Prior to this fix we would show the message after a round trip to the server. If you had a too long message error, at this point your input would be empty and we would show an error in chat. It's important to have this server side safety net, but we can have a better UX by showing an error on the frontend before sending the message, that way you can correct your message before sending it and not lose it.
Fixes a flaky test. `afterUpdate` from chart.js is not integrated with the runloop and component lifecycle, as a result we have no guarantee on when it will happen. The easiest change we can do for now is ensuring we actually have the DOM we expect to have, and if not, we exit early.
Co-authored-by: Joffrey JAFFEUX <j.jaffeux@gmail.com>
Co-authored-by: merefield <merefield@gmail.com>
The "migration to Glimmer" has been broken out here from #27155 to make the review process less onerous and reduce change risk:
* DEV: migrates most of the widget code to Glimmer in prep for IRV additions
* NB This already incorporates significant amounts of review and feedback from the prior PR.
* NB because there was significant additional feedback relating to older Poll code that I've improved with feedback, there are some additional changes here that are general improvements to the plugin and not specific to IRV nor Glimmer!
* There should be no trace of IRV code here.
Once this is finalised and merged we can continue to progress with #27155.
Follow up to #27631 to account for group mentions in channels.
We only want to show mentions for groups that are currently mentionable and those that the current user belongs to.
I am changing many of these to notes or resolving them as is,
most of these I have not actively worked on in years so someone
else can work on them when we get to these areas again.
A previous refactor has prevented errors to show correctly. The guilt of the issue is that we were not calling the error variable correctly in the templates.
This commit also adds a spec for this case, and removes the need for `I18n.backend.store_translations` in specs so we don't have to write too much boilerplate each time we write a spec.
We want to allow admins to make new required fields apply to existing users. In order for this to work we need to have a way to make those users fill up the fields on their next page load. This is very similar to how adding a 2FA requirement post-fact works. Users will be redirected to a page where they can fill up the remaining required fields, and until they do that they won't be able to do anything else.
In 53b3d2f0dc we introduced a stricter BBCode Tag parser. It prevents having "values" with spaces when they're not surrounded by a valid pair of quotes.
The `[details=` BBCode Tag is popular enough that it's worth adding a special case for it (especially since it doesn't support other parameters).
This also adds the Finnish pair of quotes.
Context - https://meta.discourse.org/t/details-accepts-only-one-word-as-summary/313019
* DEV: Upgrade Rails to 7.1
* FIX: Remove references to `Rails.logger.chained`
`Rails.logger.chained` was provided by Logster before Rails 7.1
introduced their broadcast logger. Now all the loggers are added to
`Rails.logger.broadcasts`.
Some code in our initializers was still using `chained` instead of
`broadcasts`.
* DEV: Make parameters optional to all FakeLogger methods
* FIX: Set `override_level` on Logster loggers (#27519)
A followup to f595d599dd
* FIX: Don’t duplicate Rack response
---------
Co-authored-by: Jarek Radosz <jradosz@gmail.com>
* DEV: Upgrade Rails to 7.1
* FIX: Remove references to `Rails.logger.chained`
`Rails.logger.chained` was provided by Logster before Rails 7.1
introduced their broadcast logger. Now all the loggers are added to
`Rails.logger.broadcasts`.
Some code in our initializers was still using `chained` instead of
`broadcasts`.
* DEV: Make parameters optional to all FakeLogger methods
* FIX: Set `override_level` on Logster loggers (#27519)
A followup to f595d599dd
* FIX: Don’t duplicate Rack response
---------
Co-authored-by: Jarek Radosz <jradosz@gmail.com>
* Revert "FIX: Set `override_level` on Logster loggers (#27519)"
This reverts commit c1b0488c54.
* Revert "DEV: Make parameters optional to all FakeLogger methods"
This reverts commit 3318dad7b4.
* Revert "FIX: Remove references to `Rails.logger.chained`"
This reverts commit f595d599dd.
* Revert "DEV: Upgrade Rails to 7.1"
This reverts commit 081b00391e.
There is currently only one scenario when both the composer and a user card would be present at the same time:
if you have the composer open and then you click on something outside it that triggers a card. Which implies intent to see the card (unobstructed by the composer 😉)
The reverse doesn't happen because opening the composer would close an existing user card.
