To avoid event handlers being lost when we restore a message edit
form we go through the process of creating the edit form for the
message again. This has the additional benefit of updating the
status of buttons based on org settings like GIF or upload button.
Fixes#30083.
We have used the dark theme color for the left sidebar row hover:
`--color-background-active-narrow-filter`. We are not using that variable
directly since we don't want a change in that color affecting the color in
recent conversations.
We've also used a single color on hover compared to 2 different colors
for unread and read row before. Light mode also uses a single color on
hover.
Relevant CZO conversation:
https://chat.zulip.org/#narrow/stream/101-design/topic/Hover.20color.20in.20Recent.20conversations.20dark.20mode.2E/near/1797727
Adds description in views styled like stream descriptions also adds a
help center link to the appropriate page at the end of each
description.
Fixes#29769.
Change type annotations of get_user_ids and create_item_from_group_name
to accept CombinedPillContainer and InputPillItem<CombinedPillItem>
respectively.
Change type annotations of get_user_ids and create_item_from_stream_name
to accept CombinedPillContainer and InputPillItem<CombinedPillItem>
respectively.
Earlier, the compose textarea and the preview (container) occupied the
same grid area, 1 at a time. Now, they are both children of a common div
which is assigned the same grid area. This allows adding more elements
to the same grid area in the future.
This is purely a code change and should have no visible effects.
This is a prep commit for relocating the compose expand / collapse
buttons to the top right inside corner of the textarea / preview.
Reproducer:
* Go to a stream narrow.
* Scroll up to select a previous message.
* Click on a different stream in the left sidebar.
* Click browser back button, you land at first unread message instead
of your selected message.
Fixed by updating the hash before we render the new message list.
With combined feed no longer cached, this test became flaky as
it took longer for combined feed to load.
Verified that other tests waiting on `.message-list` don't need
this change.
When the request to mark a thread as unread fails, we should use the
value of include_anchor that was sent in the request, not false.
Modified handle_mark_unread_from_here_error to take an object instead of
parameters.
Since there are no rows to set focus in recent view before initial
fetch, we were falling back to set focus on input filter after
rendering recent view.
Waiting for initial fetch to complete helps us set focus on the
table easily while recent view is also open for user to play with
if the load takes longer.
The `margin-bottom` was removed for the last element in the preview in
e55f5a1b59 to remove vertical shifts when
toggling preview mode, but it is not needed for image / video previews,
so now `margin-bottom` is not set to 0 for the last inline preview.
In 61b3c698af we stopped restoring widgets
from the cache, so the only purpose of the cache remaining was to access
the event handler patched onto the widget jquery.
Now instead of the entire jquery object, we cache just the event handler
since that's the only thing we need from the cache, and do not patch it
onto the widget jquery object.
Use the "/json/messages/flags" instead of "/json/messages/flags/narrow"
when we have fully fetched the current view.
Update messages on the basis of id instead of updating the entire
narrow.
Fixes#28945.
Since the reference for these typeaheads is has a fixed position
unlike say message controls which move on scroll and need to be
repositioned, we can use `fixed` position for typeaheads too so
there is no jumps when user is trying to scroll when a typeahead
is active.
This basically replaces the commit where where we remove `fixed`
as an argument to `Typeahead` library. We can add it later if
any of our typeaheads don't need to be fixed positioned.
When compose box is expanded, typeahead overlfows the top of the window.
We move the typeahead to the bottom of the screen and let
preventOverflow shift it into the visible area.
To better make use of server load when fetching message history for
users, we limit how much history we fetch in the initial loading
experience based on how much time in the organization's message
history it corresponds to.
Co-authored-by: Tim Abbott <tabbott@zulip.com>
For more discoverability, the keyboard shortcuts for preview, and bold,
italic and link formatting are now displayed in the tooltips of those
compose buttons.
Earlier we did not activate widgets if their rendered HTML was cached,
and also when narrowing to the combined feed view. This caused bugs with
widgets not being activated, and so not responding to any interactions.
Now we unconditionally render and activate widgets in the current view,
irrespective of the cached HTML or the view being narrowed to.
`source` is expecting a list so `false` doesn't make sense here.
It was working before because the code that calls `source` in
`bootstrap_typeahead` just does falsey checks.
This commit replaces the fontawesome "edit" and "copy-link" icon found
in the user profile modal with the corresponding zulip icons.
It also moves the functional classes like
user-profile-manage-own-copy-link-button from the icon element to its
container, including adding a new div wrapping the icon in
user-profile-manage-others-edit-button.
