Roan Kattouw [Sun, 2 Oct 2011 20:02:35 +0000 (20:02 +0000)]
(bug 31293) If Special:Userlogin is loaded over HTTPS, display MediaWiki:loginend-https instead of MediaWiki:loginend, if it exists
Aaron Schulz [Sun, 2 Oct 2011 19:44:31 +0000 (19:44 +0000)]
* Added isCacheWorthy() optimization (checks if the file exists, stale or not)
* Made isCached() use process cache
* Added MISS_TTL_SEC constant and tweaked MISS_FACTOR constant
Roan Kattouw [Sun, 2 Oct 2011 19:35:07 +0000 (19:35 +0000)]
Followup r96930: also remove the documentation that describes the reverted behavior
Aaron Schulz [Sun, 2 Oct 2011 17:53:33 +0000 (17:53 +0000)]
FileCache:
* Added FileCacheBase::*MissesRecent() functions for counting cache misses from different visitors.
* Made ObjectFileCache more generic.
* Cleaned up FileCacheBase::checkCacheDirs().
* Added FileCacheBase::typeSubdirectory() function and overwrote in HTMLFileCache. Fixes r98405 invalidating all existing cache due to directory change.
* Simplified FileCacheBase::checkCacheDirs() a bit
ResourceLoader:
* Use ResourceFileCache to handle load() requests, if $wgUseFileCache. Only caches requests for default language and skins. Single modules requests are always cached, whereas others require a certain threshold of traffic.
* Added ResourceFileCache class (functionality was initially to be in ObjectFileCache).
Daniel Friesen [Sun, 2 Oct 2011 17:42:21 +0000 (17:42 +0000)]
More RequestContext -> IContextSource.
Alexandre Emsenhuber [Sun, 2 Oct 2011 17:36:19 +0000 (17:36 +0000)]
Added missing directory to findHooks.php and a missing hook in hooks.txt
Aaron Schulz [Sun, 2 Oct 2011 17:19:19 +0000 (17:19 +0000)]
* Added end parameter and redid arguments as parameters.
* Set $wgReadOnly to avoid firing enotif stuff and such.
* Pushed commit() statements outside of batch loop...added begin().
* Removed wfWaitForSlaves(), which is just more overhead...nothing should be writing to the DB, the transaction is there just in case.
Derk-Jan Hartman [Sun, 2 Oct 2011 16:50:52 +0000 (16:50 +0000)]
Port r51457 from monobook to modern, simple and Vector.
See also Bug 18207
Takes care of disappearing symbol when clicking extiw links, and triggering the external link and https lock icon on internal links.
Daniel Friesen [Sun, 2 Oct 2011 16:37:32 +0000 (16:37 +0000)]
Merge r98679 from the pageoutput branch into trunk.
Alexandre Emsenhuber [Sun, 2 Oct 2011 16:30:59 +0000 (16:30 +0000)]
* Use local context instead of global variables
* Call Linker methods statically
Derk-Jan Hartman [Sun, 2 Oct 2011 16:10:41 +0000 (16:10 +0000)]
Follow up r98665. And remove the old css of course.
Derk-Jan Hartman [Sun, 2 Oct 2011 16:06:19 +0000 (16:06 +0000)]
Follow up to r98665.
Forgot to add the new css file.
Alexandre Emsenhuber [Sun, 2 Oct 2011 16:01:35 +0000 (16:01 +0000)]
* (bug 30466) Entries in iwlinks table are now cleared when moving a page over redirect
Derk-Jan Hartman [Sun, 2 Oct 2011 11:58:46 +0000 (11:58 +0000)]
Add tfoot emulation to tablesorter. Fixes part of the complaints in bug 31060
Derk-Jan Hartman [Sun, 2 Oct 2011 11:08:12 +0000 (11:08 +0000)]
Add the class jquery-tablesorter to all tables that are made sortable.
Use this classname to raise the specificity of the sortHeader, so that it is more difficult to accidently override with local css. Followup to r98069
Move the css into a seperate file and move the images for it to the jquery directory.
