Sam Reed [Sun, 2 Jan 2011 03:20:45 +0000 (03:20 +0000)]
ApiQueryRevisions: Switch from manual Cache checking, to use $article->getParserOutput()
ApiParse: Fix comment typo
X! [Sun, 2 Jan 2011 02:33:07 +0000 (02:33 +0000)]
Add more wfDebug tests
X! [Sun, 2 Jan 2011 02:16:32 +0000 (02:16 +0000)]
Followup r79467: Add more wfDebug() tests, complete with MockOutputPage
X! [Sun, 2 Jan 2011 01:59:52 +0000 (01:59 +0000)]
Add some wfDebug() unit tests
X! [Sun, 2 Jan 2011 01:51:44 +0000 (01:51 +0000)]
Add fallback mb_* function tests
X! [Sun, 2 Jan 2011 01:35:24 +0000 (01:35 +0000)]
Fix r79463 and r79464: Syntax error
X! [Sun, 2 Jan 2011 01:34:26 +0000 (01:34 +0000)]
Accidental syntax error. *hides*
X! [Sun, 2 Jan 2011 01:29:00 +0000 (01:29 +0000)]
Move fallback function creation out of function_exists() conditionals.
This allows for unit testing of the fallback functions to ensure that
they work like the real functions do
Aryeh Gregor [Sun, 2 Jan 2011 00:10:42 +0000 (00:10 +0000)]
Fix broken blacklisting code from r79363
Problem reported by Krinkle -- the code I copy-pasted from the wiki page
was wrong, and apparently I didn't test properly in Opera.
Sam Reed [Sat, 1 Jan 2011 23:06:04 +0000 (23:06 +0000)]
Add missing $popts parameter, and pass it from r79455
Sam Reed [Sat, 1 Jan 2011 23:01:24 +0000 (23:01 +0000)]
Allow section to work with oldid when oldid == currentrevid (worked with older id)
Sam Reed [Sat, 1 Jan 2011 22:49:47 +0000 (22:49 +0000)]
Per Platonides on CR r75334, use $article->getParserOutput() to get latest revision, rather than doing parser cache stuff ourselves
Helps reduce code duplication too
Sam Reed [Sat, 1 Jan 2011 22:46:23 +0000 (22:46 +0000)]
Document getParserOutput() return type in Article
Sam Reed [Sat, 1 Jan 2011 22:42:53 +0000 (22:42 +0000)]
Per Nikerabbit on IRC, and Bryan on CR (r79451), change back for ===, and intval( $oldid )
Sam Reed [Sat, 1 Jan 2011 22:35:16 +0000 (22:35 +0000)]
Followup r75334. Change to int == string, rather than int === string (not going to work)
Attempt to load/save an article object, rather than an title object from the parser cache.
Per Bryan, and per bug 26534 now logged, this really needs refactoring more cleanly. As if latestRevId == oldid set, it's doing exactly the same lookup/parse
Platonides [Sat, 1 Jan 2011 22:24:51 +0000 (22:24 +0000)]
rm unused global
X! [Sat, 1 Jan 2011 22:19:22 +0000 (22:19 +0000)]
Fix r79447: Remove debugging code
Platonides [Sat, 1 Jan 2011 22:16:58 +0000 (22:16 +0000)]
Remove unused globals
X! [Sat, 1 Jan 2011 22:16:54 +0000 (22:16 +0000)]
Followup r79443: Add some form of organization to the test autoloader
Raimond Spekking [Sat, 1 Jan 2011 22:09:01 +0000 (22:09 +0000)]
Localisation updates for core and extension messages from translatewiki.net (2011-01-01 21:40:00 UTC)
Platonides [Sat, 1 Jan 2011 22:04:14 +0000 (22:04 +0000)]
Use an autoloader for the tests, following the ideas from r72858.
The SeleniumTestConstants class was added in r79437
Alexandre Emsenhuber [Sat, 1 Jan 2011 21:57:32 +0000 (21:57 +0000)]
Per Platonides, fix for r79430: removed debugging code
X! [Sat, 1 Jan 2011 20:57:13 +0000 (20:57 +0000)]
Reorder functions around, add some form of order to the test suite.
It'd be nice to break some of these functions out to another class, but then again,
I'd also like to see Windows actually become stable, MediaWiki to become fully class based,
and Google to stop taking over the world. We can't all have what we want.
Platonides [Sat, 1 Jan 2011 20:56:39 +0000 (20:56 +0000)]
I am fed up of selenium constants which aren't properly defined by all files using them.
Making those of SeleniumTestConstants.php into class constants.
for i in $(grep -o 'const [A-Z_]*' SeleniumTestConstants.php | cut -c 7-); do
find -name \*.php -and -not -name SeleniumTestConstants.php -exec sed -i s/$i/SeleniumTestConstants::$i/g \{\} +
done
X! [Sat, 1 Jan 2011 18:10:28 +0000 (18:10 +0000)]
Add fuzz tests to new parser tests (someone else should make sure this works as expected)
Daniel Friesen [Sat, 1 Jan 2011 17:51:10 +0000 (17:51 +0000)]
Fix Vector to output a proper newline and indentation before the ending body and html tags... (*cough* A certain sneaker wearing rabbit made me do it)
Daniel Friesen [Sat, 1 Jan 2011 17:38:15 +0000 (17:38 +0000)]
Add a realname option to query=userinfo in the api.
