radio inputs are actually multiples, the events handling for
radio is... wonky. If we want our `<ak-radio>` component to be a
unitary event dispatcher, saying "This is the element selected," we
needed to do more than what was currently being handled.
I've intercepted the events that we care about and have placed
them into a controller that dictates both the setting and the
re-render of the component. This makes it "controlled" (to use the
Angular/React/Vue) language and depends on Lit's reactiveElement
lifecycle to work, rather than trust the browser, but the browser's
experience with respect to the `<input type=radio` is pretty bad:
both input elements fire events, one for "losing selection" and
one for "gaining selection". That can be very confusing to handle,
so we funnel them down in our aggregate radio element to a single
event, "selection changed".
As a quality-of-life measure, I've also set the label to be
unselectable; this means that a click on the label will trigger the
selection event, and a long click will not disable selection or
confuse the selection event generator.
* web: now passing the precommit phase
* web: a HACK for Storybook to inject the "use light theme" flag into the body.
This isn't really a very good hack; what it does is say that every story is
responsible for hacking its theme into the parent. This is very annoying, but
it does mean that we can at least show our components in the best light.
* web: ak-application-wizard-authentication-by-oauth, and many fixes!
1. Fixed `eventEmitter` so that if the detail object is a scalar, it will not attempt to "objectify"
it. This was causing a bug where retrofitting the eventEmitter to some older components resulted
in a detail of "some" being translated into ['s', 'o', 'm', 'e']. Not what is wanted.
2. Removed the "transitional form" from the existing components; they had a two-step where the web
component class was just a wrapper around an independent rendering function. While this worked,
it was only to make the case that they *were* independent rendering objects and could be
supported with the right web component framework. We're halfway there now; the last step will be
to transform the horizontal-element and various input CSS into componentized CSS, the way
Patternfly-Elements is currently doing.
3. Fixed the `help` field so that it could take a string or a TemplateResult, and if the latter,
don't bother wrapping it in the helper text functionality; just let it be its own thing. This
supports the multi-line help of redirectURI as well as the `ak-utils-time-delta` capability.
4. Transform Oauth2ProviderForm to use the new components, to the best of our ability. Also used
the `provider = this.wizard.provider` and `provider = this.instance` syntax to make the render
function *completely portable*; it's the exact same text that is dropped into...
5. The complete `ak-application-wizard-authentication-by-oauth` component. They're so similar part
of me wonders if I could push them both out to a common reference, or a collection of common
references. Both components use the PropertyMapping and Sources, and both use the same
collection of searches (Crypto, Flow).
6. A Storybook for `ak-application-wizard-authentication-by-oauth`, showing the works working.
7. New mocks for `authorizationFlow`, `propertyMappings`, and `hasJWKs`.
This sequence has revealed a bug in the radio control. (It's always the radio control.) If the
default doesn't match the current setting, the radio control doesn't behave as expected; it won't
change when you fully expect that it should. I'll investigate how to harmonize those tomorrow.
* web: Converted our toggle groups to a more streamlined implementation.
* web: one more toggle group.
* initial api and schema
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* separate blueprint importer from yaml parsing
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* cleanup
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: Replace ad-hoc toggle control with ak-toggle-group
This commit replaces various ad-hoc implementations of the Patternfly Toggle Group HTML with a web
component that encapsulates all of the needed behavior and exposes a single API with a single event
handler, return the value of the option clicked.
The results are: Lots of visual clutter is eliminated. A single link of:
```
<div class="pf-c-toggle-group__item">
<button
class="pf-c-toggle-group__button ${this.mode === ProxyMode.Proxy
? "pf-m-selected"
: ""}"
type="button"
@click=${() => {
this.mode = ProxyMode.Proxy;
}}>
<span class="pf-c-toggle-group__text">${msg("Proxy")}</span>
</button>
</div>
<div class="pf-c-divider pf-m-vertical" role="separator"></div>
```
Now looks like:
```
<option value=${ProxyMode.Proxy}>${msg("Proxy")}</option>
```
This also means that the three pages that used the Patternfly Toggle Group could eliminate all of
their Patternfly PFToggleGroup needs, as well as the `justify-content: center` extension, which also
eliminated the `css` import.
The savings aren't as spectacular as I'd hoped: removed 178 lines, but added 123; total savings 55
lines of code. I still count this a win: we need never write another toggle component again, and
any bugs, extensions or features we may want to add can be centralized or forked without risking the
whole edifice.
* web: minor code formatting issue.
* add new "must_created" state to blueprints to prevent overwriting objects
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: adding a storybook for the ak-toggle-group component
* Bugs found by CI/CD.
* web: Replace ad-hoc search for CryptoCertificateKeyPairs with ak-crypto-certeficate-search
This commit replaces various ad-hoc implementations of `search-select` for CryptoCertificateKeyPairs
with a web component that encapsulates all of the needed behavior and exposes a single API.
The results are: Lots of visual clutter is eliminated. A single search of:
```HTML
<ak-search-select
.fetchObjects=${async (query?: string): Promise<CertificateKeyPair[]> => {
const args: CryptoCertificatekeypairsListRequest = {
ordering: "name",
hasKey: true,
includeDetails: false,
};
if (query !== undefined) {
args.search = query;
}
const certificates = await new CryptoApi(
DEFAULT_CONFIG,
).cryptoCertificatekeypairsList(args);
return certificates.results;
}}
.renderElement=${(item: CertificateKeyPair): string => {
return item.name;
}}
.value=${(item: CertificateKeyPair | undefined): string | undefined => {
return item?.pk;
}}
.selected=${(item: CertificateKeyPair): boolean => {
return this.instance?.tlsVerification === item.pk;
}}
?blankable=${true}
>
</ak-search-select>
```
Now looks like:
```HTML
<ak-crypto-certificate-search certificate=${this.instance?.tlsVerification}>
</ak-crypto-certificate-search>
```
There are three searches that do not require there to be a valid key with the certificate; these are
supported with the boolean property `nokey`; likewise, there is one search (in SAMLProviderForm)
that states that if there is no current certificate in the SAMLProvider and only one certificate can
be found in the Authentik database, use that one; this is supported with the boolean property
`singleton`.