In theory there's also displaying a user card by clicking on a mention in composer's preview but that functionality is currently broken (and this PR is a prerequisite 😉)
---
I changed `.user-card, .group-card` to `.fk-d-menu[data-identifier="card"]` because that regressed when we moved user cards to float-kit – they are nested inside `.fk-d-menu` so its `z-index` is now important (effectively the cards had `z-index: z("dropdown")` instead of `z("usercard")`)
Wasn't quite handling the cases where a closing bracket `]` was used in the value of one of the attributes.
```markdown
[chat quote=user channel="[broken]"]
```
Would not be correctly parsed because we would _greedily_ use the first `]` as the end of the tag even though it might be a valid character when inside proper quotes.
c39a4de139/app/assets/javascripts/discourse-markdown-it/src/features/bbcode-block.js (L62)
Re-wrote the `parseBBCodeTag` to properly handle the following cases
- A closing tag (aka `[/name]`) which are easy since they don't have any attributes
- An old `[quote=...]` format we used that doesn't uses quotes but still has various attributes of the form `key:value`
- All three valid BBCode opening tag formats we support
- `[name]` without any attributes
- `[name=foo]` with a default value
- `[name foo=bar]` with some attributes
Ended up having to fix/rewrite the few bbcode rules that were using the `parseBBCodeTag` function, namely `d-wrap` and `discourse-local-dates`.
While working on this, I think I also found a way to get rid the of shims we had in place so that plugins could use the `parseBBCodeTag` function.
Reference - https://meta.discourse.org/t/having-a-right-bracket-in-a-channel-name-breaks-all-quotes-from-that-channel/308439
* remove default prop values where they're being set in constructor
* replace some `||` operators in constructors with `??` so the fallback boolean values are actually used
* remove an unused service injection (and sort the rest)
* remove unused prop
* inline an arg check
* remove an unnecessary `?.` operator
* sort element attributes
This change replaces the chat drawer tabs with new drawer routes for channels, direct messages and threads.
The main objective is to improve navigation within drawer, now that we have separation of chat sections in drawer.
The order of chat direct message groups can sometimes place usernames in an unexpected order, this change tests multiple combinations of usernames and accepts them no matter what order they are in.
In 4e7a75a7ec, we moved to a single admin plugin page and added a few fields to the "plugin serializer" but we already had a proper route with the correct serializers to properly load channels.
This fixes it by removing the "add_to_serializer" calls and changed the calls to "/admin/plugins/chat.json" to the proper "/admin/plugins/chat/hooks.json" route.
Meta - https://meta.discourse.org/t/names-are-missing-from-list-when-creating-new-chat-channel-webhooks/308481
When chat is enabled, there's a scheduled job that runs every 5 minutes to check whether we need to send a "chat summary" email to users with unread chat messages or mentions.
On Discourse with a large number of users, the query used wasn't optimal and sometimes taking minutes. Which isn't good when the query is called every 5 minutes 😬
This PR reworks the query in `Chat::Mailer.send_unread_mentions_summary`.
Instead of starting from the `users` table, it starts from the `user_chat_channel_memberships` table which is the main piece tying everything together.
The new query is mostly similar to the previous one, with some bug fixes (like ensuring the user has `allow_private_messages` enabled for direct messages) and is also slightly simpler since it doesn't keep track of the `memberships_with_unread_messages` anymore. That part has been moved to the `user_notifications.chat_summary` email method.
The `UserEmailExtension` has been deleted since that was using to N+1 update the `user_chat_channel_memberships.last_unread_mention_when_emailed_it`(quite a mouthful 😛) but that's now done directly in the `user_notifications.chat_summary` email method.
The "plat de résistance" of that PR - the `user_notifications.chat_summary` method has been re-worked for improved performances 🚀
Instead of doing everything in one query, it does 4 tiny ones.
- One to retrieve the list of unread mentions (@something) in "category" channels
- One to retrieve the list of unread messages in "direct message" channels (aka. 1-1 and group discussions)
- One to load all the chat messages for each "category" channels from the last unread mention
- One to load all the chat messages for each "direct message" channels from the last unread message
All the specs for both `Chat::Mailer` and `UserNotification.chat_summary` have been rewriten for easier comprehension and faster execution (mostly by not using chat services which makes the specs go 10x slower...)
