This is sort of a temporary fix to bring the state back to how it
was in commit: ef4337edcb. However,
long-term we will need to fix our local echo feature to do merging
of names just like we do on backend.
We don't need to get sorted streams in the "source"
function for typeahead, since we sort them later,
and we don't need to recalculate values.
This preserves the behavior that we include
unsubscribed streams in the typeahead, which is
probably intentional.
This pulls the essential bucketing/sorting logic out
of filter_table().
The diff isn't quite as clean as I'd like, but some
of the code that got added back to filter_table() can be
eliminated in the future. Basically, all the stuff
related to hidden ids can just be zapped if we go
to an approach of just re-building the DOM cleanly
whenever our filters change.
We replace two calls to stream_matches_query() with
a single call to triage_stream(), which prevents us
from doing the same is-subscribed checks twice.
We probably should have done this a while ago, even
though these functions are pretty tiny. The goal here
is to make it easier to have more consistent search
semantics.
Our first use case is subs.js. In this case we
are able to decouple a bit of generic string
matching from the subs-specific code.
We move some data code from subs.js to stream_data.js.
It's not clear we have been using the optimal sort for
dealing with locales, but this change preserves the
current behavior. The only subtle change here is that
we look up subs using a Dict now instead of a plain
JS object.
The values of this dictionary used to be raw DOM elements,
but get_row() wraps them again, so there's not a huge
reason to store them as raw DOM elements internally. It
is slightly easier to reason about the code if everything
stays at the jQuery level.
To preserve the old behavior here, we have to do something
that is kind of ugly, but at least it's explicit now. In
the old code, our cache was DOM elements, and if an id
wasn't in the cache, we would sneakily return $(undefined)
with this code in get_row():
return $(this._rows[id]);
And it turns out that $(undefined) is basically just a
zero-element jQuery object. A lot of our code depends
on this behavior and just works around the zero-element
objects as needed with checks like this:
if (this.selected_row()).length === 0) {
// don't try to get offset
}
For now we just preserve this behavior. We could eventually
be more strict here, or at least have aggressive warnings
on cache misses, but we'd need to retrofit code to be
able to call something like `has_rendered_selection()`
and/or deal with `undefined` as the return value for the case
where the selection hasn't been rendered.
Here is some example code that would cause tracebacks if
we just returned `undefined` for cache misses:
rerender_preserving_scrolltop: function () {
// old_offset is the number of pixels between the top of the
// viewable window and the selected message
var old_offset;
var selected_row = this.selected_row();
var selected_in_view = selected_row.length > 0;
if (selected_in_view) {
old_offset = selected_row.offset().top;
}
return this.rerender_with_target_scrolltop(selected_row,
old_offset);
},
This function is more cohesive and always takes in
a jQuery object containing exactly one DOM element,
and it does all stuff at the jQuery level of
abstraction (no raw DOM).
It's a pretty simple extraction--removing the level
of indentation makes the diff a bit noisy.
We shorten the name of the function and avoid having
all the callers call `.get()`. Now we mostly stay
in jQuery "space", which avoids some confusion about
when we're dealing with raw DOM elements and which
will facilitate unit testing.
Currently on zoom out from stream topics, scrollbar didn't scroll back
to opened stream. Because call to scroll-to-stream func isn't called
after all streams view is displayed. So wrong stream element is
passed to func.
Fix this by calling scroll-to-stream func after all-stream-list view
is displayed.
We use these new functions in the message compose typeahead so that they
can also be used in a PM recipients typeahead with both people and user
groups.
We now render the "skin" part of "Stream Settings" before
adding in the actual streams. The new function
populate_stream_settings_left_panel() takes care of adding
the streams. It uses a new template called
`subscriptions.handlebars`.
Splitting out this function will give us more flexibility
for various improvements.
First, we can decide to render the list after we open the
overlay, just to avoid the problem that users don't know why
the modal's opening. (And we could add a loader spinner as
needed.)
Second, we can improve our filter features so that we do
filtering in the data instead of moving DOM rows around,
which is expensive.
Third, we can eventually introduce progressive rendering.
Finally, having the function broken out will make profiling
more precise about where bottlenecks exist.
We were passing this in before, but having it as
a data member reinforces the idea that we'll want
this to be a first-class concept in the list, since
we depend on ordering for various things.
We now keep track of keys in buddy_list.js, so that
when we insert/remove items, we no longer need to
traverse all the DOM. Instead, we just find out
which position in the list we need to insert the
key in (where "key" is "user_id") and then find
the relevant DOM node directly and insert the new
HTML before that node. (And of course we still
account for the "append" case.)
There's a little more bookkeeping to make this
happen, but it should help reduce some code in
upcoming commits and pave the way toward
progressive rendering optimizations.
This commit should produce a minor speedup
for activity-related events that go through
buddy_list.insert_or_move(), since we are
not traversing the DOM to find insertion points
any more.
This will be useful for lazy rendering, where our
buddy_list widget already knows the keys (aka "userids")
it wants to render as you start scrolling them into
view.
