jenkins-bot [Fri, 1 Mar 2019 22:03:26 +0000 (22:03 +0000)]
Merge "Simplify strings in PHP code"
jenkins-bot [Fri, 1 Mar 2019 21:51:24 +0000 (21:51 +0000)]
Merge "maintenance: Replace custom .ext extraction with pathinfo()"
jenkins-bot [Fri, 1 Mar 2019 21:17:55 +0000 (21:17 +0000)]
Merge "Add mediawiki.interface.helpers.styles where it is needed"
Fomafix [Fri, 1 Mar 2019 21:15:22 +0000 (22:15 +0100)]
Simplify strings in PHP code
Change-Id: I481810ade68b0c5a5be21d22e2a107646d5813e6
jenkins-bot [Fri, 1 Mar 2019 20:25:21 +0000 (20:25 +0000)]
Merge "Add UserSendConfirmationMail hook"
jenkins-bot [Fri, 1 Mar 2019 19:31:55 +0000 (19:31 +0000)]
Merge "Hygiene: use null-coalescing operator when operating on tags"
jdlrobson [Thu, 28 Feb 2019 21:12:26 +0000 (13:12 -0800)]
Add mediawiki.interface.helpers.styles where it is needed
Bug: T212613
Change-Id: I7d5920c3464838f68dc1169779e2f2662a3e336e
Piotr Miazga [Fri, 1 Mar 2019 18:13:28 +0000 (19:13 +0100)]
Hygiene: use null-coalescing operator when operating on tags
Change-Id: I7d5b4e05ee6c010b4f9438bc4871a3369b1871a6
Thiemo Kreuz [Tue, 26 Feb 2019 12:57:24 +0000 (13:57 +0100)]
Replace lowercase {@inheritdoc} with @inheritDoc
According to the coding standards we even enforce with a custom PHPCS sniff.
It currently does not pick these mistakes up because of the curly brackets.
I'm not sure if this is worth an update of the PHPCS sniff. I wanted to
suggest this fix anyway.
Change-Id: I9041ea7a00baf7f55e0ff0e56879a89fb74bb479
Thiemo Kreuz [Thu, 28 Feb 2019 08:55:50 +0000 (09:55 +0100)]
maintenance: Add missing limit parameters to some explode()
This is, in theory, a loophole that can not only cause such code to
consume suprising amounts of memory and runtime. It can also create
suprising results. For example, an input like
-param="might contain a = char"
might result in a cut-off value.
Not so much of a problem in a maintenance script. But still good
practice, I find.
Change-Id: I14fb278e6fdb61d0c486ca7e23229851ea479408
jenkins-bot [Fri, 1 Mar 2019 03:00:21 +0000 (03:00 +0000)]
Merge "rdbms: document Database::doQuery() return value"
jenkins-bot [Fri, 1 Mar 2019 02:58:34 +0000 (02:58 +0000)]
Merge "Avoid use of deprecated ObjectCache::getMainWANInstance method"
jenkins-bot [Fri, 1 Mar 2019 02:50:11 +0000 (02:50 +0000)]
Merge "obectcache: split out some WAN cache refresh logic into scheduleAsyncRefresh()"
Aaron Schulz [Thu, 28 Feb 2019 20:42:20 +0000 (12:42 -0800)]
Avoid use of deprecated ObjectCache::getMainWANInstance method
Change-Id: I429e68c8642a00afa812f5ae900c1fb4f8308528
jenkins-bot [Fri, 1 Mar 2019 02:37:27 +0000 (02:37 +0000)]
Merge "Remove comments literally repeating the next line of code"
Aaron Schulz [Fri, 1 Mar 2019 02:21:49 +0000 (18:21 -0800)]
rdbms: document Database::doQuery() return value
Removed random @throws tag from a subclass while at it
Change-Id: I1cad1f66b62bb4a306feb5c773ed5556891f82a7
jenkins-bot [Fri, 1 Mar 2019 01:25:03 +0000 (01:25 +0000)]
Merge "Fix is_array check in ResourceLoaderImageModule"
Volker E [Fri, 1 Mar 2019 00:17:02 +0000 (16:17 -0800)]
RCFilters: Make SaveFiltersPopup header's label and icon wrap
Wrapping SaveFiltersPopup header's label and icon in languages like
Catalan and German.