These changes replace 382 lines of object-oriented invocations with 36 lines of declarative
configuration, and 98 lines for the class. Overall, the code for "find a crypto certificate" has
been reduced by 46%.
Suggestions for a better word than `singleton` are welcome!
* web: display tests for CryptoCertificateKeypair search
This adds a Storybook for the CryptoCertificateKeypair search, including
a mock fetch of the data. In the course of running the tests, we discovered
that including the SearchSelect _class_ won't include the customElement declaration
unless you include the whole file! Other bugs found: including the CSS from
Storybook is different from that of LitElement native, so much so that the
adapter needed to be included. FlowSearch had a similar bug. The problem
only manifests when building via Webpack (which Storybook uses) and not
Rollup, but we should support both in distribution.
* Fixed behavioral problem with the radio; the `if` there was
preventing the radio from reflecting the default correctly.
The observed behavior was that the radio wouldn't "activate"
until the item selected during the render pass was clicked on
first.
* Proxy Provider done.
* web: Tactical change. Put all the variants on the second page; it's
a longer list, but it's also easier to manage than all those
required sub-options.
* Rounding out the catalog.
* web: SAML Manual Configuration
Added a 'design document' that just kinda describes what I'm trying
to do, in case I don't get this done by Friday Aug 11, 2023.
I had two tables doing the same thing, so I merged them and then
wrote a few map/filters to specialize them for those two use cases.
Along the way I had to fiddle with the ESLint settings so that
underscore-prefixed unused variables would be ignored.
I cleaned up the visual appeal of the forms in the LDAP application.
I was copy/pasting the "handleProviderEvent" function, so I pulled
it out into ApplicationWizardProviderPageBase. Not so much a matter
of abstraction as just disliking that kind of duplication; it served
no purpose.
* Added SAML Story to Storybook.
* Web: This is coming together amazingly well. Like, almost too well.
* web: 80% of the way there
This commit includes the first three pages of the wizard, the
completion of the wizard framework with evented handling, and control
over progression.
Some shortcomings of this design have become evident: it isn't
possible to communicate between the steps' wrappers, as they are
POJOs without access to the context. An imperative decision-making
process has to be inserted in the orchestration layer,
which is kinda annoying.
But it looks good and it behaves correctly, to the extent that I've
given it behavior. It's an excellent foundation.
* Linting.
* web: application wizard
Found where the hook for form validity should go. Excellent! Now I just need to incorporate
that basic validation into the business logic and we're good to go.
* Turns out that was one layer too many; the topmost component was fine for
maintaining the context.
* It looks like my brilliant strategy has hit a snag.
The idea is simple. Let's start with this picture:
```
<application-wizard .steps=${[... a collection of step objects ...]}>
<wizard-main .steps=${(steps from above)}>
<application-current-panel>
<current-form>
```
- ApplicationWizard has a Context for the ApplicationProviderPair (or whatever it's going to be).
This context does not know about the steps; it just knows about: the "application" object, the
"provider" object, and a discriminator to know *which* provider the user has selected.
- ApplicationWizard has Steps that, among other things, provides Panels for:
- Application
- Pick Provider
- Configure Provider
- Submit ApplicationProviderPair to the back-end
- The WizardFrame renders the CurrentPanel for the CurrentStep
The CurrentPanel gets its data from the ApplicationWizard in the form of a Context. It then sends
messages (events) to ApplicationWizard about the contents of each field as the user is filling out
the form, so that the ApplicationWizard can record those in the ApplicationProviderPair for later
submission.
When a CurrentForm is valid, the ApplicationWizard updates the Steps object to show that the "Next
button" on the Wizard is now available.
In this way, the user can progress through the system. When they get to the last page, we can
provide in the ApplicationWizard with the means to submit the form and/or send the user back to
the page with the validation failure.
Problem: The context is being updated in real-time, which is triggering re-renders of the form. This
leads to focus problems as the fields that are not yet valid are triggering "focus grab" behavior.
This is a classic problem with "controlled" inputs. What we really want is for the CurrentPanel to
not re-render at all, but to behave like a normal, uncontrolled form, and let the browser do most of
the work. We still want the [Next] button to enable when the form is valid enough to permit that.
---
Other details: I've ripped out a lot of Jen's work, which is probably a mistake. It's still
preserved elsewhere. I've also cleaned up the various wizardly things to try and look organized.
It *looks* like it should work, it just... doesn't. Not yet.
* Late addition: I had an inspiration about how to reduce the way
reactivity broke focus by, basically, removing the reactivity and
managing the first-time-through lifecycle to prevent the update
from causing refocus. It works well! Now I just need to test it.
* This application fixes the bug with respect to the wizard-level context being updated incorrectly.
Understandings:
- To use uncontrolled inputs, which I prefer, the context object should not be a state or property
at the level of consumers; it should not automatically re-render with every keystroke, i.e. "The
React Way." We're using Web Components, [client-side
validation](https://developer.mozilla.org/en-US/docs/Learn/Forms/Form_validation) exists on the
platform already, and live-validation is problematic for any number of reasons.
- The trade-off is that it is now necessary to re-render the target page of the wizard de-novo, but
that's not really as big a deal as it sounds. Lit is ready to do that... and then nothing else
until we request a change-of-page. Excellent.
- The top level context *must* be a state, but it's better if it's a state never actually used by
the top-level context container. The debate about whether or not to make that container a dumb one
(`<slot></slot>`) or to merge it with the top-level object continues; here, I've merged it with
the top-level wizard object, but that object does not refer to the state variable being managed in
its render pass, so changes to it do not cause a re-render of the whole wizard. The purpose of the
top-level page is to manage the *steps*, not the *content of any step*. A step may change
dynamically based on the content of a step, but that's the same thing as *which step*. Lesson:
always know what your state is *about*.
- Deep merging is a complex subject, but here it's appropriate to our needs.
* web: Application Wizard
This commit combines a working (but very unpolished) version of the Application Wizard with Jen's
code for the CoreTransactionApplicationRequest, resulting in a successful round trip.
