Earlier the API endpoints related to streams accepts and returns a
field `can_remove_subscribers_group_id` which represents the ID
of user_group whose members can remove subscribers from stream.
This commit renames this field to `can_remove_subscribers_group`.
Dropping support for url_prefix for RealmPlayground, the server now uses
url_template instead only for playground creation, retrieval and audit
logging upon removal.
This does the necessary handling so that url_template is expanded with
the extracted code.
Fixes#25723.
Signed-off-by: Zixuan James Li <p359101898@gmail.com>
initialization function.
We are already importing `generated_pygments_data.json` directly in
`realm_playground.js` so there is no need to pass it via initialization
method.
This function takes the stream name (a string) and using numbers
makes it seem like it's actually using a stream id, which is
confusing. This change adjusts `stream` to always be a string.
This commit adds support for triggering audible desktop
notifications when:
* A message is sent to a followed topic with the global user setting
'enable_followed_topic_audible_notifications' enabled.
* A message with wildcard mentions is sent to a followed topic with
the global user setting
'enable_followed_topic_wildcard_mentions_notify' enabled.
This commit adds support for triggering visual desktop
notifications when:
* A message is sent to a followed topic with the global user setting
'enable_followed_topic_desktop_notifications' enabled.
* A message with wildcard mentions is sent to a followed topic with
the global user setting
'enable_followed_topic_wildcard_mentions_notify' enabled.
This commit updates the existing 'Topics' settings UI
to add support for followed topics.
Changes:
- Add "Followed" as the first option in the topic status dropdowns.
- Modify the intro text.
- Replace the "Learn more." link with a question mark linking to
'/help/mute-a-topic' on the "Topic settings" heading.
The changes are visible in the development environment only.
Fixes#25918.
Specifically when it's the last operand of the search
filter. This makes the first line of search results
more consistent with the lines that come after.
itself.
We were checking for the undefined value of `$html_item` but this
variable is always true since `$.find` never returns an undefined value
and it returns a JQuery collection instead.
So we need to check the length of that JQuery collection here to make
sure that the condition works as expected.
After migrating to TypeScript, we don't need to manually validate the
configuration passed to the `ListWidget`, that's what typescript complier
will do for us, hence I removed `validate_opts` and `valid_filter_opts`
functions.
This is a prep commit for typescript migration of this module. In this commit
I made all the users of ListWidget pass generic sort functions like 'alphabetic'
and 'numeric' explicitly to avoid type conversions which we may need when moving
to typescript, also it simplifies our `set_sorting_function` a bit more.
This is a prep commit for moving towards typescript migration of this module.
We should make `get_item` required on `list_widget` because its intended to be
used with a set of keys and then hydrate the keys lazily. Also this will help
us to avoid writing messy if-conditions while filtering and sorting and make
our life much easier when we migrate to TypeScript.
I added a `default_get_item` function which is just n identity function and
also refactored some code to make use of default get item.
Removes the `assert_number` over defensive function which was being used
for validating the type of `stream_id` but now that this module is
converted to TypeScript we do not need this function anymore since
typescript compiler will do the type checking for us.
The `stream_list.build_stream_list()` function is responsible for
rendering the stream list in the left sidebar. Previously, it uses
`topic_list.clear()` to clear the `active_widgets` that tracks active
stream(s) in the left sidebar. This led to a bug where rendering stream
list after adding or removing a stream through `build_stream_list()`
while a stream was zoomed in ("more topics" clicked) didn't fully exit
the "more topics" view. So clicking "BACK TO STREAMS" throws an exception
with the message "Error: Unexpected number of topic lists to zoom out"
because `active_widgets`, cleared by `topic_list.clear()`, became
inaccessible.
To address this issue, instead of clearing the topics list with
`topic_list.clear()`, the code now ensures the closure of the zoomed
stream by using `topic_zoom.clear_topics()`. This change ensures a
proper exit from the "more topics" view and avoids the exception.
Fixes#25670.
This commit completes the notifications part of the @topic
wildcard mention feature.