Also re-positioning close button from absolute top and remove duplicated,
inherited properties.
Change-Id: I3fd35b237f82ab95b66e6e710cba36cf1887618f
Translation updater bot [Thu, 28 Feb 2019 21:04:55 +0000 (22:04 +0100)]
Localisation updates from https://translatewiki.net.
Change-Id: Ibd3587a45d063cdd7f7276bd15319f1067819eab
Umherirrender [Thu, 28 Feb 2019 20:06:53 +0000 (21:06 +0100)]
Fix is_array check in ResourceLoaderImageModule
Checking the variable used as expression in the foreach seems wrong.
Also the var is already checked before, so this is the wrong var name
Change-Id: Ia2a7352422ace1e78949a2009b0ad1f16681fc74
jenkins-bot [Thu, 28 Feb 2019 14:25:08 +0000 (14:25 +0000)]
Merge "DateTimeFormatter: Use mw.RegExp.escape from 'mediawiki.RegExp'"
jenkins-bot [Thu, 28 Feb 2019 14:25:03 +0000 (14:25 +0000)]
Merge "Preferences: Remove href="#" from stub threshold sample link"
Fomafix [Mon, 18 Feb 2019 09:03:56 +0000 (10:03 +0100)]
Preferences: Remove href="#" from stub threshold sample link
* Prevents the sample link from being focusable and selectable via
keyboard navigation.
* Prevents the sample link from being clicked, which would switch
to the first preferences tab and scroll to the top, unhelpfully.
* Since
4f8967672d an <a> without an href also has cursor:pointer on
hover.
Change-Id: I679e277d17dc679177c29d557806c9a9e91c6703
Fomafix [Thu, 28 Feb 2019 13:16:17 +0000 (14:16 +0100)]
Add missing dependency on module 'mediawiki.RegExp'
moment-locale-overrides.js uses mw.RegExp.escape() from module
'mediawiki.RegExp'.
Change-Id: Iac669355ca1c763a34b28badc2bac3fd08e57db9
jenkins-bot [Thu, 28 Feb 2019 12:31:31 +0000 (12:31 +0000)]
Merge "mw-ui-radio, mw-ui-checkbox: Don't reset every font style"
Thiemo Kreuz [Thu, 28 Feb 2019 10:47:07 +0000 (11:47 +0100)]
maintenance: Replace custom .ext extraction with pathinfo()
Note the custom code was not working correctly for filenames that do
not contain a dot. I assume this is not relevant for the filenames this
code needs to process.
Change-Id: Id4578bc86203dcbd9f76928bbec71a2c46176e11
Fomafix [Thu, 28 Feb 2019 08:49:33 +0000 (09:49 +0100)]
DateTimeFormatter: Use mw.RegExp.escape from 'mediawiki.RegExp'
Change-Id: I7aa0264f58d7f79fd80da2a3bb37caa767d78474
jenkins-bot [Thu, 28 Feb 2019 08:42:04 +0000 (08:42 +0000)]
Merge "maintenance: Remove arguments that match their default values"
Santhosh Thottingal [Tue, 26 Feb 2019 12:02:00 +0000 (17:32 +0530)]
mw-ui-radio, mw-ui-checkbox: Don't reset every font style
The following existing style:
.client-js .mw-ui-radio:not( #noop ) {
* {
font: inherit;
}
}
is a strong reset, preventing any font style customization such
as a bold label for .mw-ui-radio.
Change it only to font-size as required for the original reason
for override. Same for .mw-ui-checkbox
I am not sure whether this override is required now, but not going
in to that now.
Change-Id: I5d1815bcfa897bdb03c81be11f63e1acb9412c10
Aaron Schulz [Thu, 28 Feb 2019 01:13:40 +0000 (17:13 -0800)]
obectcache: split out some WAN cache refresh logic into scheduleAsyncRefresh()
Change-Id: If5f895a0d38cd0d6861125000df4fcfc66f1eeb3
Roan Kattouw [Wed, 27 Feb 2019 02:16:49 +0000 (18:16 -0800)]
Add UserSendConfirmationMail hook
Allow extensions to modify the confirmation email.
Bug: T215665
Change-Id: I4bcf76699a5114292fc19085fe441de8b898a8d3
jenkins-bot [Wed, 27 Feb 2019 22:43:54 +0000 (22:43 +0000)]
Merge "API: Use log context for api-feature-usage log"
Brad Jorsch [Wed, 27 Feb 2019 20:51:38 +0000 (15:51 -0500)]
API: Use log context for api-feature-usage log
The text message is deprecated.
