Implement few optimizations for reading admin's bot dicts from database
for a constants number of requests:
- add models.get_user_profiles_by_ids() for reading bots profiles
by single query from database
- add models.get_services_for_bots() for reading services for bots
by single query from database
- add bot_config.get_bot_configs() for reading config data for bots
by single query from database
Fixes#8838
This was causing a rather confusing test flake in
test_stream_error_pm_to_bot_owner. What was happening was that if
this test (which used that code path) ran within 5 minutes of the
populate_db run, it would fail.
Applies the logic to allow community members to edit topics
of others' messages if this setting is True. Otherwise,
only administrators can update the topic of others' messages.
This logic includes a 24-hour time limit for community topic editing.
This commit migrates realm emoji to be addressed by their `id` rather
than their name. This fixes a long standing issue which was causing
an error on uploading an emoji with same name as a deactivated realm
emoji.
Fixes: #6977.
This makes this value much easier for a server admin to change than it
was when embedded directly in the code. (Note this entire mechanism
already only applies on a server open for anyone to create a realm.)
Doing this also means getting the default out of the database.
Instead, we make the column nullable, and when it's NULL in the
database, treat that as whatever the current default is. This better
matches anyway the likely model where there are a few realms with
specially-set values, and everything else should be treated uniformly.
The migration contains a `RenameField` step, which sounds scary
operationally -- but it really does mean just the *field*, in
the model within the Python code. The underlying column's name
doesn't change.
Other functions took the form of returning Sequence[T] when the QuerySet
functionality is unused beyond the function, with T being the objects
filtered for in the function body; this commit follows that practice for the
one remaining python2 comment-annotated function, completing the transition
of models.py to py3.5 function annotations.
A note is also added to another function regarding a need to return a
QuerySet, and ideally a QuerySet[T] in line with the other functions, as and
when QuerySet becomes annotated as a generic.
We also delete a couple helper functions that were only used there.
This management command was primarily used before we had a UI for
creating outgoing webhook bots.
In theory, we should be able to delete this, since first, if there are
no users in the organization, we'll end up with an equivalent value
(an empty collection of users), and second, it shouldn't be possible
for an active Zulip realm to have 0 active users in it anyway.
But the way we construct the database query in query_for_ids is such
that it's necessary to avoid a 500.
This commit refactors the bugdown to perform a lookup only on active
realm emojis. This was needed because once we migrate realm emojis
to be addressed by `id` rather than name, it will be costly to
perform a lookup on all the realm emojis.
Till now, we had been storing realm emoji's name in emoji code field
in reactions' model. This commit migrates it to store realm emoji's id.
It is a part of effort to migrate realm emojis to be referenced by their
id and not by name.
Add `translate_emoticons` to `prop_types` and `expected_keys`.
Furthermore, create a emoji-translating Markdown inline pattern.
Also use a JavaScript version of `translate_emoticons` and then use
this function during Markdown previews and as a preprocessor. This
is only needed for previews, because usually emoticon translation
happens on the backend after sending.
Add tests for emoticon translation, a settings UI, and a /help/ page
as well.
Tweaked by tabbott to fix various test failurse as well as how this
handles whitespace, requiring emoticons to not have adjacent
characters.
Fixes#1768.
models.py should only contain thin wrapper functions. Furthermore,
this move allows us to remove the circular imports. The two moved
functions are interdependent and are thus moved in one commit.
Users having only account in one realm will not be distracted by realm
name in subject lines of every email. Users who have multiple
accounts in realms can turn this setting on and receive a
corresponding realm name in email's subject.
Tweaked by tabbott to rebase and address a few small issues.
Fixes#5489.
This is the first step for allowing users
to edit a bot's service entries, name the
outgoing webhook configuration entries. The
chosen data structures allow for a future
with multiple services per bot; right now,
only one service per bot is supported.
This commit adds a setting to limit creation of generic bots
to admins for realms that want that restriction. (Generic
bots, apart from being considered spammy on some realms,
have less locked down permissions than webhook bots).
Fixes#7066.
We no longer have a special UI setting and model
field ("emoji_alt_code") for saying users want text-only
emojis. We now instead make "text" be a fifth choice
for "emojiset".
Fixes#7406
The original logic is buggy now that emails can belong to (and be
invited to) multiple realms.
The new logic in the `invites` queue worker also avoids the bug where
when the PreregistrationUser was gone by the time the queue worker got
to the invite (e.g., because it'd been revoked), we threw an exception.