It fixes a number of bugs with the way ContextProducer decorators were being processed, such that
they just weren't working with our current configuration (although they did work fine in Storybook);
consumers didn't need to be fixed.
It also *removes* the steps-aware context from the Wizard.
That *may* be a mistake. To re-iterate, the `WizardFrame` provides the chrome for a Wizard: the
button bar div, the breadcrumbs div, the header div, and it takes the steps object as its source of
truth for all of the content. The `WizardContent` part of the application has two parts: The
`WizardMain`, which wraps the frame and supplies the context for all the `WizardPanels`, and the
`WizardPanels` themselves, which are dependent on a context from `WizardMain` for the data that
populates each panel. YAGNI right now that the panels need to know anything about the steps, and the
`WizardMain` can just pass a fresh `.steps` object to the `WizardFrame` when they need updating.
Using props drilling may make more sense here.
It certainy does *not* make sense for the panels. They need to be renderable on-demand, and they
need to make sense of what they're rendering on-demand, so the function is
```
(panel code) => (context) => (rendered panel)
```
(Yes, that's curried notation. Deal.)
* This commit includes the first WDIO test for the ApplicationWizard. It doesn't do much right now, but
it does log in and navigate to the wizard successfully.
* web: completed test for single application, provided new programming language to make it easier to write tests.
* Almost there.
Missing: The validation is currently not working as expected, and I cannot get the backend
to give me meaningful data helping us "go back" to the field that wasn't valid. I really
don't want to put all the meaningful validation on the front-end; that's the road to -
perdition, the back-end must be usable by people less assiduous than we are.
Also: Need to make the button bar work better; maybe each panel can provide a custom button
bar if one is needed?
* web: Test harness
We have an end-to-end test harness that includes a trivially correct DSL for "This is what a user would do, do this":
```
const deleteProvider = (theSlug) => ([
["button", '>>>ak-sidebar-item a[href="#/core/providers"]'],
["deletebox", `>>>a[href="#/core/applications/${theSlug}"]`],
["button", '>>>ak-forms-delete-bulk button[slot="trigger"]'],
["button", '>>>ak-forms-delete-bulk div[role="dialog"] ak-spinner-button'],
]);
```
It's now possible to target individual sequences of events this way. With a little creativity, we could have standalone functions that take parameters for our calls and just do them, without too much struggle.
* web: Revised navigation
After working with the navigation for awhile, I realized that it's a poor map; what I really wanted was
a controller/view pair, where events flow up to the controller and then messages on "what to draw" flow
down to the view. It work quite well, and the wizard frame is smaller and smarter for it.
I've also moved the WDIO-driven tests into the 'tests' folder, because it (a) makes more sense to put
them there, and (b) it prevents any confusion about who's in charge of node_modules.
* web: Simplify, simplify, simplify
Sort-of.
This commit changes the way the "wizard step coordinator" layer works, giving the
wizard writer much more power over button bar. It still assumes there are only
three actions the wizard frame wants to commit: next, back, and close. This empowers
the steps themselves to re-arrange their buttons and describe the rules through which
transitions occur.
* web: resetting the form is not working yet...
I vehemently dislike the object-oriented "reset" command; every wizard should start with
an absolutely fresh copy of the data upon entry. Refactoring the wizard to re-build its
content from the inside is the correct way to go, but I don't have a good mental image
of how to make the ModalButton and the component it invokes interact cleanly, which
frustrates the hell out of me.
* web: reset
As I said, I greatly dislike having to be dependent upon "resets"; I prefer my
data to be de novo going into a "new" transaction. That said, we work with
what we've got; I've created an event generated by the wizard that says the
modal just closed; anything wrapping and implementing the wizard can then
capture that event and reset the data. I've also added a pair of functions
that create the two states (what step, what form data) anew, so that resetting
is as trivial as initializing (and is exactly the same, code-wise).
* web: Without error handling, this is complete, but I still need @BeryJu (Jens)
for help with the SAML Upload (it doesn't appear to be correctly handled?) and
the error handling.
* web: revise tests for wizard
This commit replaces the previous WDIO instance with a more formal and straightforward process using
the [pageobjects](https://martinfowler.com/bliki/PageObject.html). In this form, every major
component has its own test suite, and a test is a sequence of exercises of those components.
A test then becomes something as straightforward as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
expect(await UserLibraryPage.pageHeader).toHaveText("My Applications");
await UserLibraryPage.goToAdmin();
expect(await AdminOverviewPage.pageHeader).toHaveText("Welcome, ");
await AdminOverviewPage.openApplicationsListPage();
expect(await ApplicationsListPage.pageHeader).toHaveText("Applications");
ApplicationsListPage.startCreateApplicationWizard();
await ApplicationWizard.app.name.setValue(`Test application ${newId}`);
await ApplicationWizard.nextButton.click();
await (await ApplicationWizard.getProviderType("ldapprovider")).click();
await ApplicationWizard.nextButton.click();
await ApplicationWizard.ldap.setBindFlow("default-authentication-flow");
await ApplicationWizard.nextButton.click();
await expect(await ApplicationWizard.commitMessage).toHaveText(
"Your application has been saved"
);
```
Whether or not there's another layer of DSL in there or not, this is a pretty nice idiom for
maintaining tests.
* web: updating with forms and fixes for eslint complaints.
* web/add webdriverIO testing layer
This commit adds WebdriverIO as an end-to-end solution to unit testing. WebdriverIO can be run both
locally and remotely, supports strong integration with web components, and is generally robust for
use in pipelines. I'll confess to working through a tutorial on how to do this for web components,
and this is just chapter 2 (I think there are 5 or so chapters...).
There's a makefile, with help! If you just run `make` it tells you:
```
Specify a command. The choices are:
help Show this help
node_modules Runs `npm install` to prepare this feature
precommit Run the precommit: spell check all comments, eslint with sonarJS, prettier-write
test-good-login Test that we can log into the server. Requires a running instance of the server.
test-bad-login Test that bad usernames and passwords create appropriate error messages
```
... because Makefiles are documentation, and documentation belongs in Makefiles.
