Translation updater bot [Tue, 1 Sep 2015 19:26:51 +0000 (21:26 +0200)]
Localisation updates from https://translatewiki.net.
Change-Id: Ie07f3d57d81c5b1f645c95397f10a31f18c5d6d1
jenkins-bot [Tue, 1 Sep 2015 16:45:58 +0000 (16:45 +0000)]
Merge "EditPage: Add text placeholders to view source page"
Timo Tijhof [Fri, 28 Aug 2015 23:12:07 +0000 (01:12 +0200)]
resourceloader: Remove redundant type-checks in internal code
The private 'execute()' method only runs for modules that have been
implemented by mw.loader.implement, which already throws if 'messages'
or 'style' is set to a non-object value.
The execute() code for 'templates' already has this check removed.
Change-Id: I65b8a87500388d3d2ca6e495e1e50179639ba27a
jenkins-bot [Tue, 1 Sep 2015 16:32:11 +0000 (16:32 +0000)]
Merge "resourceloader: Consistently set state=ready after script execution (not before)"
jenkins-bot [Tue, 1 Sep 2015 16:27:31 +0000 (16:27 +0000)]
Merge "Add ForeignUpload and ForeignStructuredUpload"
Timo Tijhof [Fri, 28 Aug 2015 22:56:24 +0000 (00:56 +0200)]
resourceloader: Consistently set state=ready after script execution (not before)
For the case of 'module.script' being an array, mw.loader already sets state=ready
after execution. For the cases of it being a function or string, we were setting
state=ready before execution.
This looks like it may have intended to prevent double execution where e.g.
some other module may need this module and see it has state 'loading' and start
executing it. However ResourceLoader doesn't expose execute(). The only code
calling execute() is code that also sets 'state=loading', or handlePending.
In fact, this early setting of "ready" could actually cause double execution.
Not of the current module, but of dependent modules. If a module's script loads
other modules that are cache hits, that could may trigger handlePending() which
will assume the current module to be ready, when it isn't.
Now that these code paths match, re-use the markModuleReady callback instead of
repeating this code.
Update tests:
* Make assertion captions less verbose.
* Change load instruction from "test.implement" to "test.implement.import".
Module "test.implement" doesn't exist. This worked by accident because
implement() can sometimes run a module immediately if it doesn't exist.
* Update test to reflect internal detail that state=loading during script
execution.
* Assert afterwards that script ran and state=ready.
Change-Id: I6b0542edb113d58b8f24cc8587e98ee88b514c55
Mark Holmquist [Wed, 26 Aug 2015 19:42:21 +0000 (14:42 -0500)]
Add ForeignUpload and ForeignStructuredUpload
These classes will be used to send uploads from one wiki to another,
and the latter includes (partial) adherence to the Commons structured
data system [0].
[0] https://commons.wikimedia.org/wiki/Commons:Structured_data
Change-Id: I96fcb862eb854d23b6f9f553a87fa7ca65bf5a93
Bug: T105071
Vivek Ghaisas [Tue, 25 Aug 2015 20:27:41 +0000 (01:57 +0530)]
Ignore languages/messages/Messages*.php in line length sniff
Most of these files have very long lines, but that makes them easier to
read. Also, they're generated from a script. It's probably a better idea
to ignore their lengths than to change the lengths.
Bug: T103806
Change-Id: Ic6860520e4cc76f93aa13c364aae11cf14f279d0
jenkins-bot [Tue, 1 Sep 2015 12:51:26 +0000 (12:51 +0000)]
Merge "Make readonly work for OOUI forms"
Derk-Jan Hartman [Mon, 24 Aug 2015 20:58:33 +0000 (22:58 +0200)]
EditPage: Add text placeholders to view source page
These placeholders are already in use by the edit form, but were not
available to extensions. This gives us equal oppertunity to adapt the
contents surrounding the source textbox as we have in the editform.
I'm reusing the same placeholders. Most extensions fill these
placeholders through the showEditForm:initial hook however, so it
should not cause too many unexpected changes, since the readonly page
uses showReadOnlyForm:initial instead. I left out those that were very
specific to the editpage (afterEditTools etc).
This will be useful for Bug T110050.