Typing "tim " did not did not produce any match when suggesting person
in composebox typeahead or user group typeahead as the space at the
end of the "tim " string passed by the browser was a
`no break-space (U+00A0)` instead of `space (U+0020)`.
Although there are unicode characters other than `no break-space` which
represent spaces, only U+00A0 is replaced as it was the only space
character encountered when testing this issue manually.
Fixes#10039.
Currently, if you access an article link with an anchor link that isn't
featured in the sidebar, the main article won't be highlighted. Thus, we
exclude the anchor link hash from the article-searching selector if
the full article pathname wasn't found.
To reduce code duplication when creating hotkey deprecation notices,
create the `get_hotkey_deprecation_notice` function. Also, create a
`ui` testing file with a test for the new function.
Fix#10004.
This replaces some old code with calls to topic_data.js.
Now our topic typeahead uses the same data as our
sidebar, stream suggestions, and the "n" key, so any
future improvements to that data will benefit all
features the same.
This is an important piece of #9857.
Now that `emoji_collection` and `emojis_by_name` are global
datasources in the webapp we need to rename things carefully
to reflect their actual meaning. The fact that emoji code is
used as a css class for unicode emoji is one thing but it is
not its sole use so renaming it seems a good idea.
This commit moves the `emoji_collection` datasource in the emoji
picker to emoji.js and renames it to `emojis_by_name`. It is a
mapping from emoji name to object where each object describes an
emoji. This is an effort in the direction of de-duplicating and
unifying the datasets being used by various our widgets(like
emoji picker and composebox typeahead) in the webapp. Migrating
all the widgets to a single datasource will help us in removing
the whole class of annoying bugs which causes some emojis to be
missing from some widgets.
This commit closes a long pending issue which involved moving the
`EMOTICON_CONVERSION` mapping to build_emoji infrastructure so
that there is only one source of truth. This was pending from the
time when this feature was implemented.
Pressing `Esc` did not blur a contenteditable div by default, while
an input field was blurred by default. Due to this when a user tried
to unnarrow using `Esc` key when the searchbox had focus, the focus
remained stuck in the div itself and no further action was taken.
If search pills are not enabled, the text present in the search bar
will be selected on pressing '/' and writing someting without deselecting
the text will clear the search text. Since selecting the pills would
not make sense in this context, the search box is focused instead.
Adds box-shadow to `#searchbox` when either `#search_query` or any
of the pills have focus. Uses jquery instead of pure css as the
`:focus` event occurs on `#search_query`, while we want to add
box-shadow to `#searchbox`. This could have been done with
`:focus-within` CSS selector, but it is not supported in IE or Opera.
`#search_query` already had an onfocus/focusout listener, adding
listeners to `#searchbox.pills` for those events wouldn't have worked
as you don't want the focusout event to fire when the focus shifts
from input to pill.
Also adds `focusin`, `focusout` and `css()` to zjquery. `css` is
same as `val`, except it returns an empty object in case of no value
instead of an empty string. I don't think `css()` is valid syntax
in actual jquery.
After adding search pills, suggestions were based only on the
current input and no validation against the existing pills was done.
operator_subset_suggestions have been removed. Default suggestions
for base_operators have also been removed.
Handle multiple operators:
if `is:starred stream:Ver` was typed without selecting the typeahead
or pressing enter in between i.e search pill for is:starred has not yet
been added, then the description of `is:starred` will act as a prefix
in every suggestion.
Also makes changes re-enabling person suggestions for names with spaces.
This large function will need to be modified significantly as part of
the pills effort, and copying it lets us preserve behavior in
production until we're ready to cut things over.
tab_bar.js becomes redundant after implementation of search pills.
This commit adds a comment to tab_bar.initiliaze, so the event
listeners related to it do not get initiated. This does not remove
any code related to tab_bar.js.
Also adds left and right border around the search icon.
Following points have been implemented in this commit:
1.) Add search pill on selecting typeahead.
2.) Re-narrow after removing a search pill.
3.) Add quiet optional parameter to removeLastPill.
4.) Pre populate search pills in narrow.activate.
5.) Clear existing search pills on narrow.deactivate.
Description of above points:
1.) I tried out using the description from suggestions.lookup_table
to append a pill using appendValidatedData so that the description
had not to be calculated again. But the description in the suggestions
lookup contains html due to highlighting. This html is escaped when
inputed in a pill. An attempt was also made to remove the higlighting
by replacing the tags. But other espaced characters like < also
popped up, so it was better to use append_search_string.
3.) If one wants to refresh the pill using pill.clear and wants to
repopulate them, evaluating the event_handler associated with the
action of removing the pill may not be desired.
4.) Pill population code is added to narrow.activate. Pills are not
populated if the narrow was triggered by search as search handles the
addition and removal of pill by itself. The reason for not handling
search too in narrow.activate is to avoid clearing the pills and
repopulating them. Example of some of the triggers for narrow.activate
include `restore draft`, `topic change`,`sidebar`.
Also modifies tests for search.js
Adds an optional parameter `quiet` to removeLastPill and removeAllPills.
If `quiet` is a truthy value, the event handler associated with the
pill will not be evaluated. This is useful when using clear to reset
the pills.
