No longer used from there by the mediawiki.cookie module
No deployed javascript tries to use these, but a few undeployed
extensions do, so added to release notes.
Bug: T291760
Follow-up: I600332cdb738f0b443549948b9f070f3ccfa12aa
Change-Id: Ia02048137f191b950ae789a9d206c1d9afbce83a
All of my favorite text editors corrupt this test case whenever I edit
parserTests.txt. extraParserTests.txt contains other tests with weird
characters that may get corrupted by normal text editors.
(I had to use `vi` to make this patch, and I wouldn't wish this on
anyone.)
Change-Id: Id474469180fc284e3e28b55f65808be727507875
- Added a test where ParserOutput objects with CacheTime
properties set are unserialized from previous versions.
- Generate new serialization tests for 1.38
Now all serialization in production is JSON, so changing
property visibility shouldn't affect ParserCache.
Bug: T263851
Depends-On: I283340ff559420ceee8f286ba3ef202c01206a23
Change-Id: I70d6feb1c995a0a0f763b21261141ae8ee6dc570
In ParserOutput serialization tests we have a large number of examples
saved for different versions of MW. When MW version is released,
but the serialization doesn't change, we stop running some serialization
tests cause we can no longer find the latest example.
Change-Id: I01e8e0e17e2371b2cd08d200d3a4611d814e74bb
Follow up to the soft deprecation in I9822c60c180d204bd30cb4447a1120155d456da4.
Code search:
https://codesearch.wmcloud.org/deployed/?q=%28allow%7Cprevent%29Clickjacking&i=nope&files=&excludeFiles=&repos=
Note that the dependencies here are actually cleanups to
OutputPage::preventClickjacking(), not ParserOutput::preventClickjacking().
It can be hard to tell the difference between these two methods when
reading code (especially when parameter types are omitted) and so we're
being extra cautious by cleaning up both of these together.
Bug: T287216
Depends-On: I626e89a1bff2d9c535b828a5b25eff863c91f858
Depends-On: I23e8b35540d2da34c0a3d335069c518b4b72a333
Depends-On: I2538ac8f37fcde183f496e49a8de0ed25a38e508
Change-Id: I0f9b1fe755f7d3d6a9b18fa21a738dda7566c211
Fix the last drifts between the two. Custom CSS IDs/names/classes in
FileDeleteForm were replaced with the base one, as I think it's better
to be consistent, and it wouldn't be trivial to allow changing those
values in a clean way. Question for reviewers: should this part be
announced on tech news in case somebody is using those values in a
script?
Move the file delete intro above the form, like we do for normal
deletion.
For normal deletion, use the content of the wpReason field as default
value for that field, not the result of merging the dropdown and the
"other" field, which doesn't seem correct (the file version was already
doing this).
Make the file deletion form use the title as token salt like the base
form -- it shouldn't make any difference AFAICS.
After this patch, it should be possible to make DeleteAction extend
FormAction with low effort.
Bug: T288282
Change-Id: I8b115ecb1c94af63d686a575b18362ac32105ae2
I was under the impression that live preview could only be used on
wikitext pages, but this isn't the case. So pass the proper content
model to the API.
Bug: T293022
Change-Id: Icff5b8b25a4b0df3feda2093d5b9c92e6dd9be3a
The word "ဇန်နဝါရီ" means "January". It appeared in all the dates,
which is obviously wrong.
This was reported by User:Ninjastrikers.
Change-Id: I1491301d253a59468caf9b00626f341be8810073
After the mediawiki.rcfilters.filters.(dm|ui) modules were merged,
mw.rcfilters.js was just `require`ing the old init script to run some
extra setup, no need for separate files.
Bug: T256836
Change-Id: I6462151152d075af0ef8bce3a5edf0a0ba48761a
It's all great that it's up and running, but how does one stop the
containers when you're done with it and want/need to reduce load on
the computer, or re-create them after something went wrong?
Change-Id: I9ec446d73a5fac41550c19eb83912469cfdb92ee
Sending one packet per metric is inefficient. Collapse all rendered
metrics into a single array to make the best use of the max payload size
of a UDP packet.
Bug: T240685
Change-Id: I74fcc222a7e8610b09fa14c5782c317d1e697d1d
Previously, when a user with correct permissions uses completion search on a
private wiki, the results are returned and cached. Since we are on a private wiki,
we don't want to cache results since the content is not accessible to all users.
Now, content that is not accessible to all users will not be cached.
This patch achieves this by setting the appropriate Cache-Control response headers
for the MW REST Search endpoint.
Bug: T292763
Change-Id: I693b4088df9c0520d5238c286312ec52ab273604
$generateHtml is now wrapped in the ContentParseParams, referring to it
by its old name is confusing.
Bug: T287158
Change-Id: I31e0a596881ecd7d3bea709ab3f8029779cb7f02
The ::getIPBlock() method returns all current blocks affecting the IP
address.
Supporting changes:
* Extract ::getXffBlocks() from ::getAdditionalIpBlocks() and rename the
latter to ::getSystemIpBlocks()
* Add ::getIpBlock(), which returns an AbstractBlock instance that
represents all blocks affecting the IP, including range blocks and
system blocks (using ::getSystemIpBlocks())
Additional changes:
* Update DocBlocks for all methods mentioned above
Bug: T270318
Change-Id: I2009c3a7143c10167461e1ce9d4b8efe81390ff6
This is a fairly low-risk right, I think viewing
deleted content (viewdeleted) or OS logs (viewrestrictedlogs)
should be separate from this. The basic grant already has
other low-risk rights, like patrolmarks or abusefilter-log-detail,
so I think it should be fine to add this too.
Alternatively, we can create a new grant instead.
Bug: T292940
Change-Id: Ied76fe0e6f896e56029fb40f1197e87f3f5ff68a