Bug: T217162
Change-Id: Ie891257140ea19369e10b2e91463a1fb4aa5d233
jenkins-bot [Wed, 27 Feb 2019 21:22:45 +0000 (21:22 +0000)]
Merge "objectcache: simplify WAN cache unwrap() method by removing FLG_STALE"
Derick Alangi [Mon, 25 Feb 2019 17:11:57 +0000 (18:11 +0100)]
maintenance: Remove arguments that match their default values
Calls to methods with arguments as default values can safely be
removed as there is an exact match. With or without them, code
will still work as intended.
Change-Id: Id8141626a51d9cc8bf8bc0baeb5e38d120664da5
jenkins-bot [Wed, 27 Feb 2019 19:10:08 +0000 (19:10 +0000)]
Merge "Use splat operator in HamcrestPHPUnitIntegration"
Aaron Schulz [Tue, 26 Feb 2019 22:09:52 +0000 (14:09 -0800)]
objectcache: simplify WAN cache unwrap() method by removing FLG_STALE
Instead, make the "high read lag with lockTSE" case just lower FLD_TTL.
This also avoids constant regenerations by threads getting the mutex.
Added logging for TTL adjustments in set() for the lockTSE case.
Also remove some delete() calls from tests that were not needed.
Change-Id: Id7695f0377235e4a2f6e0efc88e870c8a990c3b0
jenkins-bot [Wed, 27 Feb 2019 17:46:53 +0000 (17:46 +0000)]
Merge "redis: Fix incomplete @var documentation in RedisConnRef"
Thiemo Kreuz [Wed, 27 Feb 2019 15:29:35 +0000 (16:29 +0100)]
Remove comments literally repeating the next line of code
I would argue that these comments do not add any information that
would not be there already. Having them adds mental overhead, because
one needs to read both the comment and the next line of code first to
understand they say the exact same. I don't find this helpful, but
more distracting.
Change-Id: I39c98f25225947ebffdcc2fd8f0243e7a6c070d7
jenkins-bot [Wed, 27 Feb 2019 17:25:09 +0000 (17:25 +0000)]
Merge "Add tests to ensure that retrieved actions match passed in restrictions"
David Barratt [Wed, 20 Feb 2019 03:18:14 +0000 (22:18 -0500)]
Add tests to ensure that retrieved actions match passed in restrictions
This is a theoretical issue where a passed in restriction does not match the
retrieved actions. There are no actions like this in MediaWiki core or known
extensions, but since the possibility exists, tests should exist to prevent
the issue before it happens.
Bug: T213220
Change-Id: I58da016768a3ee958baa2a25b8177a9e667fa955
Elliott Eggleston [Wed, 27 Feb 2019 15:48:58 +0000 (10:48 -0500)]
Fix warning on QUnit modules without dependencies
Check if $module['dependencies'] is set before checking whether
it's a string.
Change-Id: I9ece83d4ec92d0d3a41a1c6863492e646f011eab
Thiemo Kreuz [Wed, 27 Feb 2019 15:33:55 +0000 (16:33 +0100)]
redis: Fix incomplete @var documentation in RedisConnRef
Change-Id: Id6cf3731986a004ffa80611fcf02c43fd6609ff1
jenkins-bot [Wed, 27 Feb 2019 14:28:01 +0000 (14:28 +0000)]
Merge "Document type for class properties in ResourceLoaderImageModule"
jenkins-bot [Wed, 27 Feb 2019 04:53:16 +0000 (04:53 +0000)]
Merge "Fix read-only mode for LoadBalancerSingle"
Kunal Mehta [Wed, 27 Feb 2019 02:52:03 +0000 (18:52 -0800)]
Use splat operator in HamcrestPHPUnitIntegration
Change-Id: I1421748fd338137dd6990775e0d79a6cfb82aa9d
jenkins-bot [Wed, 27 Feb 2019 01:50:37 +0000 (01:50 +0000)]
Merge "mw.ui forms: Style individual form field warnings, as well as errors"
Roan Kattouw [Wed, 27 Feb 2019 01:24:10 +0000 (17:24 -0800)]
mw.ui forms: Style individual form field warnings, as well as errors
If you put a <span class="error"> inside a vform-field, it would display
correctly, but a <span class="warning"> wouldn't. Make it work for both.