Notifications are sent to the topic participants for the
@topic wildcard mention.
This commit adds a boolean field `mentions_topic_wildcard`
to the `MessageRenderingResult` dataclass.
The field is set to true only if message rendering determines
the message has an actual topic wildcard mention in it (and not,
e.g., topic wildcard mention syntax inside a code block).
The rendered content for topic wildcard mention is
'<span class="topic-mention">{wildcard}</span>'.
The 'topic-mention' class is the identifier for the wildcard
mention being a topic wildcard mention.
We don't use 'data-user-id="*"' and "user-mention" class for
topic wildcard mentions and eventually plan to remove them for
stream wildcard mentions too in a separate mini-project.
This is a prep commit to replace 'wildcard' with 'stream_wildcard'.
This wasn't included in 179d5cb because we didn't decide to
use a different rendered_content for topic wildcard mention,
i.e., ''<span class="user-mention topic-mention">{wildcard}</span>'.
Our intention was not to create separate tests for both stream
and topic wildcard mentions, as they were expected to have the
same rendered content format.
This includes changing the URL to #settings/preferences, with a
transparent redirect so that existing links, like the one from Welcome
Bot, continue to work.
We use `get_by_user_id` instead of directly accessing the global dict, because
when accessing person objects directly from one of the global dicts we
need callers to check for undefined values, this can be fixed by using
`get_by_user_id` method to get person objects because that functions
makes sure to assert that we indeed have a valid user id, so it will
never return undefined values.
Co-authored-by: Zixuan James Li <p359101898@gmail.com>
This commit removes "@" from name of role-based system groups
since we have added a restricion on having user group names
starting with "@" in the previous commit as they look odd in
mention syntax.
We also add a migration in this commit to update the name of
role-based system groups in existing realms to remove "@"
from the name. This migration also updates the names of
non-system user groups by removing the invalid prefixes
from their names and if there is a group already with that
name, we insted name the group as "group:{group_id}".
Fixes#26148.
There was no function to retrieve users while excluding bots,
this was needed to allow for typeahead user fields that require
to exclude bots from the autocomplete suggestions.
Fixes part of #25092.
Rename `get_active_human_ids` to `get_realm_active_human_user_ids`
to better fit naming conventions and to be more clear about the
detail that it only includes active users.
Fixes part of #25092.
Users are permitted to see wildcard mentions for all/everyone
during message composition in large streams according to
`restrict wildcard mentions` policy.
Fixes#25429
With the previous commit, we now have two buttons. One button cancels
the upload, while the other simply hides the upload banner. This
commit renames the switch case inside upload.js so that instead of
'upload_banner_close_button', it is now called
'upload_banner_hide_button', which appears to be more aligned with
the actual functionality of the icon.
This commit adds a cancel button to the upload banner, replacing
the previous close icon. Now, the cancel button is used to cancel
the upload process, while the close icon is used to remove the
upload banner without interrupting the upload.
A new case has been added to the switch statement in the 'upload.js'
file to handle the functionality of hiding the banner called
'upload_banner_hide_button'.
Replaced the functionality of the 'compose_banner_close_banner' case
inside the switch statement with a new case called
'upload_banner_cancel_button'. The cancel button is now assigned
the selector 'upload_banner_cancel_button'.
`Cancel` button is only preset for banner which tracks
progress while a file is being uploaded.
To maintain consistency with other banners, the cancel button's
dimensions and color have been adjusted to match the style of other
buttons present in different banners.
Fixes: #21156
`update_recipient_bar_background_color` changes in a paint
after change in theme without using `requestAnimationFrame` to
make sure they happen in the same paint..
Replaced `setTimeout` with `requestAnimationFrame` in
`server_events_dispatch` since `requestAnimationFrame` already
ensures that they happen in the next paint, assuming that was the
intention of `setTimeout` being preset there.
We should make `get_by_user_id` type-safe in the sense that it's caller
should not expect undefined values and hence it's caller should not
need to validate the user ids.
