| Commit message (Collapse) | Author | Age | Lines |
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
| |
Instead of a many-many relationship, one filterlist has multiple
filters. Nested serialization is read-only by default, so not
all CRUD methods are implemented yet for the FilterList viewset.
|
| | |
|
| | |
|
| |
|
|
| |
This also allows us to simplify the str dunder for a FilterList.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
This will take the currently defined filter list and put them inside the new schema while trying to keep defaults similar to our current setup.
|
| | |
|
| |
|
|
| |
This commit adds new filter schema as described in #479
|
| |
|
|
|
|
| |
This was changed due to performance reasons, but after some tweaking in the database, such as increasing work memory and adding an index, this query runs much faster now.
To test this, I want to revert this change, so that we can stop the materialised view from refreshing, to see if the act of refreshing is what's causing this query to seem faster when runing against the database.
|
| |\ |
|
| | | |
|
| |/
|
|
|
|
| |
We have added a new view to metricity that will keep track of an approximate message count, updating every 10 seconds.
By doing this, we avoid running a query against the whole message table every time we want to get a user's messages.
|
| |
|
|
|
| |
From now on the signal will only get executed after the Role has been deleted
The commit also introduces minor changes in the tests of roles
|
| | |
|
| |
|
|
| |
Create a test that checks if a role gets deleted it will also get unassigned from the user
|
| |
|
|
|
| |
Add a signal to the api app that automatically unassigns
deleted roles from users that have them
|
| |\ |
|
| | |\ |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | |
| | | |
Previously this query would convert each row to an array just to check if it matched or not. By changing EXCLUDE_CHANNELS to a tuple instead of a list, it doesn't get passed as an array, so we can do a simple NOT IN check.
This will also allow us to add an index with this condition to speed it up further.
|
| | | | |
|
| | | |
| | |
| | |
| | |
| | |
| | | |
We only actually care that thee key:value pairs are equal, the order of them isn't actually important.
The naming of `assertCountEqual` is a little misleading, since it actually tests that the first sequence contains the same elements as second, regardless of their order. See https://docs.python.org/3/library/unittest.html#unittest.TestCase.assertCountEqual
|
| | | |
| | |
| | |
| | |
| | |
| | | |
We only truly care about how many activity blocks a user has when they have a small number of messages, as we only use this for the voice gate.
If a user has more than say 1k messages, then we really don't need to calculate their activity blocks, as it's quite an expensive query with many messages.
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | | |
Adds a new filter list for URLs which should be treated as
redirects and unfurled.
Signed-off-by: Hassan Abouelela <[email protected]>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Adds an integer field to reminders, to keep track of how many failed
attempts at delivering the reminder were made.
Updates documentation and serialization.
Signed-off-by: Hassan Abouelela <[email protected]>
|
| | | |
| | |
| | |
| | |
| | |
| | | |
Dynamically adds static HTML redirects for static builds.
Signed-off-by: Hassan Abouelela <[email protected]>
|
| | | | |
|
| | |/
|/|
| |
| |
| |
| |
| | |
- Add newlines so it's not a big unfriendly paragraph
- Fix gitpod workspace link
- Clarification on terminal commands
- Clarification on test server and bot
- Add links for that ^ and for environment variables
|
| | | |
|
| |\ \ |
|
| | | |
| | |
| | |
| | |
| | |
| | | |
Fixes a test that fails on Windows due to differing path separators.
This is currently the only failing test.
Signed-off-by: Hassan Abouelela <[email protected]>
|
| | | |
| | |
| | |
| | |
| | |
| | |
| | |
| | | |
Ref:
https://discord.com/channels/267624335836053506/429409067623251969/896753201645232169
I think it should be updated so newer members won't get confused when
they don't see a #media-branding.
|
| | | |
| | |
| | |
| | |
| | | |
- succint -> succinct
- seperated -> separated
|
| | | | |
|
| |/ /
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
./pydis_site/apps/resources/resources/tools/ides/thonny.yaml:1: specifically
./pydis_site/apps/content/resources/guides/pydis-guides/helping-others.md:115: considered
./pydis_site/apps/content/resources/guides/pydis-guides/contributing/issues.md:59 labels
./pydis_site/apps/content/resources/guides/pydis-guides/contributing/sir-lancebot.md:99: recommend
./pydis_site/apps/content/resources/guides/pydis-guides/contributing/site.md:111: particularly
./pydis_site/apps/content/resources/guides/pydis-guides/contributing/sir-lancebot/env-var-reference.md:29: Integer
./pydis_site/apps/content/resources/guides/pydis-guides/contributing/sir-lancebot/env-var-reference.md:67: calculating
./pydis_site/apps/api/tests/test_off_topic_channel_names.py:157: response
I didn't touch the code jam and game jam typos because I'm not sure if
they should be preserved as is.
There were a few 'seperated' typos which I didn't change because I
*think* it's just another way of spelling it?
In the offensive words test there was a keyword argument named `fied`
which I didn't touch because I wasn't sure where that was from.
|
| | | |
|
| |\ \ |
|