Internal ref - t/129848
When adding custom translations for tests using `I18n.backend.store_translations`,
we need to remove the custom translations at the end of each test to
prevent the custom translations from leaking to other tests.
This change prevents explicitly declaring each route that should be intercepted for chat drawer mode.
In theory all chat drawer routes should be intercepted from the main chat routes file and therefore we would only need to add new drawer routes directly within chat-drawer-router.js.
This change allows chat drawer users to edit channel settings and members without leaving drawer mode. If a channel is open within chat drawer and the user clicks the Channel name, it will load channel settings within the drawer.
This change allows the correct number of members to be added when creating a group direct message, based on the site setting chat_max_direct_message_users.
Previously we counted the current user within the max user limit and therefore the count was off by 1.
Before this fix we could only list messages of a thread if it was part of a `threading_enabled` channel or if the thread was set to `force`.
Due to our design of also using a thread id when this is just a chain of replies so we can switch from threading enabled to disabled at any time, we will allow `Chat:: ListChannelThreadMessages` to list the messages of any thread, the only important requirements are:
- having a thread id
- being able to access this thread
To allow this, this commit simply removes the check on `threading_enabled` or `force`.
Under some circumstances, the TextField component could trigger a `Assertion Failed: You attempted to update attrs on ..., but it had already been used previously in the same computation...` error, causing the Ember app to crash.
A few follup changes after changing to the chat footer split for drawer:
* Fixing a bug that stretched the unread indicator on mobile
* Minor style changes in hover/focus behaviour for chat drawer
* Repositioning of unread indicator so it has more space at the top of the footer
* Using the `c-unread-indicator` mixin
Was "removing" (rather not re-applying) the `[spoiler]` BBCode because we were testing the **whole** class of the `span`/`div` was `spoiled` but we added another class and thus broke this functionnality.
In order to fix this issue, the test to determine whether a `span`/`div` is a spoiler, now uses a regular expression to check whether the `class` **contains** the word `spoiled`.
Reference - https://meta.discourse.org/t/quoting-spoiler-text-doesnt-include-spoiler-tags-in-the-quote/170145
Adds a placeholder image + CTA in chat, for empty channel and DM lists.
On desktop with drawer mode, we split chat into tabs (like mobile).
---------
Co-authored-by: Joffrey JAFFEUX <j.jaffeux@gmail.com>
Co-authored-by: David Battersby <info@davidbattersby.com>
Co-authored-by: Régis Hanol <regis@hanol.fr>
In this PR service objects were moved to Core https://github.com/discourse/discourse/pull/26506
However, ServiceRunner should be moved as well. Mostly for CI to run effortlessly without loading plugins.
Prior to this fix we had too logic to detect if a user is active or not:
- idle codepath on the frontend
- online user ids on the backend
The frontend solution is not very reliable, and both solution are just trying to be too smart. Making a lot of people questioning why they receive a notification sometimes and sometimes not. This commit removes all this logic and replaces it with a much more simpler logic:
- you can't receive notifications for channel you are actually watching
- we won't play a sound more than once every 3seconds
The `onNotification` signature is:
```
onNotification(data, siteSettings, user, appEvents)
```
And we were not passing `appEvents`.
Also explicitly inject `currentUser` in `chat-notification-manager` service.
We consider that you should always receive a notification sound when someone speaks directly with you in chat.
This commit also refactors the way we play audio in chat to make it simpler and throttle it to 3 seconds.
We also added a safeguard to ensure we won't play sounds for old messages, this case can happen when message bus is catching up the backlog (eg: in an inactive tab for example).
When users click a link that points to an existing group chat, we should reopen that chat instead of creating a new group chat so users can more easily continue ongoing conversations.
activeChannel is something we should use less and less as it could not exist, in this case we have the channel right here in the function so there's no reason to reach for `this.chat.activeChannel`.
Prior to this fix we wouldn't intercept it, and we also wouldn't handle it, which in result would cause us to handle as a full page interaction and open the full page chat even if you were in drawer mode.