I've chosen to go with a PageObject-oriented low-level DSL; what that means is that for each major
components (a page, a form, a wizard), there's a class that provides human-readable names for
human-interactable and human-viewable objects on the page. The LoginPage object, for example, has
selectors for the username, password, submit button, and the failure alert; accessing those allows
us to test for items as expected., and to write a DSL for "a good login" that's as straightforward
as:
```
await LoginPage.open();
await LoginPage.login("ken@goauthentik.io", "eat10bugs");
await expect(UserLibraryPage.pageHeader).toHaveText("My applications");
```
There was a *lot* of messing around with the LoginPage to get the username and password into the
system. For example, I had to do this with all the `waitForClickable` and `waitForEnable` because
we both keep the buttons inaccessible until the form has something and we "black out" the page (put
a darkening filter over it) while accessing the flow, meaning there was a race condition such that
the test would attempt to interact with the username or password field before it was accessible.
But this works now, which is very nice.
``` JavaScript
get inputUsername() {
return $('>>>input[name="uidField"]');
}
get btnSubmit() {
return $('>>>button[type="submit"]');
}
async username(username: string) {
await this.inputUsername.waitForClickable();
await this.inputUsername.setValue(username);
await this.btnSubmit.waitForEnabled();
await this.btnSubmit.click();
}
```
The bells & whistles of *Prettier*, *Eslint*, and *Codespell* have also been enabled. I do like my
guardrails.
* web/adding tests: added comments and cleaned up some administrative features.
* web/test: changed the name of one test to reflect it's 'good' status
* core/allow alternative postgres credentials
This commit allows the `dev-reset` command in the Makefile to pick up and use credentials from the
`.env` file if they are present, or fallback to the defaults provided if they are not. This is the
only place in the Makefile where the database credentials are used directly against postgresql
binaries. The syntax was tested with bash, zsh, and csh, and did not fail under those.
The `$${:-}` syntax is a combination of a Makefile idiom for "Pass a single `$` to the environment
where this command will be executed," and the shell expresion `${VARIABLE:-default}` means
"dereference the environment variable; if it is undefined, used the default value provided."
* Re-arrange sequence to avoid recursive make.
Nothing wrong with recursive make; it just wasn't essential
here. `migrate` is just a build target, not a task.
* Cleanup according to the Usage:
checkmake [options] <makefile>...
checkmake -h | --help
checkmake --version
checkmake --list-rules Makefile linting tool.
* core: added 'help' to the Makefile
* get postgres config from authentik config loader
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* don't set -x by default
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* sort help
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update help strings
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: test LDAP wizard sequence
* web: improve testing by adding test admin user via blueprint
* This commit continues the application wizard buildout. In this commit are the following changes:
- Added SCIM to the list of available providers
- Fixed ForwardProxy so that its mode is set correctly. (This is a special case in the committer;
I'm unhappy with that.)
- Fixed the commit messages so that:
- icons are set correctly (Success, Danger, Working)
- icons are colored correctly according to state
- commit message includes a `data-commit-state` field so tests can find it!
- Merged the application wizard tests into a single test pass
- Isolated common parts of the application wizard tests to reduce unnecessary repetition. All
application tests are the same until you reach the provider section anyway.
- Fixed the unit tests so they're finding the right error messages and are enabled to display them
correctly.
- Moved the test Form handlers into their own folder so they're not cluttering up the Pages folder.
* web: add radius to application wizard
This commit continues the application wizard buildout. In this commit are the following changes:
- Fixed a width-setting bug in the Makefile `make help` feature (i.e "automate that stuff!")
- Added Radius to the list of providers we can offer via the wizard
- Added `launchUrl` and `UI Settings` to features of the application page the wizard can find
- Changed 'SAML Manual Configuration' to just say "SAML Configuration"
- Modified `ak-form-group` to take and honor the `aria-label` property (which in turn makes it
easier to target specific forms with unit testing)
- Reduced the log level for wdio to 'warn'; 'info' was super-spammy and not helpful. It can be put
back with `--logLevel info` from the command line.
* fix blueprints
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* update package name
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add dependabot
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* prettier run
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* add basic CI
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* remove hooks
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: application wizard refactor & completion
This commit refactors the various components of the Wizard and ApplicationWizard, creating a much
more maintainable and satisfying Wizard experience for both developers (i.e, *me* and *Jens* so
far), and for the customer.
The Wizard base has been refactored into three components:
**AkWizardController**
The `AkWizardController` provides the event listenters for the wizard; it hooks them up, recevies the
events, and forwards them to the wizard. It unwraps the event objects and forwards the relevant
messages contained in the events. It knows of three event categories:
- Navigation requests (move to a different step)
- Update requests (the current step has updated the business content)
- Close requests (close or cancel the wizard).
**ak-wizard-frame**
The `ak-wizard-frame` is the ModalButton interface. It provides the Header, Breadcrumbs (nee`
"navigation block"), Buttons, and a DIV into which the main content is rendered.
**AkWizard**
`AkWizard` is an *incomplete* implementation of the wizard. It's meant to be inherited by a child
class, which will implement the rest. It extends `AKElement`. It provides the basic content needed,
such as steps, currentStep (as an index), an accessor for the step itself, an accessor for the
frame, and the interface to the `AkWizardController`.
**ApplicationWizard**
The `ApplicationWizard` itself has been refactored to accommodate these changes. It inherits from
`AkWizard` and provides the business logic for what to do when a form updates, some custom logic for
preventing moving through the wizard when the forms are incomplete, and a persistence layer for
filling out different providers in the same session. It's simplified a *lot*.
The types specified for `AkWizard` are pretty nifty, I think. I could wish the types being passed
via the custom events were more robust, but [strongly typed custom
events](https://github.com/lit/lit-element/issues/808) turn out to be quite the pain in the, er,
neck. As it is, the `precommit` pass did very good at preventing the worst disasters.
The steps themselves were re-written as objects so that they could take advantage of their `valid`
and `disabled` states and provide more meaningful buttons and labels. I think it's a solid
compromise, and it moved a lot of display logic out of the core `handleUpdate()` business method.
The tests, such as they are, are passing.
* Added comment describing new test.