Attempting to convert .data("draft-id") to .attr mysteriously breaks
draft saving, so I’ll leave that for future work.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
e.target is the wrong element, and even if this were to successfully
add the _full-height-no-scroll class to <body>, that’s not a child of
.portico-hello-page.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
Its selector looks like it’s missing a ‘.’, but “fixing” it breaks the
page even more, and trying to decipher this mess of event handlers
that poorly replicate what should be native behavior is making my
brain hurt.
Signed-off-by: Anders Kaseorg <anders@zulip.com>
Combined feed always narrowed to the first unread id regardless
of if it was muted or not since we were not actually using the msg_id
that we found after all the hard work of doing all the checks for it.
Originally, we only wanted to do local echo in the current
view. However, now that we're looking at navigating the user to the
conversation that they send a new message to, it's going to be quite
common that we immediately visit a destination different from the
current view, where local echo in that different view would be
valuable.
The most interesting block was added in
af188205cb / #8989. But in
6637f2dbb7, the key logic for checking
`msg_list.data.fetch_status.has_found_newest` was duplicated in the
`add_new_messages` code path, which critically also updates
`update_expected_max_message_id` and thus may close a race with
fetching message history for a view we're being navigated to, where
the locally echoed message might fail to appear at all.
This change does come with a slight regression: If we are looking at a
search view where the filter is one that we cannot apply locally, a
newly sent message will now be locally echoed (returning the compose
box for drafting another) even though it cannot be displayed in the
current view, which means that the message will not appear in either
the compose box or the current view for the brief period before we get
a reply from the server in this scenario. This is a minor detail,
likely not worth troubling ourselves over.
Given our intent to experiment with navigating the user out of the
search view in this scenario, this is likely not important.
Co-authored-by: Tim Abbott <tabbott@zulip.com>
The `at_rendered_bottom` check was incorrect since it includes
bottom whitespace in the calculation to check if we are at the rendered
bottom. This is can cause us to not fetch new messages if user stops
scrolling when bottom whitespace is partially visible leading to
a bad UX.
If user goes to combined feed view on escape keypress, then
we restore their narrow pointer and offset by restoring it from
browser history if the their last narrow was the combined feed view.
This is no longer required since narrow state is restored by
`reload` library and narrow history restores the narrow pointer
and offset when going back to a view.
Only regression here is we will no longer restore the pointer when
user navigates to the combined feed view without using browser back
button and combined feed view is the default view. This is fixed
in the next commit.
We now use "settings-field-label" and "modal-field-label"
class on label elements in settings and modals respectively
to add the bottom margin which was previously done by
"dropdown-title" class as it is better to have more
generic names to use on label for all type of inputs.
This commit adds bottom margin to label elements of
different inputs by adding "modal-field-label" class
to the label elements. Some of them already have the
margin due to having dropdown-title class.
This commit adds bottom margin to label elements of
settings inputs in personal, organization, stream
and group settings using the recently added
"settings-field-label" class.
Most of the settings are dropdown settings, so the
label for them already had a margin as they use
"dropdown-title" class.
There is no need to have hr element below the "Field choices"
heading in custom profile edit form and this also makes it
consistent with the new custom profile field form.
This commit removes classes like "title", "input-label"
and "emoji-theme" from label elements in settings and
modals as these are classes are neither used in JS
nor to apply any CSS.
This only every did a single extra fetch, and now that this fetch is
in the same direction as our main fetching sequence, no longer is
meaningfully different from just specifying a different value of
consts.maximum_initial_backfill_size.
Previously, when a message is edited, it is locally echoed with its
pre-edit content.
This is because previously, when we tried to render the edited
message of the edit box during local echo, in order to update
the content, flags, and is_me_message properties of the message
object with that of those returned is markdown.render(), we used
the spread operator and created a new message object, and updated
the existing message object with this new one.
This was misconverted, since edit_locally() method already has a
fully-rendered message object to start with, and is just doing a
rerendering, it should be mutating what message it received, rather
than constructing a new local variable.
Anders Kaseorg
Aman Agrawal
Sahil Batra
Shubham Padia
Tim Abbott
nimishmedatwal
afeefuddin
evykassirer
N-Shar-ma
Temidayo32
Prakhar Pratyush
Kenneth Rodrigues
Prakhar Pratyush
Karl Stolley
Pratik
Lauryn Menard
Alex Vandiver
roanster007