To do this I extracted a method `maybe_get_user_by_id` which is
type-unsafe version of this function which will allow undefined values
and will behave exactly the same as the previous version. So the callers
which expects the undefined values from this function should use this
new function `maybe_get_user_by_id`.
Probably about half of these callers are implicitly iterating through
all users in the people.js data set and thus by construction should
never fail, but it's simpler to just convert all existing callers than
do that audit.
This commit adds code to include can_mention_group_id field to
UserGroup objects passed with response of various endpoints
including "/register" endpoint and also in the group object
send with user group creation event.
Fixes a part of #25927.
This is logic from 10 years ago (dbc4798594)
that is no longer relevant. It seems like we used to show the
search bar open all the time and only showed the buttons when
there was focus in the search bar. Now we close the search bar
when it's not being used, and no longer need to update button
visibility or disable the search close button.
In commit 92ad988144, we removed the local echo code that tried
to re-render locally echoed messages that were rendered right before
a server/browser restart. These changes removed the only event
triggered by `"home_view_loaded.zulip"`.
This removes the remaining `"home_view_loaded.zulip"` event trigger
in web/src/server_events.js since it no longer triggers any events.
This commit introduces a seperate `page_params` file for billing module
since we only use certain page_params in billing module only that it does
not make sense to include them in the main `page_params.ts`.
Also introduced `zpage_billing_params` for proper testing new `page_params`
in billing module.
The `initialize` function registers a click handler, and
previously it was being registered another time each time
one of these calls to `initialize()` happened. The only
other thing that happens in `initialize` is a call to
`render_title_area`, so this commit replaces the extra
calls to initialize with that.
Refactored `recipient` family of variables to better names like
`recipient_ids` which also aligns with the type of these variables.
Also, refactored `typing_status.test.js` to use array of user ids
instead of string names like `alice` and `bob` to stay consistent with
the actual type of these parameters.
fields set.
This commit refactors the `state` object to either be null or has all
the fields set at the same time because they are note really decoupled.
This helps us to clean our code a little bit as well as makes it easier
to migrate this module to TypeScript.
This commit cuts the dependency on `compose.js` module for
`compose_actions.js` module by introducing a hook system for
registering different hooks from external modules.
This commit adds a "Followed topics" row to the 'Notification Triggers'
table in the 'Personal settings > Notifications' panel and the
'Organization settings > Default user settings' panel.
This adds support to control email, push, wildcard mention,
visual desktop, and audible desktop notifications for messages
sent to followed topics by toggling corresponding global
notification settings.
The "Followed topics" row is available in the development
environment only.
Previously, we have duplicate stylings in compose banner and unread banner.
This is rather messy and creates a lot of styling rules. We should define
an abstraction for them. This change will rename compose_banner_close_button
to main-view-banner-close-button.
This in-progress feature was started in 2018 and hasn't
been worked on much since. It's already in a broken state,
which makes it hard to iterate on the existing search bar
since it's hard to know how those changes will affect search
pills.
We do still want to add search pills eventually, and when
we work on that, we can refer to this diff to readd the
changes back.
Previously, if the drafts modal was open in two different tabs and
if the user deletes one or more drafts in one tab, then the count of
drafts wouldn't get updated on the other tab.
The right long-term fix likely involves moving drafts from local storage to
the cloud. But we can greatly reduce the impact of the bug by updating the count
whenever the overlay is opened or closed.
Setting empty hash `#` scrolls user to the top of message feed if
done via `window.location.hash` or using browser back / forward button.
To avoid this, we set don't set `hash` after org URL for default view
when user uses `escape` key or clicks on org logo.
In other situations, we explicitly set the hash of the view.
This commit addresses the issue of relying on `compose_state` for
retrieving the `stream_id` to display warning banners. Previously,
warnings were shown for syntax in the message edit box based on
whether that syntax would trigger a warning for the draft content (if
any) currently in the compose box.
We fix this by using a new `get_stream_id_for_textarea` function to
obtain the correct `stream_id` value for the check being done.
Fixes: #25410.