Aaron Schulz [Sun, 2 Oct 2011 07:29:39 +0000 (07:29 +0000)]
Reverted accidental change in r98405
Max Semenik [Sun, 2 Oct 2011 06:10:55 +0000 (06:10 +0000)]
Revert r98355 and r98357; drop the field instead.
Aaron Schulz [Sun, 2 Oct 2011 06:10:45 +0000 (06:10 +0000)]
Don't show 'autoblock disabled' for IP blocks
Aaron Schulz [Sat, 1 Oct 2011 23:39:36 +0000 (23:39 +0000)]
FU r98559: added cannotchangeemail
Santhosh Thottingal [Sat, 1 Oct 2011 22:34:54 +0000 (22:34 +0000)]
mediaWiki.language.convertPlural not required since it is same as the default implimentation in mediawiki.language.js
Santhosh Thottingal [Sat, 1 Oct 2011 18:42:36 +0000 (18:42 +0000)]
Fix Bug 29299 - Zero should be considered plural for pt-br
Krinkle [Sat, 1 Oct 2011 18:20:31 +0000 (18:20 +0000)]
Add support for namespaces in Special:RecentChanges subpage filter (bug 31297)
Aaron Schulz [Sat, 1 Oct 2011 06:16:08 +0000 (06:16 +0000)]
Fixed case of ID in selector
Aaron Schulz [Sat, 1 Oct 2011 05:55:12 +0000 (05:55 +0000)]
Fixed bogus addModules() call which broke the JS on the block form (didn't load)
Aaron Schulz [Sat, 1 Oct 2011 05:11:21 +0000 (05:11 +0000)]
Tweak r98602: keep the other IP normalization for non-ranges, such as upper case for v6. This keeps it a bit more consistent.
Aaron Schulz [Sat, 1 Oct 2011 05:03:10 +0000 (05:03 +0000)]
Reverted stuff from r98602 committed by mistake
Aaron Schulz [Sat, 1 Oct 2011 04:54:44 +0000 (04:54 +0000)]
Made provideIPCIDRs() not add a '/' if there is no prefix length. Fixes bug 31234.
Aaron Schulz [Sat, 1 Oct 2011 04:15:07 +0000 (04:15 +0000)]
Refactored out new sendResponseHeaders() and tryRespondLastModified() functions out of respond() code. These changes were made for $wgUseFileCache support (haven't committed the other stuff here).
Brion Vibber [Sat, 1 Oct 2011 02:03:05 +0000 (02:03 +0000)]
* (bug 31289) Fix fatal error on special pages using $this->msg() on PHP 5.2
According to online PHP docs, func_get_args() doesn't work in function parameter lists until PHP 5.3.0.
It does, however, work as the *first* parameter, which happens here and there in older code (eg array_slice(func_get_args(),1)).
Assigning to a variable and passing through works fine for this case on PHP 5.2.17.
Santhosh Thottingal [Sat, 1 Oct 2011 01:35:06 +0000 (01:35 +0000)]
Revert r97788
Brion Vibber [Fri, 30 Sep 2011 22:50:48 +0000 (22:50 +0000)]
* (bug 12130) Initial newlines are now preserved correctly during editing
HTML browsers strip the first newline from the literal contents of a <textarea>, but keep any additional newlines beyond that.
Prepending an extra newline in Html::textarea() when the contained text started with a newline makes our output basically look like this:
<textarea> <-- (this added newline is ignored)
<-- (any original newline here is preserved)
blah blah</textarea>
This seems to consistently resolve the stripping of single initial newlines from every edit operation as seen on bug 12130; as noted on comments there <https://bugzilla.wikimedia.org/show_bug.cgi?id=12130#c17> this had deleterious effects on Wikisource, where transcription/proofreading tends to involve breaking up lots of little pages, which may have a significant newline at the start of a page boundary.
Text that didn't have initial newlines won't see any difference in the HTML output.
Followup to test cases in r98576, which confirm that supported browsers consistently have this behavior.