When a user had the chat option "Show activity indicator in header" set to "all new messages", and they would get a reply to a thread they're part of, the chat icon in the header would not show the unread bubble indicator.
In order to fix this, the `ChatHeaderIconUnreadIndicator` component will now `showUnreadIndicator` whenever there is either one unread public channel or there are unread threads.
I only added a system spec for this very specific path because I don't want to slow down the whole suite to test for all the various combination of the `chat_header_indicator_preference` values.
Internal ref - t/128874
Not 100% sure why the changes in `PrettyText.format_for_email` raised this issue, but we were missing adding a link to the Discourse instance whenever we are replacing the elided part of a post with a link to either the post or the Discourse instance in the email.
Also reformated the specs using better variable names (sometimes a variable named `md` would contain some html) and used the `match_html` helper for all the tests.
Given this is currently buggy and we have the cancel button Im not sure this is actually necessary. I feel like it's adding a lot of noise of low value.
For now at least, it's better without it.
This commit reuses the existing codepath in desktop-notifications and make it available to use to chat.
primaryTab was too hard to test if not impossible in this service test, however isIdle and disabled notifications are correctly tested.
Prior to this change, only mentions would get a notification and a sound. This change will not create a notification for this case, but will play a sound. This is still respecting notification settings, not playing the sound when you are viewing the channel or not following it.
---------
Co-authored-by: Régis Hanol <regis@hanol.fr>
(experimental)
The initial implementation of glimmer topic-list and related components. Does not include new APIs and isn't compatible with existing customization. That's gonna come in future PRs.
Enabled by adding groups to `experimental_glimmer_topic_list_groups` setting.
Debouncing the audio was causing the audio to be lost sometimes, somewhat randomly. It's not supposed to be necessary.
The commit also refactors the code to async/await.
Prior to this fix we were using `topic.current_post_number` which is coming from the server side and represents the initial topic scroll position when initially rendered to the end user. However, this value is not dynamic and is not updated when the user scrolls, `topic.currentPost` is the dynamic equivalent.
No test as there are very low chances a test like this one based on scroll position, will be reliable over time.
... wasn't properly escaped so it would should html entities (like `'` instead of the apostrophe `'`).
I checked all the other places we show an excerpt and this was the only one that was missing the call to `htmlSafe` -> `replaceEmoji`.
Internal ref - t/128877
The translations of the labels of some of the chat preferences dropdowns were not customizable via the "site texts" feature.
This was because they were declared outside of a Controller class and were thus not taking into account the customization through "site texts".
Internal ref - t/128859
When you reply to a chat message, we [always create a thread][1]. But when the channel we're in doesn't have threading enabled, the reply is _technically_ not a thread.
This changes the `in_thread?` method to check for both the presence of a `thread_id` and to ensure that the channel has `threading_enabled`.
Internal ref - t/128103/3
[1]: e6e3eaf472/plugins/chat/app/services/chat/create_message.rb (L110-L115)
Whenever you get a bookmark notification, a mention notification, or click on a bookmark on a message in a long thread, we should ensure we always highlight and show the proper message.
Before this fix, we would correctly load the thread, but would always start at the bottom.
Internal ref. t/128103
For both `chat_allowed_groups` and `chat_message_flag_allowed_groups`,
this commit removes the `is_staff?` guardian check, and instead
adds both `moderators` and `admins` auto groups as `mandatory_values`
to those settings, as part of an ongoing effort to do this for
group-based setting values.
The automation plugin has 4 custom field types that are array typed. However, array typed custom fields are deprecated and should be migrated to JSON type.
This commit does a couple of things:
1. Migrate all four custom fields to JSON
2. Fix a couple of small bugs that have been discovered while migrating the custom fields to JSON (see the comments on this commit's PR for details https://github.com/discourse/discourse/pull/26939)
Some combinations of start_date and frequency/interval values can cause a recurring automation rule to either trigger before its start_date or never trigger. Example repros:
- Configure a recurring automation with hourly recurrence and a start_date several days ahead. What this will do is make the automation start running hourly immediately even though the start_date is several days ahead.
- Configure a recurring automation with a weekly recurrence and a start_date several weeks ahead. This will result in the automation never triggering even after the start_date.