Also adds the file to the static asset pipeline.
search_pill_widget.js will be used to access the pills object for
the search query box. It will act in a similar way to
compose_pm_pill.js. Why is this needed: Consider you've initiated
a pills object in search.js for the search query box. Now you want to
also access that pills object to pre-populate pills after a reload in
hashchange.js. search_pill_widget.js makes this easy without the use
of events.
When Pan & Zoom (canvas) is enabled, the `v` hotkey does not work due to
`LightboxCanvas` overriding the `keydown` event. Add `v` as an option in
the new listener.
Fix#9777.
This commit updates the `emoji-datasource` packages to version 4.0.4.
This update brings following changes to emoji infra:
1: Fix for the bleeding sprite sheets.
2: The category of some emojis has been changed. Categorywise breakup of
net gain or loss is as follows:
Travel & Places: 58 (gain)
Symbols: 47 (loss)
Smileys & People: 52 (gain)
Objects: 11 (loss)
Food & Drink: 3 (gain)
Animals and Nature: 46 (gain)
Activities: 9 (loss)
3: There were some changes in the image farm of the package which were
breaking our old emoji farm. I fixed them by modifying the remapped
emoji map.
Fixes: #8235.
This behavior was originally implemented in commit 6993f89, but due to not
specifying a toggle option, the Subscribed/All streams switcher tab was
focused after the input was focused, leading to the input's loss of focus.
Fixes#9981.
The current code for detecting which image to add the `selected` class
to is `preview_source.match(src)`. With the new thumbnails, this no
longer works because thumbnail URLs include a `?`, which has its own
RegEx significance. To solve this, check for equality instead of using
RegExes.
Various pieces of our thumbor-based thumbnailing system were already
merged; this adds the remaining pieces required for it to work:
* a THUMBOR_URL Django setting that controls whether thumbor is
enabled on the Zulip server (and if so, where thumbor is hosted).
* Replaces the overly complicated prototype cryptography logic
* Adds a /thumbnail endpoint (supported both on web and mobile) for
accessing thumbnails in messages, designed to support hosting both
external URLs as well as uploaded files (and applying Zulip's
security model for access to thumbnails of uploaded files).
* Modifies bugdown to, when THUMBOR_URL is set, render images with the
`src` attribute pointing /thumbnail (to provide a small thumbnail
for the image), along with adding a "data-original" attribute that
can be used to access the "original/full" size version of the image.
There are a few things that don't work quite yet:
* The S3 backend support is incomplete and doesn't work yet.
* The error pages for unauthorized access are ugly.
* We might want to rename data-original and /thumbnail?size=original
to use some other name, like "full", that better reflects the fact
that we're potentially not serving the original image URL.
Previous commits have fully implemented the logic for stream email
notifications; this final commit adds support for configuring it to
the UI.
Fixes#6511.
This migrates Zulip to use a dramatically better set of names and
aliases for our emoji set, defined in emoji_names.py (which is in turn
manually generated from our hand-curated CSV file).
This should significantly improve the experience of using Zulip's
emoji picker and emoji typeahead for finding what one is looking for.
When you hover above the navbar, the cursor focuses on the page
body and scrolls the entire page, breaking the positioning of the
fixed sidebar and page content.
We disable scrolling on the body but allow the sidebar and Markdown
content page elements to be scrolled to fix this bug.
Fixes#7665
In case of invitation events, 'invites_changed' event without
any real payload is sent to all the realm admins and the user.
The event is handled by reloading the list to view recent changes.
Commit tweaked by shubhamdhama:
* Send an `invite_changed` event when an user accept an invite.
Also, added the test for the same.
* No need to delete the invite list in frontend, current logic
handles the case when the invite data is changed properly.
* Extracted the common logic for sending an event into
`notify_invites_changed`.
We can now theoretically use this for any textarea
that supports our markdown (besides the compose box),
plus we keep the RTL code a bit more self-contained.
The JavaScript click handler for this feature was fragile in a way
that would break with upcoming changes to how we display the X
element. We clean this up with a replacement implementation that
should be much less fragile.
Fixes#9842.
Enables avatar images in pills wherever user_pill.js is used.
(e.g composebox, user group settings)
Changes to search_pill.js are not made as search pills haven't been
added yet completely and search_pill.js just contains the preparatory
code right now.
No change to compose_pm_pill.js is not required as it uses
`user_pill.create_item_from_text` in its `create` function.
Allow passing image link in the item passed to appendValidatedData.
When passing image link via any of the append* functions, make sure
that create_item_from_text for that pill also adds the image link to
the item created.
This commit does not make any visual change to the current app.
Changes to user_pill.js are necessary to enable user avatars for
pills.
We now use narrow_state.filter() everywhere. The
two functions did the same thing, and I slightly
prefer the concise name, which was already in use
in lots of places.
This implements right-to-left message automatic detection support in
the compose box as well as the message feed. Full unit tests and
support in the message-editing UI are for future work (as are
potentially more fancy things like supporting things like
right-to-left multi-word names for users/streams/etc.).
