Also adds a custom rule to eslint. Since the recommended way of extending
eslint is to create plugins as standalone npm packages, the separate rule
is published as 'eslint-plugins-empty-returns'.
Fixes#8669.
This change prepares us to have the server send avatar_url
of None when somebody wants a gravatar avatar (as opposed
to a user-uploaded one).
Subsequent commits will change behavior on both the server
and client to have this happen. So this commit has no-op
code for now, but it will soon use the fallback-to-gravatar
logic.
This commit is easy to revert if we want to tone down errors
to warnings for the short term, while our codepath still does
proper handling for adding users when they come in messages.
This logic used to be in extract_people_from_message(), but
we are deprecating extract_people_from_message(), whereas
the maybe_incr_recipient_count() function has logic that we
want to keep.
This change is the first step in making it so that we load
non-active users at page load time in the webapp.
Before this change, we would reactively handle deactivated
users when we saw them as senders in messages. Subsequent
changes will make it a warning if we see unknown senders
in messages.
We were incorrectly reporting active bots as non-active in
popovers, and we had no test coverage for cross-realm bots.
We also rename the function to is_active_user_for_popover,
since the old name, realm_user_is_active_human_or_bot, suggested
the wrong semantics for cross-realm bots.
Last but not least, we only do a blueslip warning if a user id
is not found. When lookups fail, we are pretty confident that
the user is not active, so an error is overkill. We can change
that as part of issue #7120.
Fixes#7153
It's easier to unit test logic inside of people.js than compose.js.
We allow users to compose emails to any of our cross-realm bots.
Someday we may tighten up which cross-realm bots are valid targets,
since it's not necessarily the case that those bots do anything
useful when you send them messages.
This dictionary includes bots, so the reference to
"people" in the name `realm_people_dict` was misleading.
We omit `realm` for brevity sake--it's usually the case
that folks implementing new features can safely ignore
cross-realm bots, and it's on our roadmap to move those
bots into the realm.
The function name `get_realm_human_user_ids` was a lie--it
includes active bots as well.
The only user of this function is `activity.js`, which wasn't
impacted by the misleading name, because we eventually filter
out bots in the `info_for` function.
It's possible that we actually want to include bots in the right
sidebar, since they can be difficult to discover in other parts
of the UI. Or, if we want to keep the right sidebar as all
human users, we may eventually want to make the logic to exclude
bots happen higher in the stack (but for real, this time).
Apparently this is a bug that slipped in when we started showing
normal users as deactivated in the user popovers: all bot users were
treated that way as well.
We'll want to do #7153 as a follow-up to get things fully working how
we want them.
This allows user to view all group private conversation messages
with a specific user. That is, it views all the the group private
messages from groups which include the given user.
Add search suggestion for group-pm-with. Add operator name
and description in "Search operators" tab.
Add change in tab name to "Group Messages" when using this operator.
Add frontend_tests for group-pm-with search operator.
Fixes: #3882.
It's not always clear whether user_ids are strings or integers, so
we explicitly convert them to integers for sorting when creating
keys for PMs.
To keep the tests passing, this commit removes some unneeded
defensive code in message_store.js that only applies to contrived
test input.
We had a theory that get_user_id() errors were often due to race
conditions related to reloads, so we would only report missing
user ids if subsequent lookups failed 5 seconds later. It turns
out we still get the blueslip errors, and now we don't get
meaningful tracebacks. This change makes it so that errors
get reported immediately again.
In f75af94984 I added some
lines of code that made it so that live updates for avatar
urls would affect messages currently in the browser.
This change worked well when the live update actually happened,
but then the next time the user would reload, the avatar in
the message pane would regress back to showing the avatar urls
from the server (which could have caching issues of their own).
This fix removes a couple lines of code that had the intended
effect of making all of your messages from any given sender
show the same url (good) but which generally grabbed
the url from an old message (bad).
After this fix, we go back to having old messages possibly
showing the old avatar urls, but new messages will display the
new avatar.
(There are lots of moving parts in the avatar system, because
not only do browsers cache image urls, but our server caches
messages and recipient info, so there have been "fixes" to
avatars since this change that are valid fixes in their own
right but not directly relevant to this commit.)
We use to have client-side logic that would append timestamps
or random numbers to avatar URLs to force browsers to
refresh their cache.
We no longer need this now that the back end maintains
versions for avatar changes and puts the version in the URLs.
When we process messages for unread counts, we now call
people.pm_reply_user_string() to get a string of user ids,
rather than using emails that may have changed since the
message was originally created.
We now convert our pm-with search operand to a list of user ids
for matching against messages, rather than using emails. On the
message side we look at user ids from display_recipient.
We have added people.pm_with_url(message), which computes a
PM url from a private message using user ids rather than emails.
We call this in add_message_metadata(), since the slugs will
be valid even if emails change, so we don't need to compute
them on the fly during message rendering.