Find: /isset\(\s*([^()]+?)\s*\)\s*\?\s*\1\s*:\s*/
Replace with: '\1 ?? '
(Everywhere except includes/PHPVersionCheck.php)
(Then, manually fix some line length and indentation issues)
Then manually reviewed the replacements for cases where confusing
operator precedence would result in incorrect results
(fixing those in I478db046a1cc162c6767003ce45c9b56270f3372).
Change-Id: I33b421c8cb11cdd4ce896488c9ff5313f03a38cf
The replaces the hacky use of onTransactionIdle(), which no longer runs
immediately in explicit transaction rounds since d4c31cf841.
Also clarified TransactionRoundDefiningUpdate comment about rounds.
Change-Id: Ie17eacdcaea4e47019cc94e1c7beed9d7fec5cf2
Since it calls LBFactory::commitAndWaitForReplication, blocking for
an unknown amount of time, it is a bit "heavyweight" to be the sort
of quick update that uses onTransactionIdle().
Change-Id: Ic3476de5b61a7cd5f31ba487b186f9605a3b2774
The "AUTO" means AUTOCOMMIT, not "automatic transactions"/DBO_TRX,
which is basically the opposite concept. The new name does not
suffer from that ambiguity.
Keep the old constant as an alias for backwards compatibility.
Also remove LoadBalancer comment about non-existing field
Change-Id: I63beeb061fc9be73f320308e4d6393b58628b8c8
In order to switch non-idempotent jobs without losing the
backlog in redis we should support read-only mode in the queue,
where the messages only gets written into the new queue.
Change-Id: I2e9cb2d9cbbd2d657d042b55d4ea0819d21cdd6f
Storing the user name or IP in every row in large tables like revision
and logging takes up space and makes operations on these tables slower.
This patch begins the process of moving those into one "actor" table
which other tables can reference with a single integer field.
A subsequent patch will remove the old columns.
Bug: T167246
Depends-On: I9293fd6e0f958d87e52965de925046f1bb8f8a50
Change-Id: I8d825eb02c69cc66d90bd41325133fd3f99f0226
This could possibly use LoggerFactory & friends to use them as
parameters, but either way let's sort the entries so they're
at least remotely capable of self-grouping.
Change-Id: Iaf435093d70add02e2f82e3037c64fec11870979
This pattern is already used elsewhere and seems like a more efficient
way to acquire locks in a non-blocking way.
Change-Id: Idb369e7cb03b793d5f8295e956fecd8d1f849e17
Each expired row has to be fetched from the user_groups table, deleted
from that table and added to the user_former_groups table. Per Jaimes
request, let's not do this for all rows at once but for smaller chunks
and wait for replication to catch up after each chunk has been
processed. In addition the function to purge the expired rows now sets a
lock so that there won't be multiple concurrent runs.
Also, cleaning this table up isn't urgent and thus should be done in a
job and not a deferred update, so let's move it there.
Bug: T176754
Change-Id: I671d4b9d09677a2f474477ba7fea33a44d6318aa
Clean up use of @codingStandardsIgnore
- @codingStandardsIgnoreFile -> phpcs:ignoreFile
- @codingStandardsIgnoreLine -> phpcs:ignore
- @codingStandardsIgnoreStart -> phpcs:disable
- @codingStandardsIgnoreEnd -> phpcs:enable
For phpcs:disable always the necessary sniffs are provided.
Some start/end pairs are changed to line ignore
Change-Id: I92ef235849bcc349c69e53504e664a155dd162c8
* Remove some creation dates, they are not protected by GPL
* Remove duplicate @defgroup API
* Remove @ingroup from some @file doc comments on class files. It is not
useful to list class files alongside classes in the doxygen module menu.
Add @ingroup to some more class files that had @ingroup on their file,
that was probably the author's intent.
* In PackedOverlayImageGallery, use the file comment as a class comment
* Don't put @defgroup and @file in the same comment. @defgroup makes the
whole doc comment describe the group.
* Instead of putting AnsiTermColorer in two groups, use hierarchical
groups.
Change-Id: If54f6e0b2bc1ea6de42045885cf836ee67b8e961
Use $wgUpdateRowsPerQuery instead, and use its value at run-time
instead of inside the job parameters. Also helps with deduplication
if batchSize is changed.
Change-Id: Ifef25a3ae5ae2418359a41eba05778613fbb548f
Previously, the INNER JOIN had "rc_timestamp >= rev_timestamp" which
complicates query planning. Even with "equals" it still was techinally
ambiguous. Instead, just use EXISTS and an exact equality operator.
Bug: T180793
Change-Id: I1e9ae7c2ce0f95484e09e867550283d816d151f0
Several classes have a "selectFields()" static method to tell callers
which fields to select from the database. With the recent comment table
change and the upcoming actor table change, this pattern has become too
simplistic as a SELECT will need to join several tables to be able to
retrieve all the needed fields.
Thus, we deprecate the selectFields() methods in favor of getQueryInfo()
methods that return tables and join conditions in addition to the
fields.
Change-Id: Idcfd15568489d9f03a7ba4460e96610d33bc4089
This allows de-duplication of single page jobs for the
same page due to edits to different templates. This is
the same logic that RefreshLinksJob already has.
Also fix a bug in that method in RefreshLinksJob.
Change-Id: I2f79031c945eb3d195f9dbda949077bbc3e67918
These were meant as sanity checks, but would fail in those
unusual cases anyway with exceptions. Instead, have an
early check to make sure no explicit transaction rounds
are active when JobRunner:run is called.
Change-Id: I723c77c8d3ef7ec4dcf09ce6d549b4fd57bdf1c2
BacklinkJobUtils consistently uses the "pages" field in leaf
jobs, even when there is only one page per leaf job.
RefreshLinksJob already has this logic for de-duplication.
Change-Id: Ia189bbc9df44f2161cfed4192c23b2ac3cfa65ce
* The condition timestamp should be the root job timestamp,
as the comments imply, though that was not the case. Only
the SET timestamp should be the current timestamp. As it
was, jobs effectively did not "see" each other since each
one would be expecting newer than the ones that ran prior.
The page_touched condition for the UPDATE query was mostly
useless, since it was always the current timestamp.
* This problem was a regression from f598ca81e3.
Change-Id: I1398366e87c959be1c98e4db959c41309b0ac1b8
These are there to handle regeneration race conditions due to
replication lag. If there happens to already be a huge amount
of purges, the cost of this can be too high. Also, if the queue
is backlogged, then it is already delayed, so extra delayed jobs
do not help and just add to the backlog in that scenario.
Bug: T173710
Change-Id: Ida0bf44233072e1356a9fe63866d522e4bb2d0ca
Do not count jobs that just make subdivide as having any
"work items". This makes $wgJobBackoffThrottling less
overzealous when used to limit these type of jobs.
The main reason to limit htmlCacheUpdate would be for
CDN purge rate limiting. For refreshLinks, it would
mostly be lag, though that is already handled for
leaf jobs and JobRunner itself.
Bug: T173710
Change-Id: Ide831b555e51e3111410929a598efb6c0afc0989
This is useful for things like SqlBagOStuff and JobQueueDB
being able to write in auto-commit mode while the main
transaction round still goes on.
SqlBagOStuff already had this sort of logic.
JobQueueDB now also takes this approach so it does not have to
defer insertion except for sqlite. This makes callers more likely
to know when insertion failed or not.
Make sure LoadBalancer sets "master"/"replica" LB info itself;
this fixes a bug found in the new tests.
Bug: T140338
Bug: T42451
Change-Id: I4199a9598d7afbf976a6efa8ed84b85b56da02bd