Fixes#3123.
This commit fixes some modules that were erroneously left out while
transitioning app.js to webpack. This commit exposes them using
expose-loader or setting them directly to window.
This commit moves all files previously under the 'app' bundle in
the Django pipeline to being compiled by webpack under the 'app'
entry point. In the process, it moves assets under the app entry
to a file called app.js that consumes all relevant css and js files.
This commit also edits the webpack config to be able to expose certain
variables for third party libraries that are currently required by
some modules. This is bad coding form and should be refactored to
requiring whatever dependencies a module may have; we're just
deferring that to the future to simplify the series of transitions we
need to do here. The variable exposure is done using expose-loader in
webpack.
The app/index.html template is edited to override the newly introduced
'commonjs' block in the base template. This is done as a temporary
measure so as not to disrupt other pages on the app during the transition.
It also fixes the value of the 'this' context that was being inferred
as window by third party libraries. This is done using imports-loader
in the webpack config. This is also messy and probably isn't how we
want things to work long term.
This commit prepares the frontend code to be consumed by webpack.
It is a hack: In theory, modules should be declaring and importing the
modules they depend on and the globals they expose directly.
However, that requires significant per-module work, which we don't
really want to block moving our toolchain to webpack on.
So we expose the modules by setting window.varName = varName; as
needed in the js files.
This cleans up some leftover js and css from the effort of
redesign the rows of the #subscriptions table. Redesign happened
in commit 368b5859 and but we forgot to clean up these js and css
pieces.
squash to subs.js.
The error handling for delete/reactivate was broken.
The old code related to appending id_suffix to the ids of
the per-bot error divs did not have corresponding
selectors in the actual error handling.
Things still aren't great, but there's a bit more
encapsulation now, and you'll see errors for the
delete/reactivate cases.
The user can also edit the question after adding it.
The question in the poll can only be added/edited
by the user who started the poll.
The input bar will be disabled for the other users
if the question is not yet added. If the question is
added, the input bar will not be visible to the other
users.
Even when admin removes all custom fields from org, custom
fields header "Profile" doesn't get removed.
Render header "Profile" whenever custom fields data get changed.
This commit fixes a regression which was introduced while
we were removing icon-vector and replacing the same with
font-awesome. We forgot to update the toggle icons from the
JS file.
Fixes#9822.
fuzzysearch matched query if the query letters appeared in sequence.
Here we use the extracted phrase_match to match query with the prefixes
of words.
This disables `ctrl + shift + [`, while `ctrl + [` will still trigger
an action.
Also, add a test for ensuring that the `ctrl + shift` combinations fall
through.
This disables `cmd-or-ctrl + shift + k` and `cmd-or-ctrl + shift + s`,
while `cmd-or-ctrl + k` and `cmd-or-ctrl + s` will still trigger
actions.
Also, add tests for ensuring that the `cmd-or-ctrl + shift`
combinations fall through.
Fix#9779.
Adds search_pill.js to the static asset pipeline. The items
for search pill contain 2 keys, display_value and search_string.
Adding all the operator information i.e the operator, operand and
negated fields along with the search_string and description was tried out.
It was dropped because it didn't provide any advantage as one had to
always calculate the search_string and the description from the operator.
The slash in command is stripped in the backend,
rather than in the client to make the client code
cleaner.
This would make client code cleaner in the slash
commands which include parameters.
NOTE: To test this locally I've used Google Chrome input tool.
This change will not affect users who don't use input tools.
Here is the algorithm used to deal with this case and other important
points:
* Here I've used `compositionend` event which is triggered
as soon as an input tool completes a word or user press "enter"
to get the suggested text. (There was a situation where it is
triggered even when input tool wasn't closed, that is when we
press space, but it also triggers another `compositionstart`
event simultaneously so our logic can't be affected by this.)
* We are using a variable `is_using_input_method` which sets to
`true` when `compositionend` event is triggered.
* Basically our searching is initiated by `keyup` event which
is triggered by the same keypress which triggers `compositionend`
event to get the text, so our main goal is to suppress the searching
triggered by this key pressing.
* Observation shows that `compositionend` is triggered before the
`keyup` and calling of callback `narrow_or_search_for_term`
used by typeahead.
i.e. chronological order of triggering of this event is
`compositionend` > calling of `narrow_or_search_for_term` > `keyup`.
* So the main logic is to set `is_using_input_method` to `false`
by default and if used the input tool then when we press enter
to get the suggested text we set it to `true` which indicate
further events triggered after it to skip the searching and
finally in `keyup` we set it to default `false` so when pressed
enter again we have it set to false and we have a successful
search.
Fixes: #9396.
Whenever a link is clicked, the page link changes, and the content
of the `.markdown .content` node updates, preventing the old
listener to catch any future anchor link clicks.
We attach the listener to the document instead and only activate
it when the target element is a proper anchor link heading.
Fixes#9767.
This makes sure that CSRF token is available while initializing
Socket, irrespective of the order of execution of deferred callbacks
after document becomes ready.
This is part of #9416.