Brion Vibber [Fri, 30 Sep 2011 22:16:37 +0000 (22:16 +0000)]
QUnit test case to confirm consistent browser behavior with initial newlines inside <textarea> HTML source. (for bug 12130)
Aaron Schulz [Fri, 30 Sep 2011 21:40:03 +0000 (21:40 +0000)]
FU r98559: added 'cannotchangeemail' msg
Aaron Schulz [Fri, 30 Sep 2011 20:43:45 +0000 (20:43 +0000)]
(bug 31283) Made ChangeEmail check $wgAuth->allowPropChange( 'emailaddress' ). Fix for r92924. Mark beat me to committing the rest of the fix in r98558, which just hid the link to the special page.
Mark A. Hershberger [Fri, 30 Sep 2011 20:36:52 +0000 (20:36 +0000)]
re r92924 andFix Bug #31283 - "Change E-mail" option in Preference should be hidden
when AuthPlugin#allowEmailChange() returns false
Patch from Rusty Burchfield
Aaron Schulz [Fri, 30 Sep 2011 17:31:36 +0000 (17:31 +0000)]
Made addNewUserLogEntry() return the ID of the new log entry to be more useful
Chad Horohoe [Fri, 30 Sep 2011 15:57:32 +0000 (15:57 +0000)]
Syntax error fix for r98528
Niklas Laxström [Fri, 30 Sep 2011 15:45:48 +0000 (15:45 +0000)]
Adding two maintenance classes to autoloader - Translate extension is using them
Roan Kattouw [Fri, 30 Sep 2011 13:57:06 +0000 (13:57 +0000)]
Revert tests
Roan Kattouw [Fri, 30 Sep 2011 13:52:00 +0000 (13:52 +0000)]
Demonstrating that committing to SVN works for me, Sam was complaining about it being broken.
Roan Kattouw [Fri, 30 Sep 2011 10:14:39 +0000 (10:14 +0000)]
Address fixme on r96978: expand URLs fed to the client-side loader
Sean Colombo [Thu, 29 Sep 2011 23:11:31 +0000 (23:11 +0000)]
Revision::newFromTitle() returns null sometimes. Protect against the errors.
Sam Reed [Thu, 29 Sep 2011 22:28:10 +0000 (22:28 +0000)]
Kill the updater remnants from r98467
Sam Reed [Thu, 29 Sep 2011 22:17:20 +0000 (22:17 +0000)]
More spurious code r98467
Sam Reed [Thu, 29 Sep 2011 22:16:00 +0000 (22:16 +0000)]
And another brace... r98467
Sam Reed [Thu, 29 Sep 2011 22:14:44 +0000 (22:14 +0000)]
And an extra brace from r98467
Sam Reed [Thu, 29 Sep 2011 22:11:51 +0000 (22:11 +0000)]
Another reverse merge artefact from r98467
Sam Reed [Thu, 29 Sep 2011 22:10:06 +0000 (22:10 +0000)]
Fix comment syntax error from r98467
Sam Reed [Thu, 29 Sep 2011 22:08:00 +0000 (22:08 +0000)]
Back out r95396 and friends
Aaron Schulz [Thu, 29 Sep 2011 21:34:34 +0000 (21:34 +0000)]
FR r98459: forgot to commit this!
Raimond Spekking [Thu, 29 Sep 2011 21:33:17 +0000 (21:33 +0000)]
Localisation updates for core and extension messages from translatewiki.net
Aaron Schulz [Thu, 29 Sep 2011 21:30:00 +0000 (21:30 +0000)]
FU r98412: added version to wfDeprecated call
Aaron Schulz [Thu, 29 Sep 2011 21:20:51 +0000 (21:20 +0000)]
Refactored HistoryPage to use RequestContext
Aaron Schulz [Thu, 29 Sep 2011 21:17:43 +0000 (21:17 +0000)]
FU r98405: various document cleanup
Sam Reed [Thu, 29 Sep 2011 21:11:03 +0000 (21:11 +0000)]
Followup r98452, move RELEASE-NOTES from r97403
Sam Reed [Thu, 29 Sep 2011 21:08:04 +0000 (21:08 +0000)]
Fix eaten line from r98447
Also remove from RELEASE-NOTES-1.19
Sam Reed [Thu, 29 Sep 2011 21:00:17 +0000 (21:00 +0000)]
Followup r98446, move RELEASE-NOTES from r95318
John Du Hart [Thu, 29 Sep 2011 20:21:32 +0000 (20:21 +0000)]
Followup r95921, clearer PHPDoc and better variable names per CR
Bryan Tong Minh [Thu, 29 Sep 2011 19:08:08 +0000 (19:08 +0000)]
Use data providers
Alexandre Emsenhuber [Thu, 29 Sep 2011 19:03:57 +0000 (19:03 +0000)]
Per Aaron, fix for r98012: make Title object be handled correctly and document where there can now be a Title object
Bryan Tong Minh [Thu, 29 Sep 2011 19:00:45 +0000 (19:00 +0000)]
(bug 30202) Restrict file names on upload to 240 bytes, because wfTimestamp( TS_MW ) . '!' . $fileName should fit in oi_archive_name, which is 255 bytes, and also the maximum file name length on many file systems is 255 bytes.