When there are only muted unread mentions in a stream, show `@` icon
and unread count in faded style, also align the `@` on more topics
with no unead counter on it.
If there are only muted unread messages without mentions don't show
the unread counter on the stream.
Fixes#25382.
We need to append the `version` parameter when constructing the urls for
medium-sized images so that the browser updates the image in real time when
the user uploads a new avatar.
Fixes#25558.
This commit updates the Help Center links in all relevant empty message
list views to open in a new tab by default. This prevents users from being
navigated away from the app.
Fixes#25337.
This commit moves `maybe_get_stream_name` function from `stream_data` to `sub_store`
as it didn't had any dependency on `stream_data` and it also helps us to cut off
dependency on `stream_data` for some of the modules including `user_topics`.
This removes the previous "or forward" text from the message
actions popover, and keeps the documentation in sync with the
new text. Internationalization tests are updated, too.
Fixes#25603.
Prior this commit, changing the message type from a stream (where posting
was not allowed) to a direct message using the compose box dropdown, did not
changed the state of the send button from disabled to enabled even though
direct messages were allowed in the organization.
This was happening because `check_stream_posting_policy_for_compose_box` was
only for streams.
Now, function is updated to check for both streams and direct
messages, as it checks if direct messages are allowed or not, and depending on
that, it updates the send button's state, tooltip and displays a relevant banner.
This likely needs further refactoring to switch to using stream IDs
rather than names in this code path, but this change fixes an
exception that would be throw when opening the compose box while
viewing a narrow to an invalid stream name/ID.
Keyboard navigation has been added to the scheduled messages modal
in this commit. The solution is based on the `modals_handle_events`
function from the `messages_overlay_ui.js` module, as well as some
helpful functions from the same module.
Fixes: #25181.
This is a preparatory commit to implement keyboard navigation in the
message scheduling modal. The main goal is to make the
`modals_handle_events` function reusable. To achieve this, we have
extracted all the context-related variables and replaced all
mentions of "draft" with the more neutral term "item". The
`modals_handle_events` function now also has a context parameter,
which contains all the necessary methods and properties to work in
different modal contexts.
This changes to not report any `call POST /json/users/me/presence`
spans, which we previously reported despite not including the inner
auto-instrumented HTTP spans.
Previously, when a user enters a empty dm-including view, they'll notice
the "Why not start the conversation" action line and click on the link.
When this happens, the compose box would open but the receipent box is
never populated.
Since the dm-including view is a search view, we should drop that phrase
from dm-including views altogether. It also isn't super natural to
have a button that starts the conversation with the user anyways.
Fixes: #25524.
`update_or_append_banner` abstracts the logic for not creating another compose
banner if one is already present in the DOM, it just replaces the content of the
old banner with the new one.
This commit introduces a new container parameter for functions that
can be used for both compose and edit mode. It provides the function
with information about the context in which it is being used.
This introduces a function that checks for both the existence and the
expiration of the `selected_send_later_timestamp`.
The logic it supports prevents users from scheduling a message to send
in the past or less than five minutes into the future at the level of
the UI (specifically the popover on the \vdots component of the Send
button). That can happen if a user attempts to edit a previously
scheduled message.
Fixes#25439.
In the "mentions" test, an additional unread message with the type
"private" and directly mentioning me has been added. This test case
checks for the scenario when the stream_id is null during the
reverse_lookup, which would have caused the test to fail before the
bug fix was implemented which now passes after the fix is applied.
This commit ensures that user-locale and 24-hour preferences are
respected in the message-scheduling modal. It also simplifies the
translation of text strings in the scheduling modal.
Available scheduling options and their time values, including whether
the options are allowed, are now calculated every time a user opens
the scheduling modal.
In order to achieve those things, additional interrelated fixes here
accomplish the following:
1. Modal-scheduling opts now have data- attributes containing
timestamps for the time a message will be scheduled to send.
2. With those timestamps in place, the logic for setting the
scheduled send-time is simplified.