Change-Id: Ia03265a907212ee7bac1b839dcdd4d2f24c79bbe
Glaisher [Tue, 1 Sep 2015 12:37:54 +0000 (17:37 +0500)]
Make readonly work for OOUI forms
This seems like a hack. Copied from HTMLTextField.
Change-Id: I46f3dbd4b60fd9a39ecd7167ec903ae7d1d804bc
Timo Tijhof [Fri, 28 Aug 2015 12:00:58 +0000 (14:00 +0200)]
resourceloader: Don't add "&*" query parameter to load.php urls
Follows-up r88883 (
8dab43f703). The IEUrlExtension library only considers the
query string if the base path has no extension (or if the dot is urlencoded).
Neither of which is the case for load.php urls.
If the server doesn't expose undecoded request URIs (per IEUrlExtension)
then it will look in the query string always. However even then it is
only an issue if the last query parameter in the url contains a dot,
which is very unlikely for load.php (they typically end in 'skin' or
'version' which don't contain plain dots).
Ref T30840.
Change-Id: I4a4ef7ef5ad7d74ecb47a44342e6324fa9844173
jenkins-bot [Tue, 1 Sep 2015 00:45:12 +0000 (00:45 +0000)]
Merge "Fixed and suppressed some thumb.php IDEA errors"
jenkins-bot [Tue, 1 Sep 2015 00:44:09 +0000 (00:44 +0000)]
Merge "Clarified WANObjectCache::delete() comments"
jenkins-bot [Mon, 31 Aug 2015 23:01:26 +0000 (23:01 +0000)]
Merge "Allow to export null values"
Stanislav Malyshev [Mon, 31 Aug 2015 22:45:06 +0000 (15:45 -0700)]
Allow to export null values
Bug: T110960
Change-Id: Idd90c8b6370547b1d9caa8b93e3d5b27942418d8
jenkins-bot [Mon, 31 Aug 2015 21:20:16 +0000 (21:20 +0000)]
Merge "resourceloader: Improve moduleRegistry documentation"
jenkins-bot [Mon, 31 Aug 2015 21:10:50 +0000 (21:10 +0000)]
Merge "resourceloader: Remove global startUp() callback"
jenkins-bot [Mon, 31 Aug 2015 21:01:26 +0000 (21:01 +0000)]
Merge "objectcache: Correct tempTTL comment for getWithSetCallback()"
Aaron Schulz [Mon, 31 Aug 2015 20:35:09 +0000 (13:35 -0700)]
objectcache: Correct tempTTL comment for getWithSetCallback()
* This applies regardless of lockTSE
Change-Id: I5adb62fbb956407cd978211a0497e8bd9de3c1a5
Aaron Schulz [Mon, 31 Aug 2015 20:26:22 +0000 (13:26 -0700)]
Clarified WANObjectCache::delete() comments
Change-Id: I8c1cffa284cf92461045f9163ff1d1e9bf8c2244
Aaron Schulz [Wed, 26 Aug 2015 23:13:49 +0000 (16:13 -0700)]
Removed READ_LATEST from revision load in getMsgFromNamespace()
* The key salting of the invidial keys via the WAN cache makes
it OK to use slaves here. The salting happens via the delete()
in the MessageCache::replace() call, which is hit when the
message page is edited.
Bug: T92357
Change-Id: Ic191f6c3fc49c4c58461d26468dd0fa94a52051e
Translation updater bot [Mon, 31 Aug 2015 20:05:26 +0000 (22:05 +0200)]
Localisation updates from https://translatewiki.net.
Change-Id: I734ddd0b235fdfc0983bad6683f8e90367862177
jenkins-bot [Mon, 31 Aug 2015 19:56:11 +0000 (19:56 +0000)]
Merge "Changing user and user talk namespace name for Latvian wikis"
jenkins-bot [Mon, 31 Aug 2015 19:30:44 +0000 (19:30 +0000)]
Merge "Don't free the parser in the catch() block of parse()"
Aaron Schulz [Wed, 26 Aug 2015 19:19:50 +0000 (12:19 -0700)]
Fixed and suppressed some thumb.php IDEA errors
Change-Id: If5ec6fd1d6d272633296b74d2d8e146feb924229
Timo Tijhof [Fri, 28 Aug 2015 22:54:27 +0000 (00:54 +0200)]
resourceloader: Improve moduleRegistry documentation
Rename local 'msgs' variable to 'messages' for consistency.