Set the initial visibility of the page content to hidden via
the stylesheet, and allow any data fetching and rendering to
complete before making content visible.
Adjust the conditional logic within `render` to first check for
the case in which a user goes to a doc link, the case in which
we asynchronously fetch data prior making content visible.
Fixes#9577.
actions_menu_handle_keyboard now only gets the action menu items
and passes them to the newly added popover_items_handle_keyboard.
popover_items_handle_keyboard takes the key and menu items as its
parameters. The function can be used when handling keyboard input
like user profile popover. Similar refactor has been carried out in
focus_first_action_popover_item. This refactor is a part of adding
the missing support of keyboard navigation to user profile popover.
The function `settings_account.add_custom_profile_fields_to_settings`
called twice, which resulted in two templates objects being
rendered.
The function also didn't check whether settings overlay was open or
not when processing new events, and thus would throw an "undefined"
error if a custom profile field was editing while the overlay was not
open.
Fixes#9668.
Explaining the problem a bit: When we narrow to a stream/private message
using `q+Enter`/`w+Enter` compose box opens which isn't desirable here.
The bug here was the propagation of event after getting handled in
`keydown_util.handle` to `hotkeys.process_enter_key`.
Fixes: #9679.
We should only open the default article heading which is "Guides",
if the user is on the index page i.e. /help/. For non index pages
we don't need to open this heading.
This commit adds a minor improvment in clicking
the left sidebar behaviour. So if you click on
an article heading then other opened headings will
be closed automatically. This makes the toggle
experience better.
This is less than perfect, but for most settings hitting
the enter key will now take you to the first element in
the right panel.
The two exceptions are below. They have checkboxes with
kind of strange markup:
Notifications
Authentication methods
If you toggle between Settings and Organization now, it
will remember where you were the last time (not counting
reload). Likewise if you go in and out of settings.
The old code always put you in the first section, which I
think was an accident of implementation. Of course, we'll
continue to default to the first row if you haven't gone
anywhere else.
This is mostly a code move, but because things are more
modular now, we don't need the two conditionals to find
out what kind of panel menu we're inside of, and our
selectors are less brittle.
The list with the options for normal settings now has
the class normal-settings-list.
The list with the options for org settings now has
the class org-settings-list.
The new markup helps us avoid code like this:
$(".settings-list li:not(.admin)")
We also have funny hacks in our key handlers related
to the old combined-list approach, which we can
eventually eliminate.
My recent refactoring that split out MessageListData
introduced a nasty bug where we were putting muted
messages into the "All Messages" view even though
the underlying list was correctly filtering
them, so the symptoms were two-fold:
- muted messages cluttered up your feed
- replying to the message caused a traceback (since
it wasn't actually in the underlying data
structure)
This has to do with what MessageListData.add_messages()
was passing back to MessageList to orchestrate drawing
in MessageListView.
I think what happened here is I got this working kind
of sloppily but correctly for the non-muting case and
then got in the weeds of some other stuff. Not my
finest moment.
The actual correct code here is simple enough. We
triage top, interior, and bottom, and then the respective
methods that put the data into the data structure
return the filtered lists (i.e. not muted) and put them
into the info structure.
Fixes#9656
We want to avoid doing too much setup for the info overlay widget
during initialization, since we don't really need it, and side
effects like focusing a modal can cause hard-to-detect
glitches for other features.
In our toggler component (the thing that handles tabs in things
like our markdown/search help, settings/org, etc.), we have
a callback mechanism when you switch to the tab. We were
being tricky and only calling it when the tab changed.
It turns out it's better to just always call the callback,
since these things are often in modals that open and close,
and if you open a modal for the second time, you want to do
the callback task for whichever setting you're going to.
There was actually kind of a nasty bug with this, where the
keyboard handling in the keyboard-help modal worked fine the
first time you opened it, but then it didn't work the second
time (if you focused some other element in the interim), and
it was due to not re-setting the focus to the inner modal
because we weren't calling the callback.
Of course, there are pitfalls in calling the same callbacks
twice, but our callbacks should generally be idempotent
for other reasons.
We want the Botserver to not only work with the
botserverrc, but also with a zuliprc of an outgoing
webhook. Because the Botserver uses the outgoing
webhook token for authentication, we need to include
it in the zuliprc for outgoing webhooks.
This is preparation for an upcoming refactoring where we pass a bot
ID, not the email/api_key, into the zuliprc generation functions in
the bots code path.
`format_as_suggestion` formats a list of operators into a
suggestion using the Filter.describe and Filter.unparse methods.
This change aims to increase readability.
In admin UI for creating new choice type of custom field, the behavior
of trash icon for removing choice field is buggy.
When admin clicks on trash icon it disappears, but the row does not
and admin end up being unable to create the field.
Fix this by selecting proper element to find and delete choice row.
These two slash commands now use zcommand to talk to
the server, so we have no Message overhead, and if you're
on a stream, you no longer spam people by accident.
The commands now also give reasonable messages
if you are already in the mode you ask for.
It should be noted that by moving these commands out of
widget.py, they are no longer behind the ALLOW_SUB_MESSAGES
setting guard.