Commit to fix UploadTest to use @dataProvider will follow
Bryan Tong Minh [Thu, 29 Sep 2011 18:35:34 +0000 (18:35 +0000)]
(bug 31213) Exception thrown when trying to move file cross-namespace.
* Don't try to construct a File object when the namespace is wrong
* Start on some tests
Aaron Schulz [Thu, 29 Sep 2011 17:54:19 +0000 (17:54 +0000)]
FU r98421: forgot to move mParserOptions field
Aaron Schulz [Thu, 29 Sep 2011 17:51:27 +0000 (17:51 +0000)]
Moved getParserOptions() crap back to Article.php due to $wgUser dependency. Callers updated.
Aaron Schulz [Thu, 29 Sep 2011 17:44:43 +0000 (17:44 +0000)]
Made getParserOutput() actually use the right parserOptions for the user
Aaron Schulz [Thu, 29 Sep 2011 17:39:14 +0000 (17:39 +0000)]
Fix for r97091: Extensions can cause anons to have their own language settings. Distinguish anon options from canonical options.
Alexandre Emsenhuber [Thu, 29 Sep 2011 15:21:54 +0000 (15:21 +0000)]
Per Aaron, follow-up r96859: add @deprecated comment, a wfDeprecated() and mark it for removal in next version since it's not used anywhere
Brian Wolff [Thu, 29 Sep 2011 15:16:35 +0000 (15:16 +0000)]
Fix handling of qplimit/qpoffset and cachedtimestamp in QueryPage api module
Basically api was thinking qplimit was the offset, and qpoffset was the limit. Also remove an extra ! when checking for the cached timestamp which was stopping the display of cachedtimestamp property in the api output.
I'm going to tag this 1.18 since this module was introduced in 1.18, and I assume we'd want to fix the limit/offset parameter before the wrong behaviour gets released and people depend on it, etc.
Alexandre Emsenhuber [Thu, 29 Sep 2011 15:14:34 +0000 (15:14 +0000)]
Per Aaron, follow-up r93758: add coments and fix typos
Alexandre Emsenhuber [Thu, 29 Sep 2011 15:07:13 +0000 (15:07 +0000)]
Follow-up r89568: use local context instead of $wgOut
Aaron Schulz [Thu, 29 Sep 2011 08:18:20 +0000 (08:18 +0000)]
HTMLFileCache refactoring:
* Rewrote class and split into three classes: a base class, and html cache and a more generic cache to be used later.
* The new classes now use RequestContext.
* Renamed fetchPageText() -> fetchText().
* Split out new saveText() function from saveToFileCache().
* Various other cleanups and fixes.
Also fixed backwards setting of $wgDisableCounters in rebuildFileCache.php.
Krinkle [Wed, 28 Sep 2011 22:47:05 +0000 (22:47 +0000)]
clean up Html::inlineScript usage
* ResourceLoader::makeLoaderConditionalScript:
-- window.mediaWiki -> window.mw; Not just because it's shorter but because that's the variable that is actually being used inside the script.