Change-Id: I5d035204ac5d22673a9b9a68ce9ea6cf23aaeba8
Timo Tijhof [Fri, 28 Aug 2015 22:20:53 +0000 (00:20 +0200)]
resourceloader: Remove global startUp() callback
Instead of a hardcoded JSONP-style global callback, use script
load handlers. This uses the same logic as jQuery uses internally
(ajax/script.js), which we call inside mediawiki.js. Inline those
two lines here since this the code is what loads jQuery itself.
Bug: T39894
Change-Id: I02b0d16a6ec3081dc551bfd9c3ae652707b7edcf
Brad Jorsch [Mon, 31 Aug 2015 14:32:06 +0000 (10:32 -0400)]
Comment typo fix
Change-Id: I52e36f1ecbef8e8c18066d3dde84bda086ef9808
jenkins-bot [Mon, 31 Aug 2015 13:02:23 +0000 (13:02 +0000)]
Merge "Rephrase tooltip-t-contributions for better grammar"
maurelio [Mon, 31 Aug 2015 12:21:25 +0000 (14:21 +0200)]
Changing user and user talk namespace name for Latvian wikis
Added too GENDER support and kept the old NS_USER and NS_USER_TALK values
as $namespacealiases
Bug: T106513
Change-Id: Ibabbf284c5b5fef982d332d3dd5fd06fa76a69dd
Kunal Mehta [Mon, 31 Aug 2015 06:42:56 +0000 (23:42 -0700)]
Skip monolog tests if monolog is not installed
Change-Id: Ib3f4f76651da183901634a0ac6787666e47d74d4
jenkins-bot [Mon, 31 Aug 2015 00:28:44 +0000 (00:28 +0000)]
Merge "Workaround fopen lack of SubjectAltName support for instantCommons"
Translation updater bot [Sun, 30 Aug 2015 18:56:08 +0000 (20:56 +0200)]
Localisation updates from https://translatewiki.net.
Change-Id: I234526fd75bd75e927c3891988f82081dd05131e
jenkins-bot [Sun, 30 Aug 2015 07:35:34 +0000 (07:35 +0000)]
Merge "Improve wording of "size-bytes" and "size-pixel" messages"
This, that and the other [Sun, 30 Aug 2015 05:23:13 +0000 (15:23 +1000)]
Improve wording of "size-bytes" and "size-pixel" messages
"B" and "P" are vanishingly rare abbreviations for "bytes" and "pixels"
respectively. Let's use the full English terms for these, combined with
a PLURAL magic word for good measure.
Change-Id: Id59c4b9dea2c13940ae790b6a236ac08abe0a768
Brian Wolff [Wed, 1 Jul 2015 08:15:39 +0000 (02:15 -0600)]
Workaround fopen lack of SubjectAltName support for instantCommons
Hacky work around, where on php < 5.6.0 where SubjectAltName is not
supported, if a request to commons.wikimedia.org fails, retry the
request but telling php fopen wrapper to treat it as if it came
from en.wikipedia.org for validation purposes, since as of
c02fab71422a that's what the common name of the cert will be.
In the ideal world, everyone would just have curl installed.
I know this is super hacky, but I'd really like instant commons to
work out of the box even without curl installed.
Note: I'm basing the php 5.6 part on documentation, I have not tested
this with a copy of that version of php.
Bug: T75199
Change-Id: Ibde59be61a5b3d7cd5397ba352dce9be11e1b06f
Translation updater bot [Sat, 29 Aug 2015 19:33:15 +0000 (21:33 +0200)]
Localisation updates from https://translatewiki.net.
Change-Id: I11090b4bd770e52ee751b31b3788189e8d40c3f4
Bartosz Dziewoński [Sat, 29 Aug 2015 18:44:08 +0000 (20:44 +0200)]
mediawiki.ForeignApi: Use `type === 'POST'` instead of `type !== 'GET'`
Works better when 'type' is not set (which can happen if someone is
calling #ajax directly rather than #get or #post), is consistent
with code in mediawiki.api, and better reflects what we really mean.