Change-Id: I9e582ade2fca87782aa7be38774fa4943dc196da
jenkins-bot [Tue, 26 Feb 2019 23:22:59 +0000 (23:22 +0000)]
Merge "Remove unnecessary variable from UserTest"
jenkins-bot [Tue, 26 Feb 2019 22:44:51 +0000 (22:44 +0000)]
Merge "resourceloader: Add getScript() method for loading scripts with callback"
Thalia [Tue, 26 Feb 2019 22:15:31 +0000 (22:15 +0000)]
Remove unnecessary variable from UserTest
Change-Id: I2aaae36cc3102b309aa8f5ab05f512c4fbf9810c
Aaron Schulz [Wed, 23 Jan 2019 19:32:18 +0000 (11:32 -0800)]
rdbms: make safeWaitForMasterPos() handle master connection failure
Previously, in such cases where a master connection was needed
and could not be acquired, a fatal error would occur. Instead,
throw a DBReplicationWaitError exception.
Also removed a redundant documentation block for the method.
Change-Id: Ide3015d57673df0876daf265a0182fe41004052d
Petr Pchelko [Thu, 21 Feb 2019 02:09:41 +0000 (18:09 -0800)]
Emit new style API action logs into Monolog.
Bug: T214080
Depends-On: I454d3d12dcc1d7e525dc468df90424e68df59d78
Change-Id: Id0be845884cd834d89df897cc8a862904caeb09a
Amir Sarabadani [Tue, 26 Feb 2019 12:17:09 +0000 (13:17 +0100)]
Fix lowercase in namespace of LogCapturingSpi.php
Basically turning Mediawiki to MediaWiki as all other classes exist
in that namespace
Bug: T217128
Change-Id: I95039a1a54c9900a2f2414b5a6ccce5fb6e5e612
jenkins-bot [Tue, 26 Feb 2019 11:03:17 +0000 (11:03 +0000)]
Merge "selenium: improve ffmpeg reporting"
jenkins-bot [Tue, 26 Feb 2019 05:30:36 +0000 (05:30 +0000)]
Merge "DatabasePostgres: Ignore "IGNORE" option to update()"
Evad37 [Sat, 2 Feb 2019 05:42:00 +0000 (13:42 +0800)]
resourceloader: Add getScript() method for loading scripts with callback
Provides a friendly wrapper for loading scripts using $.ajax().
Returns a promise when the script dependency -- a single
script url -- has been loaded. This matches how mw.loader.load
is able to load a script url, and how mw.loader.using will
return a promised that is resolved when dependecies are loaded,
or rejected if there is an error.
Added as a separate function mw.loader.getScript(), rather
than adding this functionality to using(), as it is a separate
use case, less confusing for users, and there is no shared or
duplicated code between loading a script and loading one
or more named modules.
Bug: T27962
Change-Id: I13be426d03261a2d0c6a1631af94a9f9af58394b
jenkins-bot [Mon, 25 Feb 2019 22:00:27 +0000 (22:00 +0000)]
Merge "debug: Add basic accessibility support to debug console"
Translation updater bot [Mon, 25 Feb 2019 21:46:14 +0000 (22:46 +0100)]
Localisation updates from https://translatewiki.net.
Change-Id: Id54b521705d09e65558ed7ce20f67280c7eac13d
jenkins-bot [Mon, 25 Feb 2019 21:11:19 +0000 (21:11 +0000)]
Merge "Remove Target/User validation from Block::appliesToUsertalk()"
Umherirrender [Sat, 23 Feb 2019 20:35:15 +0000 (21:35 +0100)]
Document type for class properties in ResourceLoaderImageModule
Also declare missing properties
Change-Id: Ia63597fa996e7050b29ff6df0694dd7e036800d8
jenkins-bot [Mon, 25 Feb 2019 20:31:59 +0000 (20:31 +0000)]
Merge "Add Priority Hints support"
jenkins-bot [Mon, 25 Feb 2019 20:31:48 +0000 (20:31 +0000)]
Merge "Chrome Origin Trial support"
David Barratt [Mon, 25 Feb 2019 18:35:35 +0000 (13:35 -0500)]
Remove Target/User validation from Block::appliesToUsertalk()
This fixes a regression in I0e131696419211319082cb454f4f05297e55d22e where
an IP block that also blocks logged in users from editing, throws an exception.