This adds a /ping command that will be useful for users
to see what the round trip to the Zulip server is (including
only a tiny bit of actual server time to basically give a
200).
It also introduce the "/zcommand" endpoint and zcommand.js
module.
We had a significant amount of code for handling what seemed to be 2
cases, but which were really just a single case (if we are trying to
narrow to a specific message ID, and we end up landing on it, restore
the previous offset; with the special case that the previous offset
might be passed in from the previous call).
This cleanup also fixes a very minor bug, where our background
auto-reload (`reload.initiate({immediate: true});` in the JS console)
would incorrectly reset the pointer position to match the a near:
message ID if that was present in the narrow.
This commit fixes a couple regression related to narrowing.
For a long time we've had bugs where we too aggressively
preserve the currrent selection on topic -> stream
re-narrows ("s" key) even when the wider narrow may
have unread messages before the selection.
Also, we recently introduced a bug so that when you used
a link from the "copy link to conversation" (aka a "near"
query), it would advance you to your first unread message
despite the near:999 specifier. (The code would work for
subsequent "near" queries once you had fetched some of
your original messages).
This commit introduces a new data structure called id_info (replacing
the select_strategy data structure) in various functions and uses that
to track all the ids of relevance.
Significantly rewritten by tabbott to handle a few extra corner cases,
and add a ton of comments explaining why it works the way it does.
Fixes#2091.
Fixes#9606.
When admin user create new private stream, widget for changing privacy
of stream doesn't render. Because we render subscription-settings
template partially on subscription-add event, so this case wasn't
handled.
Fixes#9469
Previously weren't registering modal properly, which gets fixed by
using open_modal. It further fixes closing of modal by using escape
and positioning of modal.
Fixes: #9590.
This directly prevents a traceback when submessage events
arrive in the wrong order. This was probably a symptom
of not updating message.submessages for not-yet-widgeted
messages, which was fixed in an earlier commit, but we
want defensive code in case of races or other glitches, and
it's not the end of the world is somebody sees partial
survey results due to some corner case.
The "if" condition that was removed in this commit
is no longer needed, since the called code now
handles the cannot-apply-locally use case. (We
wanted the called functions to be defensive, so
they already were effectively handling the conditions
anyway, and recent commits has them returning
appropriate values and doing the right things.)
This commit makes it so that any query for
which we do a local filter leads to us
examining the full list of unread message
ids in our cache to find a potentially
unread message that passes the filter. This
can often allow us to more immediately
jump to a new narrow with an appropriately
selected message.
Fixes#9319
We want to be able to call get_first_unread_info() even
if we cannot apply a search locally. It was returning
the correct value before, but this change removes a
blueslip warning that will allow our callers to remove
some guard code in a subsequent commit.
We want to update message.submessages for new events, even
though our couple of widgets (poll/tictactoe) that can process
"update" events currently just apply events as "deltas"
to their current data.
This does fix a subtle issue where you may get incoming
events for a message that the client knows about but which
it hasn't yet activated as a widget. Up until now,
we've rarely seen the bug that's fixed here, since it's
usually the case that as soon as we receive a message, we
widgetize it right away.
The user can now specify the value while creating a stream.
An admin can later change it via `Change stream permissions`
modal. Add is_announcement_only to subscription type text.
For some reason in my original version I was sending both
content and data to the client for submessage events,
where data === JSON.parse(content). There's no reason
to not just let the client parse it, since the client
already does it for data that comes on the original
message, and since we might eventually have non-JSON
payloads.
The server still continues to validate that the payload
is JSON, and the client will blueslip if the server
regressses and sends bad JSON for some reason.
With this commit, we change how we deal with translation for strings.
Previously we used to fetch the translations data after loading which
created a lot of unpleasant race bugs.
So we changed this to use the `translation_data` sent in `page_params`
which is available at load time. The previous fetching can be useful if
we want to change the string to the changed language without reloading
the page but since we ask the user to reload the page after changing
the default language so fetching after loading isn't useful for us and
hence we can add resource only once.
Ultimately, we can remove the i18next plugins too. We leave the logic
for clearing local storage, patched to fully clear it.
Fixes: #9087.
Fixes#3380.
The blueslip warning mentioned in #3380 were from paths ending at
people.email_list_to_user_ids_string. Some additional blueslip warnings
were raised after using that function.
Although we can put a validation check somewhere in the call stack of
people.email_list_to_user_ids_string, this function itself is used to
validate the operand by the higher order functions, so it wouldn't make
sense to put a validation check before that. Instead, removing the
blueslip warning altogether was chosen.
people.email_list_to_user_ids_string was replaced by
people.reply_to_to_user_ids_string which is a blueslip-free version
of the same. Other blueslip warnings were removed.
In 1f72647a5a I accidentally
flipped a condition that made the down key "recenter" on
the "normal" case, not in the "is-at-end" case.
This commit undoes that regression, which probably only
affected czo for a weekend, and makes the logic a bit
more clear.
If atleast one of the private_message_recipients is invalid, compose
box will not be opened.
Thanks to Aastha Gupta <aastha.gupta4104@gmail.com> for some preliminary
work on this.