These 2 scenarios share the same cause which is that the automation plugin doesn't use the start_date as the date for the first run and instead uses the frequency/interval values from the current time to calculate the first run date.
This PR fixes this bug by adding an explicit check for start_date and using it as the first run's date if it's ahead of the current time.
This change adds a wrapper link around the thread list details on mobile to make the click area larger.
We also update child div elements to span to ensure valid html, since the link is an inline element and divs are block level.
---------
Co-authored-by: chapoi <101828855+chapoi@users.noreply.github.com>
We were missing two `getURL` calls.
The test is now written for subfolder but it's good enough. If it's working for subfolder, it's working for non subfolders, the opposite being false.
menus and tooltips are now appended to their own portals. The service are the only responsible for managing the instances, prior to this commit, services could manage one instance, but the DMenu and DTooltip components could also take over which could cause unexpected states.
This change also allows nested menus/tooltips.
Other notable changes:
- few months ago core copied the CloseOnClickOutside modifier of float-kit without removing the float-kit one, this commit now only use the core one.
- the close function is now trully async
- the close function accepts an instance or an identifier as parameter
Whenever a post already failed "lightweight" validations, we skip all the expensive validations (that cooks the post or run SQL queries) so that we reply as soon as possible.
Also skip validating polls when there's no "[/poll]" in the raw.
Internal ref - t/115890
- prevents re-rendering avatars while updating messages quickly in the thread preview indicator
- ensures the cancel button is shown when you are admin OR when the streamed message is a reply to the current user
Prior to this commit, only system users had this pass.
Another significant change of the PR, is to make membership of a channel the angular stone of the permission check to create/update/stop streaming a message. The idea being, if you are a member of a channel already we don't need to check if you can join it AGAIN.
We also have `Chat::AutoRemove::HandleCategoryUpdated` which will deal with permissions change so it's simpler and less prone to error to consider the membership as the only source of truth.
There's no point checking if a user can join a channel if they are already part of it. This case was frequent when using `enforce_membership: true` for custom bots for example.
This case had not been tested end to end as `Discourse.track_events` was not working when wrapping `send_message`. Because of this lack of end to end test, a regression has been created when renaming the expected context properties. This commit fixes the regression and write a slightly convulted, but effective end to end test.
This commit introduces several enhancements to the ChatSDK module, aiming to improve the functionality and usability of chat thread interactions. Here's what has been changed and added:
1. **New Method: `first_messages`:**
- Added a method to retrieve the first set of messages from a specified chat thread.
- This method is particularly useful for fetching initial messages when entering a chat thread.
- Parameters include `thread_id`, `guardian`, and an optional `page_size` which defaults to 10.
- Usage example added to demonstrate fetching the first 15 messages from a thread.
2. **New Method: `last_messages`:**
- Added a method to retrieve the last set of messages from a specified chat thread.
- This method supports reverse pagination, where the user may want to see the most recent messages first.
- Similar to `first_messages`, it accepts `thread_id`, `guardian`, and an optional `page_size` parameter, defaulting to 10.
- Usage example provided to illustrate fetching the last 20 messages from a thread.
The TextCleaner step has been moved from chat message’s validation to create_message/update_message services. It allows us to easily tweak part of its behavior depending on the needs.
For example we will now disable strip_whitespaces by default when streaming messages as we want to keep newlines and spaces at the end of the message.
Recurring automations are triggered by a scheduled job that runs every minute and checks for due automations, runs them and then marks as them as completed (by deleting the `PendingAutomation` record). However, the job is currently subject to a race condition where a recurring automation can be executed more than once at its due date if it takes more than a minute to finish.
This commit adds a mutex around the code that triggers the recurring automation so that no concurrent executions can happen for a single automation.
Meta topic: https://meta.discourse.org/t/daily-summary-9pm-utc/291850/119?u=osama.
It's currently possible to setup multiple automation rules that trigger each other resulting in an infinite loop. To prevent that, this commit adds a global "circuit breaker" that prevents all automations from triggering while an automation rule is executing.
Internal topic: t/124365.
Those were all low hanging fruits - all were already glimmer components, so this was mostly merging js and hbs files and adding imports.
(occasionally also adds/fixes class names)
This change encourages users to title their threads to make it easier for other users to join in on conversations that matter to them.