The target/user validation in Block::appliesToUsertalk() does not work well
when an IP block prevents logged in users from editing. This case throws a
validation error, when in reality it should make a judgement.
This change means that the method will now trust whatever is passed into it as
the current user's talk page.
Bug: T211578
Change-Id: I3bb27cf7bec8421d31aa4de28a0e658365bb5bf2
Gilles Dubuc [Wed, 20 Feb 2019 14:06:36 +0000 (15:06 +0100)]
Add Priority Hints support
When the option is enabled, the first non-icon
thumbnail encountered has a "high" importance.
Bug: T216499
Change-Id: I8d3c1b3e2d136ba16bd7de4809ee3ca63ab462fe
Gilles Dubuc [Wed, 20 Feb 2019 08:49:28 +0000 (09:49 +0100)]
Chrome Origin Trial support
Bug: T216595
Change-Id: I74b8824a87e6f52a84d2f8a9708501fc05f73266
jenkins-bot [Mon, 25 Feb 2019 17:05:36 +0000 (17:05 +0000)]
Merge "objectcache: improve set() slam prevention in getWithSetCallback()"
jenkins-bot [Mon, 25 Feb 2019 16:40:50 +0000 (16:40 +0000)]
Merge "media: Mark ImageHandler::getScriptParams() as protected"
jenkins-bot [Mon, 25 Feb 2019 16:40:45 +0000 (16:40 +0000)]
Merge "media: Mark public MediaHandler/ImageHandler methods as such"
Antoine Musso [Thu, 21 Feb 2019 16:00:32 +0000 (17:00 +0100)]
selenium: improve ffmpeg reporting
1) stdout/stderr log improvement
Our wdio test suite records video of Selenium test session using ffmpeg,
its stdout/stderr are relayed via console.log() however the new line was
not stripped which is confusing and in case the buffer contains multiple
lines, only the first line had the 'ffmpeg stderr: ' prefix.
Ensure each ffmpeg line is prefixed by logging them individually.
2) exit code / signal
childProcess 'close' event is being passed the exit code and the signal.
We only handled the exit code, thus when ffmpeg is send a signal we
logged:
ffmpeg exited with code null
Handle the case when the childProcess is send a signal (typically
SIGINT).
Change-Id: I9f509690baa9b7981399a09448582f45c30800d7
Derick Alangi [Mon, 25 Feb 2019 12:28:37 +0000 (13:28 +0100)]
actions: Improve Action class code for clarity & consistency
~ Move local variable $obj to one liner as it's redundant in
this case.
~ Use === check instead of == as in order case above for the
purpose of consistency.
~ Make use of single quotes for strings literals instead of
double quotes.
Change-Id: I69650cacb0658a7369147ea5e9933fb5585e3376
Thiemo Kreuz [Mon, 25 Feb 2019 09:20:28 +0000 (10:20 +0100)]
media: Mark ImageHandler::getScriptParams() as protected
Proof this method is really not called from anywhere else, except the
ImageHandler class itself, and one subclass:
https://codesearch.wmflabs.org/search/?q=%5CSgetScriptParams&i=1
Change-Id: I57e75ce0cb7c9e52d545b8e8226610bcc6e06c44
Thiemo Kreuz [Mon, 25 Feb 2019 09:16:30 +0000 (10:16 +0100)]
media: Mark public MediaHandler/ImageHandler methods as such
Note I'm intentionally not touching the entire file, but only methods
I'm absolutely sure are already called from outside, e.g. from
MediaHandlerFactory, and must be public because of this.
I'm intentionally not doing anything with private or protected in this
patch, as such changes are much more fragile.
This is a direct follow up for the changes proposed in Iaa4f60d.