* web: ensuring copy from `main` is canon
* web: fixes after merge
* web: laying the groundwork for future expansion
This commit is a hodge-podge of updates and changes to the web. Functional changes:
- Makefile: Fixed a bug in the `help` section that prevented the WIDTH from being accurately
calculated if `help` was included rather than in-lined.
- ESLint: Modified the "unused vars" rule so that variables starting with an underline are not
considered by the rule. This allows for elided variables in event handlers. It's not a perfect
solution-- a better one would be to use Typescript's function-specialization typing, but there are
too many places where we elide or ignore some variables in a function's usage that switching over
to specialization would be a huge lift.
- locale: It turns out, lit-locale does its own context management. We don't need to have a context
at all in this space, and that's one less listener we need to attach t othe DOM.
- ModalButton: A small thing, but using `nothing` instead of "html``" allows lit better control over
rendering and reduces the number of actual renders of the page.
- FormGroup: Provided a means to modify the aria-label, rather than stick with the just the word
"Details." Specializing this field will both help users of screen readers in the future, and will
allow test suites to find specific form groups now.
- RadioButton: provide a more consistent interface to the RadioButton. First, we dispatch the
events to the outside world, and we set the value locally so that the current `Form.ts` continues
to behave as expected. We also prevent the "button lost value" event from propagating; this
presents a unified select-like interface to users of the RadioButtonGroup. The current value
semantics are preserved; other clients of the RadioButton do not see a change in behavior.
- EventEmitter: If the custom event detail is *not* an object, do not use the object-like semantics
for forwarding it; just send it as-is.
- Comments: In the course of laying the groundwork for the application wizard, I throw a LOT of
comments into the code, describing APIs, interfaces, class and function signatures, to better
document the behavior inside and as signposts for future work.
* web: permit arrays to be sent in custom events without interpolation.
* actually use assignValue or rather serializeFieldRecursive
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: eslint & prettier fixes, plus small aesthetic differences.
* Restoring this file. Not sure where it disappears to.
* fix label in dark mode
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* SCIM Manuel -> SCIM
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix lint errors
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: better converter configuration, CSS repair, and forward-domain-proxy
1. Forward Domain Proxy. I wasn't sure if this method was appropriate for the wizard,
but Jens says it is. I've added it.
2. In the process of doing so, I decided that the Provider.converter field was overly
complexified; I tried too hard to reduce the number of functions I needed to define,
but in the process outsourced some of the logic of converting the Wizard's dataset
into a property typed request to the `commit` phase, which was inappropriate. All
of the logic about a provider, aside from its display, should be here with the code
that distinguishes between providers. This commit makes it so.
3. Small CSS fix: the fields inherited from the Proxy provider forms had some unexpected
CSS which was causing a bit of a weird indent. That has been rectified.
* web: running pre-commit after merge.
* web: ensure the applications wizard tests finish after current changes
* prettier has opinions.
* web: application wizard spit & polish
The "ApplicationWizardHint" now correctly uses the localstorage and allows the user to navigate back
and see the message after it's been hidden, so that it will always be available during the test
phase.
The ApplicationList's old "Create Application Form" button has been restored for the purposes of the
test phase.
The ApplicationWizard is now available on both the ApplicationList and ProviderList pages.
Tana and I discussed the microcopy, putting a stronger second-person "You can do..." twist onto the
language, to give the user the sense of empowerment.
The ShowHintController now has both "hide" and "show" operations, to support the hint restoration.
* web: updated storybook stories for the wizard, illustration how "a simple wizard" is configured in source code and tested with storybook.
* web: I hate getting spanked by prettier.
* web: sometimes I wish I had lower standards
Anyway, this was a very stupid bug, because by definition function
definition arguments don't have uses, they're being defined, not
implemented. Fixed, conf fixed to compensate, and consequences
conquered.
* move context from labs to main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* Revert "move context from labs to main"
This reverts commit 3718ee69048966d26b1c357a7d2653fbb3ab613b.
* web: reify the data loop
I was very unhappy with the "update this dot-path" mechanism I was using earlier; it was hard
for me to read and understand what was happening, and I wrote the darned thing. I decided instead
to go with a hard substitution model; each phase of the wizard is responsible for updating the
*entire* payload, mostly by creating a new payload and substituting the field value associated
with the event.
On the receiver, we have to do that *again* to handle the swapping of providers when the user
chooses one and then another. It looks clunky, and it is, but it's *legible*; a junior dev
could understand what it's doing, and that's the goal.
* Revert "web: reify the data loop"
This reverts commit 09fedcacf02a90a021ce9e18c0eb4bec1ef48302.