Change-Id: Iea265be5c0e6113e91e700124ac36a0b588ca25c
addshore [Sat, 29 Aug 2015 12:08:44 +0000 (14:08 +0200)]
Add protected getCurrentRow to TablePager
This makes unit testing TablePagers easier
Change-Id: I899ec6b056fb1f007a6caf0be320324cffe08326
Kunal Mehta [Sat, 29 Aug 2015 04:39:15 +0000 (21:39 -0700)]
Really actually fix the typo in Parser.php
Change-Id: I9d9c6f13095087ac2c2c6693c6bd1613219bf658
Aaron Schulz [Sat, 29 Aug 2015 00:06:45 +0000 (17:06 -0700)]
Fixed parser report typo
Change-Id: Ia549f4e1932bc1196e840e154b8d6fb0b608d10d
jenkins-bot [Sat, 29 Aug 2015 00:01:58 +0000 (00:01 +0000)]
Merge "Improved getMainStashInstance() docs"
Aaron Schulz [Thu, 27 Aug 2015 09:10:03 +0000 (02:10 -0700)]
Made MessageCache use the WAN cache for individual cache keys
* These keys are used for larger messages not in the main blob,
and they still use explicit purges as the pages change
Change-Id: I254ddcc9ffe7c49654788d6aabcca2ab7ed4f03f
Aaron Schulz [Fri, 28 Aug 2015 23:24:13 +0000 (16:24 -0700)]
Improved getMainStashInstance() docs
Change-Id: I49e5c112478927387244214917fb06907c22186f
Translation updater bot [Fri, 28 Aug 2015 19:59:26 +0000 (19:59 +0000)]
Merge "Localisation updates from https://translatewiki.net."
jenkins-bot [Fri, 28 Aug 2015 19:35:20 +0000 (19:35 +0000)]
Merge "Revert "Prevent the jumbling up of autocomments in RTL""
Translation updater bot [Fri, 28 Aug 2015 19:33:53 +0000 (21:33 +0200)]
Localisation updates from https://translatewiki.net.
Change-Id: I68d4a05bd1404393686d145ba5f46f93f789d52a
TheDJ [Fri, 28 Aug 2015 18:55:38 +0000 (18:55 +0000)]
Revert "Prevent the jumbling up of autocomments in RTL"
This reverts commit
6f8c3359ed6e531edd9eed81dd033514feb6a802.
which was causing page load issues due to a bug in Chrome and Safari.
Bug: T110716
Change-Id: I0ab1bf1f52474d8ab43d34628df5d972e817e8f0
Kunal Mehta [Fri, 28 Aug 2015 17:40:08 +0000 (10:40 -0700)]
CssContentHandler: Fix @return documentation
Missed this when copying the code from JavaScriptContentHandler...
Change-Id: I6b232e431b4e997b65a698de42473c9b6deae6b1
jenkins-bot [Fri, 28 Aug 2015 17:17:36 +0000 (17:17 +0000)]
Merge "API: Add ApiResult::META_KVP_MERGE"
Brad Jorsch [Fri, 28 Aug 2015 15:10:20 +0000 (11:10 -0400)]
API: Add ApiResult::META_KVP_MERGE
This allows for merging the KVP key into the value for the alternative
output format. Specifically,
{ "key": { "foo": "bar" } }
can now be turned into
[{ "name": "key", "foo": "bar" }]
instead of
[{ "name": "key", "value": { "foo": "bar" } }]
Change-Id: Ie1f9235893dbbcd2948c46e0356360b5635a3ddd
C. Scott Ananian [Fri, 28 Aug 2015 14:23:47 +0000 (10:23 -0400)]
Be a little more conservative with the IPv6 support in the Sanitizer.
This isn't strictly necessary, but it matches the production used in
Parsoid better.
Change-Id: I542b1f71fe9e917fef5718a6df8ebb98735f7fb4
jenkins-bot [Fri, 28 Aug 2015 14:00:22 +0000 (14:00 +0000)]
Merge "Add comments clarifying how calling getters on ParserOutput affects the parser cache."
jenkins-bot [Fri, 28 Aug 2015 09:39:33 +0000 (09:39 +0000)]
Merge "Fixed some doc errors in tryNormaliseRedirect()"
jenkins-bot [Fri, 28 Aug 2015 08:31:22 +0000 (08:31 +0000)]
Merge "Added jobqueue.pickup_root_age metric"
Aaron Schulz [Mon, 24 Aug 2015 20:00:23 +0000 (13:00 -0700)]
Fixed some doc errors in tryNormaliseRedirect()
Change-Id: I8f9397d05de1c0bae33497d1f9e3146939599380
jenkins-bot [Fri, 28 Aug 2015 08:23:02 +0000 (08:23 +0000)]
Merge "api: Load mediawiki.apipretty with addModuleStyles"
daniel [Wed, 26 Aug 2015 16:17:05 +0000 (18:17 +0200)]
Add comments clarifying how calling getters on ParserOutput affects the parser cache.