Change-Id: Ida817b289ddd5e9a8c162cc1fa3335c639a0bbe5
jenkins-bot [Mon, 25 Feb 2019 02:45:42 +0000 (02:45 +0000)]
Merge "Return the page_id in list=langbacklinks as an int"
Reedy [Mon, 25 Feb 2019 00:18:47 +0000 (00:18 +0000)]
Return the page_id in list=langbacklinks as an int
Bug: T216968
Change-Id: I5b16779be7b24b1e46d4787a82a8daa3611f67b1
Reedy [Mon, 25 Feb 2019 00:38:33 +0000 (00:38 +0000)]
Use (int) rather than intval()
Bug: T216969
Change-Id: I4c06716c83b69d128f26fa7f68736808aa2d3d64
jenkins-bot [Mon, 25 Feb 2019 00:31:59 +0000 (00:31 +0000)]
Merge "Return the page_id in list=iwbacklinks as an int rather than string"
setian [Sun, 24 Feb 2019 21:43:33 +0000 (16:43 -0500)]
Return the page_id in list=iwbacklinks as an int rather than string
Bug: T216968
Change-Id: I6645c5f1c6e76be3187c24053ed430e99c03bff4
jenkins-bot [Sun, 24 Feb 2019 22:49:26 +0000 (22:49 +0000)]
Merge "resourceloader: Make JSMinPlus allow reserved words as property name (ES5)"
Ebrahim Byagowi [Wed, 20 Feb 2019 23:41:08 +0000 (03:11 +0330)]
resourceloader: Make JSMinPlus allow reserved words as property name (ES5)
This makes parse of "var a = { true: 12 };" and "var a = { true: 12 };"
possible thus making JSMinPlus ES5 compatible.
Bug: T215280
Change-Id: I84bcacf26ab8900d09958b5d961cc40e1a046698
Translation updater bot [Sun, 24 Feb 2019 21:23:34 +0000 (22:23 +0100)]
Localisation updates from https://translatewiki.net.
Change-Id: Iaa33d306da2a1daf0f99f7432b542f7bdbd72eca
jenkins-bot [Sun, 24 Feb 2019 18:17:35 +0000 (18:17 +0000)]
Merge "specials: Cleanup and fix PHPDoc comments"
Derick Alangi [Sat, 23 Feb 2019 14:17:53 +0000 (15:17 +0100)]
specials: Cleanup and fix PHPDoc comments
Fix and cleanup code documentation in PHPDoc for clarity. This involves
correcting return types, parameter types and errors expected to be thrown
by methods within a class.
Change-Id: I30ba4b99d25460f84d348dd96480df67e92c69a6
jenkins-bot [Sun, 24 Feb 2019 00:38:34 +0000 (00:38 +0000)]
Merge "Upgrade moment from 2.19.3 to 2.24.0"
jenkins-bot [Sun, 24 Feb 2019 00:19:54 +0000 (00:19 +0000)]
Merge "Remove comments literally documenting unit tests being unit tests"
jenkins-bot [Sat, 23 Feb 2019 10:46:56 +0000 (10:46 +0000)]
Merge "Sanitizer: remove deprecated parameter to escapeIdReferenceList()"
Kunal Mehta [Sat, 23 Feb 2019 01:17:05 +0000 (17:17 -0800)]
registration: Fix 'ResourceLoaderWikiModule' schema validation
There is a full schema for ResourceModules when the
class: ResourceLoaderWikiModule is used. Except it doesn't actually work,
since it it fails validating against that schema, it will continue to check
the other anyOf schemas, find the arbitrary class one, and pass that one.
Blacklisting ResourceLoaderWikiModule from the arbitrary class schema means
that it has to pass the specified schema for it, or fail validation.
I verified that all extensions and skins in Gerrit pass validation with this
patch.
Change-Id: Icb862088c0ca1f89ed5a57a4286dd9049d213021
Kunal Mehta [Sat, 23 Feb 2019 01:03:01 +0000 (17:03 -0800)]
registration: Fix 'arbitrary ResourceLoaderModule definition' schema regex
We want this regex to match on everything that is not the literal string
'ResourceFileModule' or 'ResourceImageModule' (actually
'ResourceLoaderWikiModule' should be in this too, but I'll fix that in a
follow-up). But the current regex would not match if one of those was in a
substring.
The new regex should match everything that is *not* those two strings.