The creator of the chat thread will receive a toast notification prompting them to add a thread title when on mobile and the thread has at least 5 sent replies.
That could cause flakeyness in specs depending in which timing message bus would arrive and it's not necessary as it should be updated with `handleThreadOriginalMessageUpdate`.
Follow up to #26712 to account for older threads that don't have a persisted excerpt, as this was previously generated on every page load.
This change allows us to build the excerpt on the fly when none exists, fixing the issue of missing message excerpts for thread previews (within channel) and thread lists (on mobile/desktop).
This change moves the chat message excerpt into a new database column (string) on the chat_messages table.
As part of this change, we will now set the excerpt within the `Chat::CreateMessage` service, and update it within the `Chat::UpdateMessage` service.
This commit will now allow us to track read position in a thread and returns to this position when you open the thread.
Note this commit is also extracting the following components to make it possible:
- `<ChatMessagesScroller />`
- `<ChatMessagesContainer />`
The `UpdateUserThreadLastRead` has been updated to allow this.
Various refactorings have also been done to the code and specs to improve the support of last read.
It's important to close the modal or we will just remove it from screen without calling callbacks, which will cause the body to be locked on iOS.
It's hard to test this behavior, as it only happens on iOS and the modal will disappear anyways, it's only a matter of ensuring it's closed correctly.
This is reverting part of 08ff0bac29 to only have the call on exit channel. This was causing incorrect unread update. I will refactor this in another commit.
Prior to this fix, any change to an automation would reset `pending_automations`, now we only do it if any value related to recurrence (start_date, interval, frequency, execute_at...) has been changed.
It means that any trigger creating `pending_automations` now needs to manage them in the `on_update` callback.
This commit adds a new option to the `user_updated` trigger of the automation plugin to only trigger an automation for new users that join after the automation is enabled.
Internal topic: t/125829/9.
This is to enable :array type attributes for Contract
attributes in services, this is a followup to the move
of services from chat to core here:
cab178a405
Co-authored-by: Joffrey JAFFEUX <j.jaffeux@gmail.com>
Prior to this fix we were calling the action before closing the menu which could cause various callbacks, like the enable body scroll one, to not be called as some actions will do: `chat.activeMessage = null;` causing the message actions menu to be instantly destroyed.
This method name is a bit confusing; with_secure_uploads implies
it may return a block or something with the uploads of the post,
and has_secure_uploads implies that it's checking whether the post
is linked to any secure uploads.
should_secure_uploads? communicates the true intent of this method --
which is to say whether uploads attached to this post should be
secure or not.
This commit adds a new automation script for creating topics. It's very similar to the existing 'create a post' automation, except that it posts new topics in a specific category and with optional tags.
Internal topic: t/125829.
The complexity of the situation is that we don't want to load faker into production by default but fabricators and styleguide are available on production.
This is made possible through app/assets/javascripts/discourse/app/lib/load-faker.js which contains a function to ensure faker is loaded asynchronously (loadFaker) and another function to access the loaded faker (getLoadedFaker).
Note 1: this commit also refactors fabricators to have access to context and use faker where possible
Note 2: this commit moves automation to admin bundle
---------
Co-authored-by: David Taylor <david@taylorhq.com>
When selecting messages to move to a new channel, if any of the selected messages is the original message of a thread, the entire thread, including all its replies, will be moved to the destination channel
When clicking a "checklist", we're parsing the raw to change the state of the checklist that was clicked. If there was an image URL with an empty alternative text in the raw "" it would consider the "[]" as an empty checklist item.
Internal - t/124499
Available as a normal synchronous module in tests
Available as an async import in core, or via the `loadFaker` helper in themes/plugins (which cannot use async import directly)
Kris
Régis Hanol
Discourse Translator Bot
Joffrey JAFFEUX
Jarek Radosz
Alan Guo Xiang Tan
ScottMastro
Robert
chapoi
Loïc Guitaut
锦心
Krzysztof Kotlarek
Keegan George
Jan Cernik
David Battersby
Martin Brennan
Ted Johansson
pangbo13
Sérgio Saquetim
Sam
Osama Sayegh
David Taylor
Natalie Tay
Andrei Prigorshnev
Mark VanLandingham