* ProtectionForm::buildCleanupScript
-- Use a single quote string and no new line
-- Use ResourceLoader::makeLoaderConditionalScript instead of duplicating this logic everywhere
* ProtectionForm::buildCleanupScript:
-- Use Xml::encodeJsCall to render javascript code
-- Use ResourceLoader::makeLoaderConditionalScript
-- Fix bug 31230 by using the new OutputPage->addJsConfigVars( r98374 )
(had to use $wgOut for now since there wasn't an obvious route to a context that I could find here (Article has context, but ProtectionForm::__construct takes any WikiPage, not just Article)
* EditPage::getEditToolbar:
-- Use Xml::encodeJsCall
-- Use ResourceLoader::makeLoaderConditionalScript
Fixes:
* (bug 31230) ProtectionForm should set wgCascadeableLevels in mw.confg instead of globally
Krinkle [Wed, 28 Sep 2011 22:08:08 +0000 (22:08 +0000)]
New OutputPage::addJsConfigVars() method (bug 31233)
* to make the output page specific mw.config map extendable.
* fixes (bug 31233) Make OutputPage's mw.config array extendable internally instead of hard coded
* fixes weird unfinished comment from r61690
Krinkle [Wed, 28 Sep 2011 22:00:51 +0000 (22:00 +0000)]
shared.css clean up:
* Use background-color instead of background when possible (to support cascading rules and don't reset the entire thing)
* no parens around url()
Krinkle [Wed, 28 Sep 2011 20:33:27 +0000 (20:33 +0000)]
inexisting -> inexistent (typo)
* tests/qunit
* resources/mediawiki
Niklas Laxström [Wed, 28 Sep 2011 20:28:40 +0000 (20:28 +0000)]
Followup r98135
* Added gender parameter to newuserlog-create2-entry
* Hide old comments (for all log types!)
* logentry-newusers-newusers needs a message after all
* Move user tool links away from the action links and drop the User: prefix from create2
Sam Reed [Wed, 28 Sep 2011 19:38:34 +0000 (19:38 +0000)]
Followup r98339, r98355
Let's pretend the column never existed
Sam Reed [Wed, 28 Sep 2011 19:28:19 +0000 (19:28 +0000)]
Followup r98339
Don't try and remove the user_options column in SQLite
Added 'user.user_options' to ignores
Alexandre Emsenhuber [Wed, 28 Sep 2011 19:22:19 +0000 (19:22 +0000)]
* Use local context instead of global variables
* Call Linker methods statically
Robin Pepermans [Wed, 28 Sep 2011 19:02:13 +0000 (19:02 +0000)]
Release notes for r98253
Sam Reed [Wed, 28 Sep 2011 18:19:57 +0000 (18:19 +0000)]
Followup r98339
Add RELEASE-NOTES-1.19
Fix runChild in DatabaseUpdater.php
Sam Reed [Wed, 28 Sep 2011 18:08:48 +0000 (18:08 +0000)]
* (bug 31204) Remove old user.user_options
Added run of ConvertUserOptions maintenance script into update.php before dropping the column
Removed usages of user_options in code
Marked User::decodeOptions() deprecated as of 1.19.
Made ConvertUserOptions drop out early if the user_options field doesn't exist
Made ConvertUserOptions update user_options to '' after migration of user options to mOptions
Sam Reed [Wed, 28 Sep 2011 17:01:53 +0000 (17:01 +0000)]
Followup r98329
Change ApiQueryInfo to match also
Sam Reed [Wed, 28 Sep 2011 16:37:27 +0000 (16:37 +0000)]
Fix whitespace
Sam Reed [Wed, 28 Sep 2011 16:36:17 +0000 (16:36 +0000)]
Bug 31199 - Fix notification of implicit groups so it doesn't massively change json output
Needs merging to 1.18, 1.18wmf1 and pushing to site
Aaron Schulz [Wed, 28 Sep 2011 01:28:29 +0000 (01:28 +0000)]
(bug 31206) "Altering blocks doesn't appear to be working". Don't compare string to User objects :)
Siebrand Mazeland [Wed, 28 Sep 2011 01:07:55 +0000 (01:07 +0000)]
(bug 31194) Add linkTrail for Northern Sami.