[greg: fix upgrade-compatibility logic; add test; explain
revoked-invite race above]
[Modified by greg to (1) keep `USERNAME_FIELD = 'email'`,
(2) silence the corresponding system check, and (3) ban
reusing a system bot's email address, just like we do in
realm creation.]
Storage limititations are only set on the value of
a config entry, since this is the only user-accessible
part of the schema. Keys are statically set by each
embedded bot.
This restyles and rewords some of the emoji style section to look
better and fit it more with the current style guide.
Tweaked by tabbott to modify the historical migration rather than
adding a new one. This is OK because the emojiset choices text change
doesn't touch the database; it's just a Django Python code thing.
Also removed translation tags, since we don't need them for a set of
brand names.
This commit puts the guts of parse_usermessage_flags into
UserMessage.flags_list_for_flags, since it was slightly faster
than the old implementation and produced the same results.
(Both algorithms were super fast, actually.)
And then all callers use the model method now.
The main limitation of this version is that it's controlled entirely
from settings, with nothing in the database and no web UI or even
management command to control it. That makes it a bit more of a
burden for the server admins than it'd ideally be, but that's fine
for now.
Relatedly, the web flow for realm creation still requires choosing a
subdomain even if the realm is destined to live at an alias domain.
Specific to the dev environment, there is an annoying quirk: the
special dev login flow doesn't work on a REALM_HOSTS realm. Also,
in this version the `add_new_realm` and `add_new_user` management
commands, which are intended for use in development environments only,
don't support this feature.
In manual testing, I've confirmed that a REALM_HOSTS realm works for
signup and login, with email/password, Google SSO, or GitHub SSO.
Most of that was in dev; I used zulipstaging.com to also test
* logging in with email and password;
* logging in with Google SSO... far enough to correctly determine
that my email address is associated with some other realm.
This adds the data model and bugdown support for the new UserGroup
mention feature.
Before it'll be fully operational, we'll still need:
* A backend API for making these.
* A UI for interacting with that API.
* Typeahead on the frontend.
* CSS to make them look pretty and see who's in them.
The first method we extract to this library is
get_active_subscriptions_for_stream_id().
We also move num_subscribers_for_stream_id() to here, which
is slightly annoying (having the method on Stream was nice)
but avoids some circular dependency issues.
This extraction moves all the huddle logic into models.py, which
hopefully can reduce friction for things like re-organizing our
caches (there are two cache entries for every huddle) and/or
just putting huddle_id on Message directly.
Do you call get_recipient(Recipient.STREAM, stream_id) or
get_recipient(stream_id, Recipient.STREAM)? I could never
remember, and it was not very type safe, since both parameters
are integers.
Tweaked by tabbott to have the field before the invitation is
completed be called invite_as_admins, not invited_as_admins, for
readability.
Fixes#6834.
This has exactly the same behavior so long as self.subdomain contains
no colon character, ':'; and of course we don't allow those in
subdomains, because they aren't allowed by DNS.
Adds support to add "Embedded bot" Service objects. This service
handles every embedded bot.
Extracted from "Embedded bots: Add support to add embedded bots from
UI" by Robert Honig.
Tweaked by tabbott to be disabled by default.
This is a prepatory commit that adds non-active users to
the realm user cache. It mostly involves name changes and
removing an `is_active` filter from the relevant DB query.
The only consumer of this cache is `get_raw_user_data`, which
now filters on `is_active` in a dictionary comprehension (but
this will get moved around a bit in a subsequent commit).
Since a user could use the same installation of the Zulip mobile app
with multiple Zulip servers, correct behavior is to allow reusing the
same token with multiple Zulip servers in the RemotePushDeviceToken
model.
Message.get_raw_db_rows is moved to MessageDict, since its
implementation details are highly coupled to other methods
in MessageDict.
And then sew_messages_and_reactions comes along for the
ride.
We eventually want to move Reaction.get_raw_db_rows to there
as well.
Add this field to the Stream model will prevent us from having
to look at realm data for several types of stream operations, which
can be prone to either doing extra database lookups or making
our cached data bloated.
Going forward, we'll set stream.is_zephyr to True whenever the
realm's string id is "zephyr".
This removes sender names from the message cache, since
they aren't guaranteed to be valid, and they're inexpensive
to add.
This commit will make the message cache entries smaller
by removing sender___full_name and sender__short_name
fields.
Then we add in the sender fields to the message payloads
by doing a query against the unique sender ids of the
messages we are processing.
This change leads to 2 extra database hops for most of
our message-related codepaths. The reason there are 2 hops
instead of 1 is that we basically re-calculate way too
much data to get a no-markdown dictionary.