* web: revert the 'lit' to 'lit-labs' for task and context.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-10-18 19:43:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s8939f574b096054a">
<source>Not you?</source>
2023-10-19 12:53:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc4eedb434536bdb4">
<source>Need an account?</source>
2023-10-19 12:53:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s38f774cd7e9b9dad">
<source>Sign up.</source>
2023-10-19 12:53:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa03aa46068460c95">
<source>Forgot username or password?</source>
2023-10-19 12:53:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s4a87445f3108db7c">
<source>Select one of the sources below to login.</source>
2023-10-19 12:53:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s091d5407b5b32e84">
<source>Or</source>
2023-10-19 12:53:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="se5fd752dbbc3cd28">
<source>Use a security key</source>
2023-10-19 12:53:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s670ad066cc0e50a3">
<source>Login to continue to <x id="0" equiv-text="${this.challenge.applicationPre}"/>.</source>
2023-10-19 12:53:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="scf5ce91bfba10a61">
<source>Please enter your password</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s85366fac18679f28">
<source>Forgot password?</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s14c552fb0a4c0186">
<source>Application requires following permissions:</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7073489bb01b3c24">
<source>Application already has access to the following permissions:</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s98dc556f8bf707dc">
<source>Application requires following new permissions:</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sbd19064fc3f405c1">
<source>Check your Inbox for a verification email.</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s8aff572e64b7936b">
<source>Send Email again.</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sdc323c6af4ae9f01">
<source>Successfully copied TOTP Config.</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s3687049d1af562c4">
<source>Copy</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s3643189d1abbb7f4">
<source>Code</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sfe1c86b42ba13376">
<source>Please enter your TOTP Code</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc2ec367e3108fe65">
<source>Duo activation QR code</source>
2023-10-23 16:48:12 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc5668cb23167e9bb">
<source>Alternatively, if your current device has Duo installed, click on this link:</source>
2023-10-25 16:25:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s721d94ae700b5dfd">
<source>Duo activation</source>
2023-10-25 16:25:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s708d9a4a0db0be8f">
<source>Check status</source>
2023-10-25 16:25:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s31fba571065f2c87">
<source>Make sure to keep these tokens in a safe place.</source>
2023-10-25 16:25:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc0a0c87d5c556c38">
<source>Phone number</source>
2023-10-25 16:25:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s04c1210202f48dc9">
<source>Please enter your Phone number.</source>
2023-10-25 16:25:37 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="seb0c08d9f233bbfe">
<source>Please enter the code you received via SMS</source>
2023-10-27 09:39:39 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s2b7dbba348234a36">
<source>A code has been sent to you via SMS.</source>
2023-10-27 09:39:39 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa84adff85b5e505c">
<source>Open your two-factor authenticator app to view your authentication code.</source>
2023-10-27 09:39:39 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7abc9d08b0f70fd6">
<source>Static token</source>
2023-10-27 09:39:39 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s844fea0bfb10a72a">
<source>Authentication code</source>
2023-11-06 12:51:41 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s3cd84e82e83e35ad">
<source>Please enter your code</source>
2023-11-06 12:51:41 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s18b910437b73e8e8">
<source>Return to device picker</source>
2023-11-14 12:05:18 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sbcf8604929b6a27a">
<source>Sending Duo push notification</source>
2023-11-14 12:05:18 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s3b68883dda2682ed">
<source>Assertions is empty</source>
2023-11-14 12:05:18 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sbbb7318812d64e51">
<source>Error when creating credential: <x id="0" equiv-text="${err}"/></source>
2023-11-14 12:05:18 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sfe199b2564b66054">
<source>Error when validating assertion on server: <x id="0" equiv-text="${err}"/></source>
2023-11-14 12:05:18 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="se409d01b52c4e12f">
<source>Retry authentication</source>
2023-11-14 12:05:18 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s8d857061510fe794">
<source>Duo push-notifications</source>
2023-11-14 12:05:18 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s47490298c17b753a">
<source>Receive a push notification on your device.</source>
2023-11-15 15:28:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s16bc281dce5685e8">
<source>Authenticator</source>
2023-11-15 15:28:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sdefec5401bf67eba">
<source>Use a security key to prove your identity.</source>
2023-11-15 15:28:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sd6a025d66f2637d1">
<source>Traditional authenticator</source>
2023-11-15 15:28:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb25e689e00c61829">
<source>Use a code-based authenticator.</source>
2023-11-15 15:28:56 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s9e568afec3810bfe">
<source>Recovery keys</source>
2023-11-15 22:14:30 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb17e8c70f9a05c77">
<source>In case you can't access any other method.</source>
2023-11-16 23:07:21 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s97f2dc19fa556a6a">
<source>SMS</source>
2023-11-18 00:55:48 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s0e516232f2ab4e04">
<source>Tokens sent via SMS.</source>
2023-11-18 00:55:48 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6ae0d087036e6d6d">
<source>Select an authentication method.</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sac17f177f884e238">
<source>Stay signed in?</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s859b2e00391da380">
<source>Select Yes to reduce the number of times you're asked to sign in.</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc1589121ae2f5f92">
<source>Enter the code shown on your device.</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s67664f8ee9aea98d">
<source>Please enter your Code</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s455a8fc21077e7f9">
<source>You've successfully authenticated your device.</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s3ab772345f78aee0">
<source>Flow inspector</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s502884e1977b2c06">
<source>Next stage</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb3fa80ccfa97ee54">
<source>Stage name</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sbea3c1e4f2fd623d">
<source>Stage kind</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s2bc8aa1740d3da34">
<source>Stage object</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc3e1c4f1fff8e1ca">
<source>This flow is completed.</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s342eccabf83c9bde">
<source>Plan history</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb2f307e79d20bb56">
<source>Current plan context</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sa13e6c8310000e30">
<source>Session ID</source>
web/admin: revise wizard form handling (#7331)
* web: break circular dependency between AKElement & Interface.
This commit changes the way the root node of the web application shell is
discovered by child components, such that the base class shared by both
no longer results in a circular dependency between the two models.
I've run this in isolation and have seen no failures of discovery; the identity
token exists as soon as the Interface is constructed and is found by every item
on the page.
* web: fix broken typescript references
This built... and then it didn't? Anyway, the current fix is to
provide type information the AkInterface for the data that consumers
require.
* web: extract the form processing from the form submission process
Our forms have a lot of customized value handling, and the function `serializeForm` takes
our input structures and creates a JSON object ready for submission across the wire for
the various models provided by the API.
That function was embedded in the `ak-form` object, but it has no actual dependencies on
the state of that object; aside from identifying the input elements, which is done at the
very start of processing, this large block of code stands alone. Separating out the
"processing the form" from "identifying the form" allows us to customize our form handling
and preserve form information on the client for transactional purposes such as our wizard.
w
* web: multi-select, but there's a styling issue.
* web: provide a closed control for multi-select
This commit creates a new control, using the ak-form-element-horizontal as a *CLOSED*
object, for our multi-select. This control right now is limited to what we expect to
be using in the wizard, but that doesn't mean it can't be smarter in the future.
* web: hung up by a silly spelling error
* web: update the form-handling method
With the `serializeForm` method extracted, it's much easier to examine and parse
every *form* with every keystroke, preserving them against the changes that
happen as the customer navigates the Wizard. With that in place, it became
straightforward to retrofit the "handle changes to the application, to the provider, and to the providerType"
into the three pages of the wizard, and to provide *all* of the form elements in a base class
such that no specialized handling needs to happen to any of the child pages.
Fixed an ugly typo in the oauth2 provider, as well.