In user profile popover, date type of custom fields values are
not showing in correct format as "date_joined" value.
Fix this using moment.js to render date type of fields
in correct format.
We've had rare and hard-to-track-down glitches with our
old implementation for going up and down in the message
list, which was using jQuery selectors.
We now just use MessageListData under the hood to find
our next id.
The Botserver uses section headers in the flaskbotrc to
determine which bot to run. Silently setting the section
headers to a bot's username is confusing and makes it
harder for Botserver users to figure out how to get the
Botserver to run the bots they want. This commit empties
all flaskbotrc section headers and thus makes the assignment
of bots explicit and mandatory.
Fixes#6515.
New suggestions for `sender:King ha` will respect spaces and the new
suggestion will be `Sent by King Hamlet <email>` instead of `Sent by King,
search for ha`. But if first term of sender operand is a valid user email,
tokens will be seperated by spaces. e.g `sender:hamlet@zulip.com abc`
will show `Sent by King Hamlet <email>, search for abc`.
We had debug code that was reaching into msg_list._items when
it could use msg_list.all_messages() instead.
When we split out MessageListData, using _items started
breaking this code.
This should make it easier for us to iterate on a less-dense Zulip.
We create two classes on body, less_dense_mode and more_dense_mode, so
that it's easy as we refactor to separate the two concepts from things
like colors that are independent.
A "zform" knows how to render data that follows our
schema for widget messages with form elements like
buttons and choices.
This code won't be triggered until a subsequent
server-side commit takes widget_content from
API callers such as the trivial chat bot and
creates submessages for us.
This starts the concept of a schema checker, similar to
zerver/lib/validator.py on the server. We can use this
to validate incoming data. Our server should filter most
of our incoming data, but it's useful to have client-side
checking to defend against things like upgrade
regressions (i.e. what if we change the name of the field
on the server side without updating all client uses).
We should probably have a try/catch in MessageListView itself
too, for post-processing kind of stuff, but we want to make
this new module defensive in its own right.
This is a trial to have the first reply hotspot in the bottom
whitespace (and stick there until "Got it!" is pressed).
Tweaked by tabbott to clean it up a bit. Still needs more work on the
visuals.
We can have this scenario:
- somebody else creates a widget-ready message
- message arrives in storage
- (message is not yet in view, so no message.widget)
- new submessage event arrives
We want to just ignore submessage events in that case.
(There's a more complete fix coming for this scenario, where
we at least update message.submessages for the eventuality
that we do render the message later.)
This commit improves the output that blueslip produces while
showing error stack traces on the front-end. This is done by
using a library called error-stack-parser to format the stack
traces.
This commit also edits the webpack config to use a different
devtool setting since the previous one did not support sourcemaps
within stack traces. It also removes a plugin that was obviated
by this change.
Fixes#9492.
`is` operator uses predefined categories. This commit
displays an invalid operand message if the operand does not fall into
any of these categories and the `is` operator is not at the last.
e.g. `is:abc sender:abc@zulipchat.com` will have `invalid abc operand
for has operator, sent by abc@zulipchat.com` as a prefix for all its
suggestions.
Fixes#9492.
Default suggestion e.g `abc messages` as a suggestion for `is:abc`
is not shown in a new suggestion. But if the is operator is already
present before any other operator, the default message text will be
used. e.g `is:abc sender:abc@zulipchat.com` will have all the suggestions
with the prefix `abc messages, sent by abc@zulipchat.com`.
`get_containing_suggestions` was used to get the operand suggestions
for the `has` operator. `get_special_filter_suggestions` is now used
to get both the operand and operator suggestions for `has`.
Partially fixes#9461.
Negated suggestion for both operand and operators are handle in
get_special_filter_suggestions. A bug is get_operator_suggestions
causing the removal of `-` symbol from the operand was also fixed.
Now that we've moved it into a bulleted set of options inside a modal,
there's no good reason to have separate variables for the corner cases
around who can manage a stream.
Our logic for stream_has_topics never accounted for
us creating essentially "empty" stream buckets that
don't have topics, but we recently added some code
related to unread counts that violated the original
assumptions of the code.
Now we check deeper into the stream bucket to find
actual topics.
This bug manifested in the left sidebar where users
were seeing streams as recently active just because
some muted topics had unread counts, when in fact
the stream was inactive for practical purposes.
Having submessages will become common enough that the
info message here is too spammy, and there are other
ways to observe incoming submessages if you're doing dev
debugging, which this was originally written for.
We could get submessage events for messages that weren't
in our message store if somebody played with a widget
that was on an "old" message for somebody else.
Dropdown element for outgoing interface type was not showing correct
value, cause the way default value was set to dropdown was incorrect
(it should have been setting the selected parameter on the selected
option if it were going to be selected via the template code).
Fixes#9419.
Fixes#9373.
`not_subscribed` warning is not shown for bots on either private or public
streams. Some of the bots have an interface such that they receive the
message mentioning them even if on a private stream where they are not
subscribed.
We use "Everyone" for the button labels already.