We now have a dedicated cache for active_user_ids() that only
stores a list of user_ids.
Before this commit, active_user_ids() used a cache of UserProfile
dictionaries, so it incurred unnecessary deserialization costs for
all the user fields that it sliced away in a list comprehension.
Because the cache is skinnier here, we also need to invalidate it
less frequently. Basically, all we care about is new users, realm
deactivations, and user deactivations.
It's hard to measure how much this will improve performance, because
the speedup for any operation here is pretty minor, but we use this
function a lot, so hopefully it will make the overall system more
healthy.
Previously, the bot domain was calculated correctly in most
circumstances, but if you were using the root domain, it would be
e.g. ".chat.zulip.org", not "chat.zulip.org". We fix this, with
perhaps more use of setting REALMS_HAVE_SUBDOMAINS than would be ideal
if we weren't about to set that True unconditionally.
This change optimizes get_status_dict_by_realm() by
introducing query_for_ids(), which quickly computes
an "IN" clause for user ids.
This change also inlines the `two_weeks_ago` check, but
that is just for clarity, not performance.
The prior version of this function was passed in a QuerySet, which
made it difficult to effectively profile the callers, and there
is really no compelling reason to pass in a query any more.
Avoid a join to UserProfile here speeds up the query from
86ms -> 28ms when you analyze it with about 2000 mobile users
in a 5000-user realm.
We also avoid some code duplication here, since we filter
UserPresence for the same group of users as we filter
PushDeviceToken.
This avoids an O(N-squared) hit during presence queries. The speedup
here is probably negligible compared to everything else going on, but
sets are more semantically correct, anyway.
Before this commit, postgres would choose a non-optimal query
plan to find all presence rows belonging to a realm. We now
do an extra query to get the list of relevant user_ids, which allows
the next query to take advantage of UserPresence's index on
user_profile_id.
Here is the query plan for the offending query (this particular query isn't
verbatim from the code, but it's representative of the problem):
explain analyze
select client_id
from zerver_userpresence
INNER JOIN zerver_userprofile ON
zerver_userprofile.id = zerver_userpresence.user_profile_id
WHERE
zerver_userprofile.is_active and
zerver_userprofile.realm_id = 3;
Hash Join (cost=149.66..506.82 rows=5007 width=4) (actual time=48.834..121.215 rows=5007 loops=1)
Hash Cond: (zerver_userprofile.id = zerver_userpresence.user_profile_id)
-> Seq Scan on zerver_userprofile (cost=0.00..260.11 rows=5369 width=4) (actual time=0.009..24.322 rows=5021 loops=1)
Filter: (is_active AND (realm_id = 3))
Rows Removed by Filter: 3
-> Hash (cost=87.07..87.07 rows=5007 width=8) (actual time=48.789..48.789 rows=5010 loops=1)
Buckets: 1024 Batches: 1 Memory Usage: 196kB
-> Seq Scan on zerver_userpresence (cost=0.00..87.07 rows=5007 width=8) (actual time=0.007..24.355 rows=5010 loops=1)
Total runtime: 145.063 ms
You can see above that we're filtering on realm_id instead of using an index.
When you decompose the query into two queries, the total time is about 100ms, for a
savings of 33%. I imagine the savings would be even greater on an instance with lots
of realms. This was tested on dev with one really large realm and one tiny realm.
We were using `.order_by('user_profile_id', '-timestamp') in our
UserPresence query in get_status_dicts_for_query.
We don't need a full sort to produce the dictionary of statuses.
In fact the whole operation in Python is still O(N):
- divvy rows up to be per-user in an O(N) pass
- find max row for the 'aggregated' entry in an O(n) pass
per user
The one minor annoyance of this fix is that datetime_to_timestamp
is lossy, so if you naively call to_presence_dict before finding
the "max" row, you get test flakes if rows are created during the
same second. I decided to avoid calling to_presence_dict so there
are fewer moving parts, but there's still the ugly step of having
to remove the "dt" field from the final results.
novokrest
Umair Khan
Shubham Dhama
Tim Abbott
neiljp (Neil Pilgrim)
Sarah
Harshit Bansal
Greg Price
Vishnu Ks
YJDave
Marco Burstein
Robert Hönig
Aastha Gupta
Roman Godov
rht
Abijith10
Aditya Bansal
Alena Volkova
Rishi Gupta
Rhea Parekh
Brock Whittaker
Steve Howell
Abhijeet Kaur
Vishnu Ks