* web: wizard should work with multi-select and should reflect default values
(Note: This commit is predicated on both the "Extract serializeForm function from Form.ts" and
"Provide a controlled multi-select input control" PRs.)
The initial attempt at the wizard was woefully naive in its implementation, missing some critical
details along the way. This revision starts off with one stronger assumption: trust that Jens knows
what he's doing, and knew what he was building when he wrote the initial `Form` handler.
The problem with the `Form` handler, and the reason I avoided it, was simply that it does too many
things, especially in its ModelForm variant: it receives a model from the back-end, renders a
(hand-written) form for that model, allows the user to interact with that model, and facilitates
saving it to the back-end again, complete with on-page notifications of success or failure.
The Wizard could not use all of that. It needs to gather the information for *two* models (an
Application and a Provider, plus the ProviderType) and has a new and specialized end-point for a
transaction that allows the committing or roll back of both models to happen simultaneously,
predicated on success or failure respectively.
With "Extract `serializeForm` completed, it was possible to repurpose the forms that already
existed, stripping them down to just their input components, and eventing the entire thing in a
single event loop of "events flow up, data flows down." In this case, the *entire form* is
serialized on a per-event basis and pushed up the to the orchestration layer, which saves them off.
Writing a parent `BasePanel` class that has accessors for `formValues` and `valid` means that the
state of every page is accessible with a simple query. This simplified the `BaseProviderPanel` class
to just specialize the `dispatchUpdate` method to send the wizard update with the new provider
information filled out.
Because the *form* is being treated as the source of truth about the state of a `Partial<Application>`
or `Partial<*Provider>` object, the defaults are now being captured as expected.
Likewise, this simplified the `providerCache` layer which preserves customer input in the event that
the customer starts filling out the wrong provider to a simple conditional clause in the
orchestrator. The Wizard has much fewer smarts because it doesn't (and probably never did) need
them.
Along with the above changes, the following has also been done:
For SAML and SCIM, the providerMappings now works. They weren't being managed as `state` objects,
so they weren't receiving updates when the update event retrieved the information from the back-end.
In order to make clear what's happening, I have extracted the loops from the original definition and
built them as named objects: `propertyMappings`, `pmUserValues`, `pmGroupValues` and so on, which I
then pass into the new multi-select component.
I fixed a really embarrassing typo in Oauth2's "advanced settings" block.
I have extracted the CoreGroup search-select into a custom component.
I deleted the `merge` function. That was a faulty experiment with non-deterministic outcomes, and I
was never happy with it. I'm glad its gone.
I've added a title header to each of the providers, so the user can be sure that they're looking
at the right provider type when they start filling out the form.
I've created a new token, `data-ak-control`, with which we can mark all objects that we can treat as
Authentik value-producing components, the form value of which is available through a `json()`
method. I've added this bit of intelligence to the `serializeForm` function, short-circuiting the
complex processing and putting the "this is the shape of the value we expect from this input" *onto
the input itself*. Which is where it belongs.
* web: add error handling to wizard.
* web: improve error handling in light components
Rather than reproduce the error handling across all of the LightComponents,
I've made a parent class that takes the common fields to distribute between
the ak-form-element-horizontal and the input object itself. This made it
much easier to properly display errors in freeform input fields in the
wizard, as well as working with the routine error handling in Form.ts
* Added the radio control to the list of LightComponents.
* Fix bug where event was recorded twice.
* Fixed merge bug (?) that somehow deleted the Authorization Select block in OAuth2.
* web: prettier had opinions
* web: added error handling and display
* web: bump @lit-labs/context from 0.4.1 to 0.5.1 in /web
Bumps [@lit-labs/context](https://github.com/lit/lit/tree/HEAD/packages/labs/context) from 0.4.1 to 0.5.1.
- [Release notes](https://github.com/lit/lit/releases)
- [Changelog](https://github.com/lit/lit/blob/main/packages/labs/context/CHANGELOG.md)
- [Commits](https://github.com/lit/lit/commits/@lit-labs/context@0.5.1/packages/labs/context)
---
updated-dependencies:
- dependency-name: "@lit-labs/context"
dependency-type: direct:production
update-type: version-update:semver-minor
...
Signed-off-by: dependabot[bot] <support@github.com>
* web: updated wizard to run with latest package.json configuration
Apparently, there were stale dependencies in package-lock.json that were conflicting
with the requests in our package.json. By running `npm update`, I was able to resolve
the conflict.
I have also removed the default names from the context names collection; they weren't doing
any good, and they permit frictionless renaming of dependencies, which is never a good
idea.
* web: schlepping on the errors messages
During testing, I realized I was unhappy with the error messages. They're not very helpful.
By adding links to navigate back to the place where the error occurred, and providing better
context for what the error could have been, I hope to help the use correct their errors.