3. There are no more `send_later_xxx` global variables in the
`schedule_send` module.
Fixes#25403.
This commit extracts date-based logic from the popover menu file and
puts it in with the scheduled-messages logic.
The aim is for greater testability, with some initial tests now
presented on the date-based logic.
This function will allow us to adjust the codebase to write what it
means semantically -- whether a check is for the message list being
visibly empty, or completely empty.
In this commit, we leave the .empty() method incorrect, because
several other adjustments need to be made atomically with fixing it.
Apparently some past refactor caused the animation for new mentions to
be triggered when initialization the app.
(This seems pretty clearly unintentional: A user loading the app
doesn't need their attention specifically drawn to the @-mentions view
in the same way that a user who is using the app and receives a
mention right now does.)
Fixes#24641
When the user clicks on a link which has `stopPropagation`
and doesn't trigger `scroll`, then we don't hide any existing
popovers if the element being clicked doesn't hide popovers
explicitly.
To fix this, we hide all popovers on change in hash which makes sense
on its own given how we use hashes.
This commit introduces structures and logic to hide the Scheduled
messages item from the left sidebar if there are no messages
scheduled to be sent.
Test coverage has been added for counts and visibility, too.
Fixes: #25101
This commit introduces logic to present a message count with the
Scheduled messages item in the left sidebar.
The count is present on the initial load, and is updated as a user
adds or removes scheduled messages.
Previously, the unread banner templates just rendered on the contents of
the unread banner. This works fine if we don't want to make changes towards
the parent/container of the contents.
This change introduces a new container to each unread banner templates and
a rename. Thus, we can make unique styling changes to the unread banners
while also bring the structure closer to how it is for compose banners.
This commit deletes `/fixed-width` and `/fluid-width` slash commands
from the typeahead and also hides the slash commands `/light` and
`/dark` in production.
Fixes#25374.
When all the unread messages in a muted stream are in specifically
muted topics, this ensures that the total unread count for the stream
that the user sees before clicking "more topics" will match the total
unreads number for the stream itself.
This behavior is limited to muted streams, since in a normal / not
muted stream, we don't display a "muted topics only" faded unread
count by the stream's summary line to avoid distracting the user with
it, we match that behavior for the "more topics" line.
We also now display the `@` , again to ensure the stream's summary
line never displays an `@` without some topic row having one.
Earlier, the `s` hotkey just narrowed to the stream of the selected
message (to a topic), while `shift+s` narrowed to the conversation view
(topic / dm) of the selected message.
Now, the `shift+s` hotkey has been removed (but retained for toggling
subscription to a stream when the stream overlay is active), and the `s`
hotkey takes on double duty: if the current view is not topic / dm, it
narrows to that, else when in topic view, it switches to stream view. It
has no effect when in dm view. Documentation has been updated for this
both in the help center, and the in-app `Keyboard shortcuts` menu.
A deprecation notice has been added for `shift+s` as well.
Fixes: #24226.
As discussed in the comment, this resulted in the many inactive users
in chat.zulip.org with "Tim" as their full display name from being
incorrectly prioritized over me in a stream where I've been active;
and this class of problem seems like it will be common in large open
communities.
This reverts a portion of 4dc1b2f812,
with explanatory comments for why this behavior is preferred.
The stream name changes in the original PR remain very helpful.
Separating these concepts allows us to provide a much nicer format for
contexts where ultra-specific clarification is not a priority.
This new variant is currently only used in the scheduled messages UI.
Previously this dropdown was only for selecting streams, but
soon it will also be for switching to a private message. This
name helps it be clearer that the dropdown is more general
purpose.
We refactor the triage function to optionally take in a comparator
function, and use this to sort the results, except any exact match,
which is placed highest. Now we don't need to sort the results of triage
for streams, languages and slash commands since we just pass in the
comparator function. The overall effect is same as before, except that
exact matches are always shown first.
For users, we can't use the new triage feature to achieve this goal
without sorting `rest` and breaking a key optimization, so we just add
a bit of manual code for the job.
Fixes: #25123.