Soon we'll support "Everyone" meaning either the installation or the realm,
depending on the URL route used to access the stats.
When suggesting operators to chose, category wise suggestions are
shown instead of a single default suggestion. e.g suggestions for
all the categories of has operator will be show instead of `Messages
with one or more` suggestion which did not make sense.
`has` operator uses predefined categories. This commit displays an
invalid operand message if the operand does not fall in to any of
these categories and the `has` operator is not at the last.
e.g. `has:abc sender:abc@zulipchat.com` will have `invalid abc
operand for has operator, sent by abc@zulipchat.com` as a prefix for
all its suggestions.
Fixes#9384.
Default suggestion e.g `messages with one or more abc` as a suggestion
for `has:abc` is not shown in a new suggestion. But if the has operator
is already present before any other operator, the default message text
will be used. e.g `has:abc sender:abc@zulipchat.com` will have all the
suggestions with the prefix `messages with one or more abc, sent by
abc@zulipchat.com`.
This commit lays the foundation to handle submessages for
plugin widgets. Right now it just logs events, but subsequent
commits will add widget functionality.
Partially fixes#4708.
Implements a first version (v1) for the feature. The next step would be
to allow admins to toggle `is_announcement_only` in the UI.
Fixes#9182. Adds a link to the keyboard shortcuts popup at the
bottom-right corner of the right sidebar. A tooltip saying
`Keyboard Shortcuts(?)` has been added to the icon. The icon is
positioned using `position: fixed`.
We now initialize most modules in ui_init.js, which
isn't the perfect place to do it, but at least now
we have a mostly consolidated entry point.
All the new foo.initialize() methods introduced in
this module run the same order relative to each
other as before this commit. (I did some console
logging with a hacked version of the program to
get the order right.) They happen a bit later than
before, though.
A couple modules still have the `$(function() {`
idiom for miscellaneous reasons:
archive - is a different bundle
common - used elsewhere
list_render - non-standard code style
scroll_bar - no exports
setup - probably special?
socket - $(function () is nested!
transmit - coupled to socket
translations - i18n is a bigger problem
ui_init - this bootstraps everything
We now work with MessageListData objects while populating
data from local narrows, before actually making the
wrapper MessageList object.
This change will simplify unit testing (less view stuff
to fake out) in certain situations.
It will also allow us to eliminate the delay_render flag.
We used to have positional parameters for table_name
and filter, but we don't use them for message_list.all
and we're about to replace filter in some cases.
Passing everything in on opts is more consistent and
self-documenting in the calling code, plus lots of
unit tests can get away with passing in `{}` now
for situations where table_name does not matter.
All of our callers pass in muting_enabled, so we
remove the default value for it. And then the
collapse_messages variable doesn't have to live on
`this` as it's only being passed through down to the
view.
Before this change, the way to add messages had a lot
of ping-pong-ing between MessageList and MessageListData,
where first the data got triaged, but not actually
inserted into data structures, and then subsequent
calls would add the data and get filtered results.
Now we have a simple API for MessageListData.add_messages
that does all the data stuff up front. Having a fully
function MLD.add_messages not only makes the ML.add_messages
function about four lines shorter, it also sets us up
to easily build standalone MLD objects before making
the heavier ML objects.
We now only preserve the offset for the previous
selection (pre-narrow) if that is still the id
we want selected after calling maybe_add_local_messages.
Right not this does not change any behavior, but
upcoming changes to maybe_add_local_messages will
change the selected id to the first unread message
in certain circumstances, in which case preserving
the offset will possibly be confusing, since you're
not on the same message.
We will need this for cases where the topic names in
unread.js are a superset of the names we got from messages.
It's important to pass in a dict of existing dicts to avoid
expensive max() calls to get the max ids of topics (otherwise
the plan would have been to merge the lists in the caller).
Whenever custom fields templates get rendered within user account
settings, on-change event listener wasn't get created. Cause
event listener wasn't set properly.
When org admin tries to change only bots name of no-owner-bot,
It update bots name but returns error, "No such bot owner".
Cause frontend pass `null` value in `bot-owner`.
Template was rendering undefined value of `bot_id` instead of
`user_id`.
Fix this by replacing `bot_id` with `user_id` and changing
template data variable to `data-user-id` to avoid
future confusion.
Fixes#9305.
Empty operators are not allowed while parsing narrowing URLs.
`parse_narrow` stops parsing further if it encounters an empty string
operator.
Shubham Dhama
Aditya Bansal
Harshit Bansal
Steve Howell
Yashashvi Dave
Cynthia Lin
Shubham Padia
Marco Burstein
Tim Abbott
Abhigyan Khaund
Sarah
Aastha Gupta
Joshua Pan
Rohitt Vashishtha
Anupam Dagar
Rhea Parekh
Shayan Toqraee
Anders Kaseorg
Armaan Ahluwalia
Akash Nimare
Vishnu Ks
Jack Zhang
Felix Yan
Robert Hönig
Akash Nimare
Rishi Gupta
Eeshan Garg
Nikhil Kumar Mishra
Aman Ghebreweldi
Alena Volkova