* make package the same as main
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
---------
Signed-off-by: dependabot[bot] <support@github.com>
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-12-06 11:28:19 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6fe64b4625517333">
<source>Powered by authentik</source>
2023-12-06 17:06:07 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sdf34a5599d66f85c">
<source>Background image</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7fa4e5e409d43573">
<source>Error creating credential: <x id="0" equiv-text="${err}"/></source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s9d95f09deb601f34">
<source>Server validation of credential failed: <x id="0" equiv-text="${err}"/></source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6c8f05e3be04f62a">
<source>Register device</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sba65ae54d6585c1a">
<source>Unread notifications</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7cfe12cd14df9950">
<source>Sign out</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s5599c62bb78c631f">
<source>Admin interface</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s1298e361e40ee1c5">
<source>Stop impersonation</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6abff64e7ff7fde9">
<source>Avatar image</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s28cbd874ba450b4e">
<source>Less details</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s8fa26f65aed77c96">
<source>More details</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s3fb39fc45e840f78">
<source>Refer to documentation</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sc741dfb09d3395f0">
<source>No Applications available.</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sf34026321b35315c">
<source>Either no applications are defined, or you don’ t have access to any.</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s1cf2298d92c327a6">
<source>My Applications</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s2656433a3b1f7e86">
<source>My applications</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s06c92148da82be0d">
<source>Change your password</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sff50532a2d85e32e">
<source>Change password</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="saf63d34c8601dd41">
<source><x id="0" equiv-text="${prompt.label}"/></source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s045c3b86aae073c1">
<source>Delete account</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s4a6aa26413287069">
<source>Successfully updated details</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6fcd9b5a87ceccd6">
<source>Open settings</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s8c05cccd470f6b5f">
<source>No settings flow configured.</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sb546eb04425e07fa">
<source>Update details</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sea2f00b34b385a43">
<source>Successfully updated device.</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s6a406aecb2c0e5c5">
<source>Enroll</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sabb56f74492e7e96">
<source>Update Device</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s30205d424e710818">
<source>Successfully disconnected source</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s67dedada007d4067">
<source>Failed to disconnected source: <x id="0" equiv-text="${exc}"/></source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sd2208cd1a767644b">
<source>Disconnect</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7a4f059aaa029719">
<source>Connect</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sababff57115130a0">
<source>Error: unsupported source settings: <x id="0" equiv-text="${source.component}"/></source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sd1031bddc66dc495">
<source>Connect your user account to the services listed below, to allow you to login using the service instead of traditional credentials.</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s7968dbed9b106c29">
<source>No services available.</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s3a135682bd30bdbb">
<source>Create App password</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s588796ee929a2e4c">
<source>User details</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="s332a5235948c1a1d">
<source>Consent</source>
2023-12-08 01:52:59 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
<trans-unit id="sff945d3f59b93c5e">
<source>MFA Devices</source>
</trans-unit>
<trans-unit id="sc54aafeea9c9bab0">
<source>Connected services</source>
2023-06-12 13:41:44 +00:00
</trans-unit>
2023-12-20 08:45:34 +00:00
</body>
</file>
web: Replace lingui.js with lit-localize (#5761)
* \#\# Details
web: replace lingui with lit/localize
\#\# Changes
This rather massive shift replaces the lingui and `t()` syntax with lit-localize, XLIFF, and the `msg()`
syntax used by lit-localize. 90% of this work was mechanized; simple perl scripts found and replaced
all uses of `t()` with the appropriate corresponding syntax for `msg()` and `msg(str())`.
The XLIFF files were auto-generated from the PO files. They have not been audited, and they should be
checked over by professional translators. The actual _strings_ have not been changed, but as this was
a mechanized change there is always the possibility of mis-translation-- not by the translator, but by
the script.
* web: revise lit/localize: fix two installation issues.
* web: revise localization
TL;DR:
- Replaced all of Lingui's `t()` syntax with `msg()` syntax.
- Mechanically (i.e with a script) converted all of the PO files to XLIFF files
- Refactored the localization code to be a bit smarter:
- the function `getBestMatchLocale` takes the locale lists and a requested locale, and returns the
first match of:
- The locale's code exactly matches the requested locale
- The locale code exactly matches the prefix of the requested locale (i.e the "en" part of "en-US")
- the locale code's prefix exactly matches the prefix of the requested locale
This function is passed to lit-locate's `loadLocale()`.
- `activateLocale()` just calls `loadLocale()` now.
- `autodetectLanguage` searches the following, and picks the first that returns a valid locale
object, before passing it to `loadLocale()`:
- The User's settings
- A `?locale=` component found in `window.location.search`
- The `window.navigator.language` field
- English
The `msg()` only runs when it's run. This seems obvious, but it means that you cannot cache
strings at load time; they must be kept inside functions that are re-run so that the `msg()` engine
can look up the strings in the preferred language of the user at that moment.
You can use thunks-of-strings if you really need them that way.
* Including the 'xliff-converter' in case anyone wants to review it.
* The xliff-converter is tagged as 'xliff-converter', but has been
deleted.
\#\# Details
- Resolves #5171
\#\# Changes
\#\#\# New Features
- Adds a "Add an Application" to the LibraryView if there are no applications and the user is an administrator.
\#\#\# Breaking Changes
- Adds breaking change which causes \<issue\>.
\#\# Checklist
- [ ] Local tests pass (`ak test authentik/`)
- [ ] The code has been formatted (`make lint-fix`)
If an API change has been made
- [ ] The API schema has been updated (`make gen-build`)
If changes to the frontend have been made
- [ ] The code has been formatted (`make web`)
- [ ] The translation files have been updated (`make i18n-extract`)
If applicable
- [ ] The documentation has been updated
- [ ] The documentation has been formatted (`make website`)
* web: fix redundant locales for zh suite.
* web: prettier pass for locale update
* web: localization moderization
Changed the names of the lit-localize commands to make it clear they're
part of the localization effort, and not just "build" and "extract".
* update transifex config
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* fix package lock?
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* use build not compile
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
* web: conversion to lit-localize
The CI produced a list of problems that I hadn't caught earlier,
due to a typo ("localize build" is correct, "localize compile" is
not) I had left in package.json. They were minor and linty, but
it was still wise to fix them.
* web: replace lingui with lit/locale
This commit fixes some minor linting issues that were hidden by a typo in package.json. The
issues were not apparently problematic from a Javascript point of view, but they pointed
to sloppy thinking in the progression of types through the system, so I cleaned them
up and formalized the types from LocaleModule to AkLocale.
* web: replace lingui with lit/localize
One problem that has repeatedly come up is that localize's templates do not produce
JavaScript that conforms with our shop style. I've replaced `build-locale` with
a two-step that builds the locale *and* ensures that it conforms to the shop style
via `prettier` every time.
* web: replace lingui with lit-locale
This commit applies the most recent bundle of translations to the
new lit-locale aspect component. It also revises the algorithm
for *finding* the correct locale, replacing the complex fall-back
with some rather straightforward regular expressions.
In the case of Chinese, the fallback comes at the end of the
selection list, which may not be, er, politically valuable
(since Taiwan and Hong Kong come before, being exceptions that
need to be tested). If we need a different order for presentation,
that'll be a future feature.
* web: replace lingui with lit/locale
Well, that was embarassing.
---------
Signed-off-by: Jens Langhammer <jens@goauthentik.io>
Co-authored-by: Jens Langhammer <jens@goauthentik.io>
2023-06-02 15:08:36 +00:00
</xliff>