Brion Vibber [Tue, 27 Sep 2011 22:51:36 +0000 (22:51 +0000)]
* (bug 31187) Fix for user JavaScript validation to allow identifiers with valid Unicode letter characters
Followup r91591, r93020: patch to jsminplus to support Unicode chars and char escapes in identifiers
Fast-path check keeps runtime about the same on most scripts (eg jquery.js parsing was abround 4100ms both before and after on my test machine)
Slow-path code kicks in if plain ASCII word chars don't extend all the way to the next whitespace or punctuation char.
Using PCRE's Unicode properties magic to ensure that we're catching everything, following ECMA-262 edition 5.1 spec.
Note that identifiers using escapes don't get normalized to their UTF-8 form; this might be a nice thing to do as it saves a couple bytes, but currently there's no change made to output.
Added QUnit tests to verify that unicode letter & escapes work in identifiers in all supported browsers (ok back to IE 6, yay)
Robin Pepermans [Tue, 27 Sep 2011 21:00:38 +0000 (21:00 +0000)]
(bug 31193) Set fallback language of Assamese from Bengali to English
Raimond Spekking [Tue, 27 Sep 2011 19:58:34 +0000 (19:58 +0000)]
Localisation updates for core and extension messages from translatewiki.net
Roan Kattouw [Tue, 27 Sep 2011 19:21:53 +0000 (19:21 +0000)]
(bug 31191) Followup to r90934: do this in the successful login code path too. The bug didn't surface in most cases because the user is redirected immediately after a successful login, except when CentralAuth stops that from happening.
Chad Horohoe [Tue, 27 Sep 2011 18:56:44 +0000 (18:56 +0000)]
Also add $wgRequest to the forbidden globals...thought it was common sense.
Aaron Schulz [Tue, 27 Sep 2011 18:55:15 +0000 (18:55 +0000)]
(bug 31088) [Regression] "Create account/log in" should use the current title and query in returnto
Sam Reed [Tue, 27 Sep 2011 16:15:29 +0000 (16:15 +0000)]
Fix character access
Alexandre Emsenhuber [Tue, 27 Sep 2011 15:13:37 +0000 (15:13 +0000)]
* Use local context instead of global variables
* Call Linker methods statically
* Changed link to Special:Whatlinkshere on Special:Mostlinkedtemplates to pass the page in the subpage parameter instead of "target" to match Special:Mostlinked's behaviour
Alexandre Emsenhuber [Tue, 27 Sep 2011 14:47:42 +0000 (14:47 +0000)]
Per OverlordQ, fix for r87164: let's use an aggregate function that exists in postgresql
Daniel Friesen [Tue, 27 Sep 2011 07:58:23 +0000 (07:58 +0000)]
Followup r98210; Add an ugly hack to support old extensions using the SkinTemplateToolboxEnd hook, erm... 'thanks' Tim.
((Also fix an extra htmlspecialchars that shouldn't have been there))
Aaron Schulz [Tue, 27 Sep 2011 06:45:58 +0000 (06:45 +0000)]
Improved filecache short-circuit slightly (catching NS_MEDIA and such)
Daniel Friesen [Tue, 27 Sep 2011 05:31:57 +0000 (05:31 +0000)]
Add getSidebar method to BaseTemplate to simplify the sidebar boilerplate.
Brion Vibber [Tue, 27 Sep 2011 00:41:24 +0000 (00:41 +0000)]
Debug logging and possible fix for bug 31177: mystery sidebar message failures
This may be related to things failing on load from External Storage; however we have not yet been able to verify this.
Tweaks MessageCache::loadFromDB() and MessageCache::getMsgFromNamespace() to avoid storing empty cache entries when loading text fails.
When building initial cache if we get a failure we'll log and store a '!TOO BIG' message which requests on-demand load later.
If an on-demand load failures, we'll log and return the false through but won't update the cache with the bad value.
To enable the logging in production, set up a $wgDebugLogFiles entry for 'MessageCache'.
Note that MessageCache::loadFromDB() bypasses Revision's text entry memcaching and may cause a lot of ES fetches at once.
However any ES failures *should* already be logged in the 'ExternalStoreDB' log file.