Change-Id: Ie6a4a8989df218f88cfc7c1cf1d4b68cf871abbb
jenkins-bot [Fri, 28 Aug 2015 07:08:01 +0000 (07:08 +0000)]
Merge "SpecialVersion: Force text credits and license to LTR direction"
jenkins-bot [Fri, 28 Aug 2015 07:05:56 +0000 (07:05 +0000)]
Merge "Add date format for Latvian language"
Amir E. Aharoni [Thu, 27 Aug 2015 21:23:18 +0000 (00:23 +0300)]
SpecialVersion: Force text credits and license to LTR direction
It's safe to assume that text files with credits or licenses are written
in English, so it should be explicitly set to English/LTR
so that English won't appear in RTL wikis as RTL.
Change-Id: Ic9db006b2d3509de32f01ae04782535ab0183993
Aaron Schulz [Tue, 25 Aug 2015 08:17:43 +0000 (01:17 -0700)]
Added --status and --limit filters to showJobs.php
Change-Id: Ibb3e907061c04a4dd2af23789fd59685240f0876
jenkins-bot [Fri, 28 Aug 2015 04:35:59 +0000 (04:35 +0000)]
Merge "Add ParserOutput cache and expiry times to NewPP report"
jenkins-bot [Fri, 28 Aug 2015 03:14:27 +0000 (03:14 +0000)]
Merge "Use Parsoid v3 API; emulate the RESTBase v1 API"
jenkins-bot [Fri, 28 Aug 2015 02:18:43 +0000 (02:18 +0000)]
Merge "Removed useless $wgUseLocalMessageCache check in MessageCache"
Ori Livneh [Fri, 28 Aug 2015 02:05:27 +0000 (19:05 -0700)]
Add ParserOutput cache and expiry times to NewPP report
The labels are not localized, because I think this ought to be outputted as a
JSON blob, with uniform field names. But not doing that in this patch.
Change-Id: I235839b276632308ddeac7afe763d355b73c2a25
This, that and the other [Fri, 28 Aug 2015 01:36:48 +0000 (01:36 +0000)]
Merge "Update namespace names for Sanskrit"
Aaron Schulz [Fri, 28 Aug 2015 01:31:35 +0000 (18:31 -0700)]
Removed useless $wgUseLocalMessageCache check in MessageCache
* The localCache member will use CACHE_NONE if this is off,
so the set()/get() calls are already no-oped
* Also allow using the local cache if $hash is false
in "stale mode", which handles the case where the hash
key fell out of memory for some reason more gracefully
Change-Id: Ie12efcd4088a6dc4a4cdd2fd06646f2881df53d7
jenkins-bot [Thu, 27 Aug 2015 23:33:07 +0000 (23:33 +0000)]
Merge "Make ParserOptions fields private"
Brad Jorsch [Thu, 27 Aug 2015 13:38:53 +0000 (09:38 -0400)]
Make ParserOptions fields private
Now that Ib58e8020 and Ie644854 are merged, there are no direct accesses
to these fields left in core or extensions in Gerrit. So let's make them
private to make sure no more get added before we decide whether we're
going to eliminate them entirely.
Bug: T110269
Change-Id: I82041b88dd0f194716f54d3207649388328805ca
Kunal Mehta [Thu, 27 Aug 2015 23:13:42 +0000 (16:13 -0700)]
registration: Remove unused 'wgExtraNamespaces' merge strategy
As of
9df0672255afd, this is no longer used.