Bug: T216861
Change-Id: I7d888adfdc67b2075028e49432710308aa0125af
jenkins-bot [Fri, 22 Feb 2019 23:32:54 +0000 (23:32 +0000)]
Merge "resourceloader: Remove back compat for old packageFiles format"
James D. Forrester [Fri, 22 Feb 2019 10:34:40 +0000 (10:34 +0000)]
Upgrade moment from 2.19.3 to 2.24.0
Bug: T191339
Change-Id: I15034d88fabd9185f3eb5ffff98ee6f9922a7bc2
jenkins-bot [Fri, 22 Feb 2019 20:54:04 +0000 (20:54 +0000)]
Merge "Add .phar to $wgFileBlacklist as a paranoia measure"
jenkins-bot [Fri, 22 Feb 2019 19:33:48 +0000 (19:33 +0000)]
Merge "foreign-resources: Pull all moment locales, even the ones we don't yet use"
jenkins-bot [Fri, 22 Feb 2019 19:33:43 +0000 (19:33 +0000)]
Merge "foreign-resources: Add moment to our list of resources"
jenkins-bot [Fri, 22 Feb 2019 19:30:40 +0000 (19:30 +0000)]
Merge "resourceloader: Change 'packageFiles' format to be JSON-compatible"
James D. Forrester [Fri, 22 Feb 2019 10:30:11 +0000 (10:30 +0000)]
foreign-resources: Pull all moment locales, even the ones we don't yet use
This makes things simpler for configuration at a trivial cost of bytes.
Change-Id: I2d7a0dd4136ccccf1eba6c93af4ba34acfbf0886
James D. Forrester [Wed, 23 Jan 2019 00:51:45 +0000 (16:51 -0800)]
foreign-resources: Add moment to our list of resources
Because we don't pull down all the locales, we current have to list
each of them. To be fixed in a follow-up.
Also add the changelog for in-repo visibility about changes, as we
have begun to make standard.
Change-Id: I839301d05877eb63279cc1dc94ffbe05f8bb0bfc
Roan Kattouw [Fri, 22 Feb 2019 01:31:22 +0000 (17:31 -0800)]
resourceloader: Remove back compat for old packageFiles format
Was introduced as a transitional measure in Ic566a1cd7efd075c3.
Depends-On: I717f03caf9ea8266e6a4d2b6daf4c543c0815931
Change-Id: I6ae615ea38572042f8ba705338067b393827153a
Roan Kattouw [Fri, 22 Feb 2019 01:26:23 +0000 (17:26 -0800)]
resourceloader: Change 'packageFiles' format to be JSON-compatible
The module definition format for 'packageFiles', as initially designed,
mixes sequential and associative arrays. This works in PHP, but not in
JSON. To make the format JSON compatible, use a 'name' key instead of
using the key in the main array.
Leave backwards compatibility in place so that extensions using the old
format can be migrated. This will be removed in the next commit.
Before:
'packageFiles' => [
'script.js',
'script2.js',
'config.json' => [ 'config' => [ 'Foo', 'Bar' ] ],
'data.json' => [ 'callback' => function () { ... } ],
],
After:
'packageFiles' => [
'script.js',
'script2.js',
[ 'name' => 'config.json', 'config' => [ 'Foo', 'Bar' ] ],
[ 'name' => 'data.json', 'callback' => function () { ... } ],
],
This can then be written in extension.json as:
"packageFiles": [
"script.js",
"script2.js",
[ "name": "config.json", "config": [ "Foo", "Bar" ] ],
[ "name": "data.json", "callback: [ "MyExtHooks", "getData" ] ]
]
Change-Id: Ic566a1cd7efd075c380bc50ba0cc2c329a2041d7
jenkins-bot [Fri, 22 Feb 2019 17:37:38 +0000 (17:37 +0000)]
Merge "Return early in Title::checkUserBlock() if user does not have a block."
Antoine Musso [Fri, 22 Feb 2019 13:58:27 +0000 (14:58 +0100)]
selenium: prevent webdriverio automatic screenshot
When the wdio configuration has screenshotPath, when a test fail a
screenshot is automatically taken. However there is no support to
specify a filename, we thus went with our own handler in afterTest hook
to have the screenshot names after the full test name.
As a result we had two screenshots taken:
A) CHROME-Error-xxx.png
B) Suite-Sometest.png
Set screenshotPath to null which disable the automatic screenshoting.
Adjust rest of code to use logPath as a base path to save screenshots.
As a side effect, it works around a nasty race condition in the
reporting. The report is generated while some test still has an
afterTest hook running. The report fail to generate and cause nodejs to
exit. The exact fix still has to be figured out though. T216818.
Bug: T216818
Change-Id: I4b53c98bd4de0d9ff850bb883ba899dd1dce0274