This fixes a regression introduced in
9954db4b59, where the realm's default
language would be ignored for users created via API/LDAP/SAML,
resulting in all such users having English as their default language.
The API/LDAP/SAML account creation code paths don't have a request,
and thus cannot pull default language from the user's browser.
We have the `realm.default_language` field intended for this use case,
but it was not being passed through the system.
Rather than pass `realm.default_language` through from each caller, we
make the low-level user creation code set this field, as that seems
more robust to the creation of future callers.
Making request a mandatory kwarg avoids confusion about the meaning of
parameters, especially with `request` acquiring the ability to be None
in the upcoming next commit.
Letting the clock tick without a reason introduces the
possibility of nondeterministic test failures depending on the execution
time. The default value when not specified is tick=True, which makes it
easy to miss.
The rule doesn't prohibit setting tick=True, as perhaps there will be
tests wanting to use that feature on purpose, but such a test should
explicitly set it to make the intent clear.
None of these tests seem to want to have tick=True, which is the
default. Letting the clock tick without a reason introduces the
possibility of nondeterministic test failures depending on the execution
time.
Earlier, followed topics in muted streams were not listed in
the 'Recent conversations' table when 'Include muted' filter is
unchecked.
They should be listed as we currently list unmuted topics.
The "test_filter_all" includes tests for muted and not muted topics,
but the 'is_spectator' field is set to 'true'.
Logically, it is incorrect because a topic can't have a
visibility policy set in spectator view.
This commit fixes the incorrect behavior and adds a separate test
for the case 'is_spectator=true'.
The issue was introduced in d0a697f.
Refactor tools/lib/provision_inner to conditionally write activation commands to user's bash profile based on the host machine type. Automatic activation now skipped for native linux containers.
Fixes#15029
If you focus on inbox-header name by right arrow then down and then
left, it will focus on unread count instead of ellipsis icon.
Hard to explain how this fixes the bug here, next commit adds
explanatory comments to the function.
This reverts b8581e2895. The mobile
client on Android parses this field using:
```kotlin
timeMs = data.require("time").parseLong("time") * 1000
```
This throws an error if value is not `long` (i.e. an integer),
resulting in dropped notifications on Android from servers which had
deployed b8581e2895.
Switch back to sending an integer, but keep the behaviour from
fd6091ad17 where we send the timestamp in the payload of both
Android and Apple push notifications.
Rather than fetch all UserMessage rows for all streams, and subtract
those out in Python-space from the list of all Message rows the user
may have received -- do this via a "NOT EXISTS" subquery. This is
much better indexed (performing in fractions of milliseconds rather
than hundreds), and also consumes much less memory.
Adds support for bulk-adjusting a single user's membership in multiple
user groups in a single transaction in the low-level actions
functions, for future use by work on #9957.
This commit resolves an issue where the edit button was not visible
during message hover for new messages. The problem occurred when a
message was sent after the hover event, resulting in the hovered
message being a local copy with the `locally_echoed` attribute set
to true. This caused the `is_message_editable_ignoring_permissions`
function to return false, subsequently causing `is_content_editable`
to return false as well. As a result, the `is_content_editable`
argument was false, leading to the edit button not being displayed.
The issue was caused due `$current_message_hover` being set to the
locally echoed version of `$message_row`. When the message was
successfully sent and the message was rerendered, the `message_hover`
function was called again on the properly sent message row.
However, the conditional check `($current_message_hover &&
rows.id($current_message_hover) === id)` returned true because
`$current_message_hover`, set to the locally echoed version, had the
same ID as the actual `$message_row`. Consequently, this condition
returned true at the start of the `message_hover` function,
preventing further execution of message_hover function and,
consequently, the updating of hover buttons.
To address this, an `else if` condition was introduced to check if
the message is locally echoed. If true, the function is terminated
to prevent `$current_message_hover` from being set to the locally
echoed version of `$message_row` until the message is sent correctly.
Once the message is sent correctly, the `message_hover` function is
executed again, setting `$current_message_row` to the properly sent
version of `$message_row`, and resulting in updating of hover buttons.
Fixes#25312
This fixes the explanation of the setting's syntax to be more precise
(which doesn't mean "easily understandable" - because the setting is
a bit tricky) as well as an example to illustrate it.
This ensures uniform display and flexing of different bits of
banner content, including:
* Banners like Sent! Scroll down to view...) that have only a
classless `<p>` marking their content
* Banners like the unscheduled message allert, which puts a
.banner_content class right on the paragraph
* Banners like the Your message was sent to a stream you have
muted", which tuck a `<p>` into a div with the .banner_content
class
Uptil now, a dash `-` was used to separate the task from its (optional)
description. But a colon `:` was used in the `/todo` command syntax to
indicate where the task ends and it's description starts.
To make the separator symbol consistent for both the syntax and widget,
`:` is used for both. It was chosen over `-` as `:` is encountered less
often in normal language, and so accidental usage of it when using the
`/todo` command syntax is less likely. It is also is impossible to
confuse with the bullet list syntax which is allowed for tasks.
Earlier the tasks in a list were sorted alphabetically and on marking
one complete, it was pushed under any incomplete tasks. This behaviour
can be unexpected and confusing, so now each task is appended to the
bottom of the list on being added, and no shifting takes place on
marking it completed.
Fixes part of #20213.
The bot settings avatar row was using `fa-clipboard` icon. Changed
the icon to use `{{> copy_to_clipboard_svg }}` SVG icon that is used
everywhere else in the app as copy icon.
Abbreviated long URL for custom field in the user profile modal
and user card popover. Added a copy-to-clipboard icon at
the end of the URL. When hovering over the copy button, it displays
a tippy tooltip with the text `Copy URL`. Clicking the copy button
displays a `Copied` tippy tooltip for 1 second.
Fixes#21680
This is a prep commit. This commit renames
`custom_profile_fields_link` to `custom-profile-fields-link`
to follow kebab-case, which is the preferred style in Zulip.
When the user is typing a numbered / bulleted list in the compose box,
on pressing Enter the list syntax will be inserted automatically on
a new line, and removed if Enter is pressed after an empty line.
Fixes: #20305.
A formatting button below the compose box can format the selected text
by numbering or unnumbering the selected (partially or completely)
lines. The behaviour is inspired by GitHub's bulleted list formatting.
Fixes part of: #20305.
A formatting button below the compose box can format the selected text
by bulleting or unbulleting the selected (partially or completely)
lines. The behaviour is inspired by how GitHub's bullet list format
button works.
Also adds a new breakpoint for showing vdots for formatting buttons.
Fixes part of: #20305.
Semgrep 0.118.0 changed the default of --scan-unknown-extensions to
false. It also seems that it no longer respects --lang (or never
did), so rename the config file to reflect that it only includes
Python rules, to make it clear that additional languages will require
separate config files.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
The internal design here was that we had no-select on the outer
element and auto-select on these inner elements to override it.
But, `auto-select` class seems to have no effect so we are
removing it.
Tim Abbott
Anders Kaseorg
Mateusz Mandera
Alya Abbott
Alex Vandiver
Prakhar Pratyush
Sudip Bose
Aman Agrawal
Hardik Dharmani
Karl Stolley
N-Shar-ma