Change-Id: I7e40096cb5173626ba7f2cd3d2be121c8a04768a
C. Scott Ananian [Wed, 27 May 2015 21:46:45 +0000 (17:46 -0400)]
Use Parsoid v3 API; emulate the RESTBase v1 API
Update the ParsoidVirtualRESTService and the
RestbaseVirtualRESTService to use Parsoid's v3 API, instead of the
deprecated v1/v2 APIs. Since Visual Editor still issues requests
using the Parsoid v1 API, convert Parsoid v1 API requests into Parsoid
v3 API requests when needed for a smooth transition. We also add
support for converting RESTBase v1 API requests to Parsoid v3 API
requests.
The next step will be to convert Visual Editor to issue RESTBase v1
API requests (https://gerrit.wikimedia.org/r/217995), and then the
Parsoid v1 conversion code added here can be removed (T100681).
Tested Parsoid v1->v3 conversion, Parsoid v1->RESTBase conversion,
plus Parsoid v3 and RESTBase v1->Parsoid v3 conversion using VE
patched to issue RESTBase v1 API requests.
Bug: T100681
Change-Id: I07ac60cdec7a52ef93187d40099325a069e3239a
jenkins-bot [Thu, 27 Aug 2015 21:10:22 +0000 (21:10 +0000)]
Merge "XMP: log structured data for halting errors"
Mukunda Modell [Thu, 27 Aug 2015 19:28:51 +0000 (14:28 -0500)]
Don't free the parser in the catch() block of parse()
This is the solution proposed by @bd808 to address T89532, also, it's
completely untested.
Bug: T89532
Change-Id: I9b5b47eb9d64a3d80f9c49618a362415d15071b4
(cherry picked from commit
d34ed1acbe2ea58e5de59d3a18e278be0e384f8f)
Timo Tijhof [Wed, 26 Aug 2015 17:37:54 +0000 (19:37 +0200)]
api: Load mediawiki.apipretty with addModuleStyles
Follows-up
d25cb99. addModules() is async and causes a flash
of unstyled content. E.g. the h1.firstHeading renders and then
disappears on pages like:
<https://en.wikipedia.org/w/api.php?action=purge>
Change-Id: I74c6386d2c5f318987dac103a76c0bc2af6e2653
Bryan Davis [Thu, 27 Aug 2015 19:56:57 +0000 (13:56 -0600)]
XMP: log structured data for halting errors
Log structured data for halting errors in XMPReader::parse at warning
level.
Bug: T89532
Change-Id: Iba38fedb4255687742228cf656218b4b633a7f40
jenkins-bot [Thu, 27 Aug 2015 19:25:11 +0000 (19:25 +0000)]
Merge "Fix spaces in MigrateFileRepoLayoutTest::tearDown"
Translation updater bot [Thu, 27 Aug 2015 19:16:32 +0000 (21:16 +0200)]
Localisation updates from https://translatewiki.net.
Change-Id: I5d73ea5a5527d5ba495d9fba466b7f418288dff9
umherirrender [Thu, 27 Aug 2015 19:03:32 +0000 (21:03 +0200)]
Fix spaces in MigrateFileRepoLayoutTest::tearDown
Use tabs for indent
Change-Id: I021b7ff41be23e46fe49dc552e697c7b8f78622e
Aaron Schulz [Thu, 27 Aug 2015 09:20:54 +0000 (02:20 -0700)]
Cleaned up int|INF doc type to avoid IDEA errors
* INF is not a class (technically a float, which would look confusing here)
Change-Id: I9c437ecd0f2fe1a2a3e15197d50829a638f3073a
Southparkfan [Thu, 27 Aug 2015 16:13:37 +0000 (18:13 +0200)]
Enable autocompletion for Special:ComparePages
Bug: T26235
Change-Id: Ia55ff14ceeef4eb49facd42a71c6330dd3384eb8
Ricordisamoa [Thu, 27 Aug 2015 14:19:20 +0000 (16:19 +0200)]
Add space between classes in HTMLButtonField
Classes such as 'wb-input-buttonmw-ui-button' were being generated
when $wgUseMediaWikiUIEverywhere = true, see:
https://integration.wikimedia.org/ci/job/mwext-Wikibase-repo-tests-mysql-hhvm/4217/console
Change-Id: If40b712ffb8e31d27380bf481333eecb7b95f9de
jenkins-bot [Thu, 27 Aug 2015 17:54:35 +0000 (17:54 +0000)]
Merge "Revert "Load all CSS in the top queue""
Ori.livneh [Thu, 27 Aug 2015 17:02:11 +0000 (17:02 +0000)]
Revert "Load all CSS in the top queue"
I don't want to roll this out until our WebPageTest infrastructure is ready
and can show whether or not this is beneficial. I should have indicated this
clearly on the change rather than leave it open, sorry.
This reverts commit
b7c0e537ebb7216cd12a6c48fbaee7c2c7757c55.
Change-Id: I43e3295873a94da1c2febe2c1a19ad20ca914254
jenkins-bot [Thu, 27 Aug 2015 14:16:45 +0000 (14:16 +0000)]
Merge "Load User via READ_LATEST in ApiOptions to avoid CAS errors"
jenkins-bot [Thu, 27 Aug 2015 09:48:15 +0000 (09:48 +0000)]
Merge "Load all CSS in the top queue"
jenkins-bot [Thu, 27 Aug 2015 09:40:35 +0000 (09:40 +0000)]
Merge "Move job running till after deferred updates"
jenkins-bot [Thu, 27 Aug 2015 09:38:27 +0000 (09:38 +0000)]
Merge "Ignore DBPerformance thresholds in Special:ConfirmEmail"
jenkins-bot [Thu, 27 Aug 2015 09:36:34 +0000 (09:36 +0000)]
Merge "Moved onArticleEdit() lines around for clarity"
jenkins-bot [Thu, 27 Aug 2015 08:09:41 +0000 (08:09 +0000)]
Merge "Add @throws tag to Database::insert"
jenkins-bot [Thu, 27 Aug 2015 08:08:08 +0000 (08:08 +0000)]
Merge "Fixed headers sent in addMissingMetadata()"
addshore [Thu, 27 Aug 2015 07:55:53 +0000 (09:55 +0200)]
Add @throws tag to Database::insert
This was already documented but not with @throws
Change-Id: Ib55c2efb9993c5d939fab1ed5d6ade3fc277902f
Kunal Mehta [Wed, 26 Aug 2015 21:18:39 +0000 (14:18 -0700)]
registration: Fix namespaces added through the ExtensionProcessor
Using $wgExtraNamespaces overrides any localized namespaces with the
canonical form, which is not ideal.
Namespaces added through extension.json will now store the canonical
form and numerical id in a 'ExtensionNamespaces' attribute that is read
by MWNamespace::getCanonicalNamespaces().
Also fix the documentation on $wgExtraNamespaces, as using
$wgCanonicalNamespaceNames has not been possible since r85327.
Bug: T110389
Change-Id: I5bd9a7258f59d8c4a7ad0543d2115960fbea9b3a
Aaron Schulz [Thu, 27 Aug 2015 01:40:43 +0000 (18:40 -0700)]
Ignore DBPerformance thresholds in Special:ConfirmEmail
Bug: T92357
Change-Id: I81a1a31de522c2390e9d4d515677731cce4fcf1f
jenkins-bot [Thu, 27 Aug 2015 01:17:50 +0000 (01:17 +0000)]
Merge "Avoid using DB_MASTER in SpecialEditWatchlist::getWatchlist()"
jenkins-bot [Thu, 27 Aug 2015 01:11:51 +0000 (01:11 +0000)]
Merge "Defer spreadAnyEditBlock() post-send"
Aaron Schulz [Wed, 26 Aug 2015 23:54:23 +0000 (16:54 -0700)]
Avoid using DB_MASTER in SpecialEditWatchlist::getWatchlist()
* Only use it on POST requests, where the submit callback really
wants the latest data. The form data will be lagged by user
delay anyway, so using the slave there does not change much.
ChronologyProtector already handles the user seeing their
*own* changes for consecutive updates.
Bug: T92357
Change-Id: I50274ad5f67f6445a89c9d8d6f01d3fca1e9002b
Aaron Schulz [Thu, 27 Aug 2015 00:54:58 +0000 (17:54 -0700)]
Defer spreadAnyEditBlock() post-send
* No need for these to be perfectly synchronous
Bug: T92357
Change-Id: I6863960c8170ef4c3f2e356841b4086ac98f0f03
jenkins-bot [Thu, 27 Aug 2015 00:09:50 +0000 (00:09 +0000)]
Merge "Removed deprecated $wgMasterWaitTimeout; not used by LoadBalancer"