X! [Sat, 1 Jan 2011 17:05:08 +0000 (17:05 +0000)]
Don't use the --verbose option anymore
Alexandre Emsenhuber [Sat, 1 Jan 2011 16:58:00 +0000 (16:58 +0000)]
Made the page_count update defered, as for the site_stats update. I just removed Article::incViewCount() since nothing else was calling this function.
Daniel Friesen [Sat, 1 Jan 2011 16:44:58 +0000 (16:44 +0000)]
Fix modern to use the tooltiponly attribute.
Daniel Friesen [Sat, 1 Jan 2011 16:41:45 +0000 (16:41 +0000)]
Convert the common bottomscripts, reporttime, and debug boilerplate into a printTrail() helper method.
Note that in vector this does move the fixalpha stuff above bottomscripts while it was after it before, theoretically though this shouldn't cause any trouble.
Markus Glaser [Sat, 1 Jan 2011 15:52:52 +0000 (15:52 +0000)]
SimpleSeleniumTest now has its own resources (database and images dir)
Markus Glaser [Sat, 1 Jan 2011 15:51:54 +0000 (15:51 +0000)]
test data for SimpleSeleniumTest
Markus Glaser [Sat, 1 Jan 2011 15:51:37 +0000 (15:51 +0000)]
test data for SimpleSeleniumTest
Markus Glaser [Sat, 1 Jan 2011 15:49:58 +0000 (15:49 +0000)]
typo in comment
Markus Glaser [Sat, 1 Jan 2011 15:49:33 +0000 (15:49 +0000)]
changed path after relocation
Markus Glaser [Sat, 1 Jan 2011 15:47:43 +0000 (15:47 +0000)]
now uses separate database and images dir for selenium tests if $wgSeleniumUseTestResources is set
Alexandre Emsenhuber [Sat, 1 Jan 2011 12:40:21 +0000 (12:40 +0000)]
Update copyright year
Shinjiman [Sat, 1 Jan 2011 10:59:32 +0000 (10:59 +0000)]
Localisation updates Cantonese, Chinese and Literary Chinese
X! [Sat, 1 Jan 2011 06:49:00 +0000 (06:49 +0000)]
Add parsertest specific help to PHPUnit.php
X! [Sat, 1 Jan 2011 05:53:04 +0000 (05:53 +0000)]
More work on new parser tests:
-No need to specify a new function for each argument, use PHP5 magic to automatically set it
-Create (g|s)etCliArg() functions
-Implement some features from the old parser tests
X! [Sat, 1 Jan 2011 03:39:37 +0000 (03:39 +0000)]
Start work on porting ParserTests completely into PHPunit.
Disable temporary tables, a few reports of them not working. The tables get deleted anyway, so why risk it?
Daniel Friesen [Sat, 1 Jan 2011 01:05:39 +0000 (01:05 +0000)]
*sigh* I wish this were git... Followup r79398, small mistake that slipped through when I was fixing an incorrect method use.
Daniel Friesen [Sat, 1 Jan 2011 01:03:02 +0000 (01:03 +0000)]
Implement "relevant" title and user in the skin system and update undelete, log, contributions, blockip, and movepage to use it.
A "Relevant" is used by the skin to determine what title to display tabs for. This setting allows pages like Special:MovePage (which is linked to from the tabs themselves) to retain the tabs specific to the page relevant to it when switching to the special page.
Similaly a "Relevant" user is used by the skin to display things in the toolbox which would usually only be displayed on the user's userpage and talkpage, pages like Special:Contributions which are linked to by the toolbox can use this to retain the toolbox links when switching between the user pages and these special pages.
Daniel Friesen [Sat, 1 Jan 2011 00:59:12 +0000 (00:59 +0000)]
Fix php notice in vector from r79383.
Daniel Friesen [Fri, 31 Dec 2010 23:48:41 +0000 (23:48 +0000)]
Followup r79383, missing message documentation.
Daniel Friesen [Fri, 31 Dec 2010 23:30:00 +0000 (23:30 +0000)]
Merging Vector's navigation_urls and SkinTemplate's content_actions code into content_navigation. content_actions is now built by folding content_navigation and cleaning it up a bit.
content_actions hooks no longer work and have been dropped from the code, the hooks that affected vector before now affect all skins.
A few logic changes were made to make for a clean merge:
- vector was using vector-???-??? messages while SkinTemplate was using '???' messages. So as a side effect of merging that together all skins now support messages like '$skinname-view-history' which will fallback to the standard message if not defined.
- For MediaWiki: pages where the page does not exist but the message does in the i18n system SkinTemplate displayed "Edit" while Vector displayed "Create"; All skins now display "Edit".
- For users without undelete permissions SkinTemplate displayed an "Undelete" tab if the user had deletedhistory and deletedtext permissions. Vector would only display the tab for users with both deletedhistory and undelete permissions; The new behavior in all skins is to always display a tab if you have deletedhistory (since Special:Undelete will always have something to display) but display a "View ... deleted" instead of "Undelete" message if you do not have undelete permissions.
- Skins no longer need to hardcode tests for the &action= to decide if they should ignore the accesskey on some tabs, tabs which should not have an accesskey in the current page now have a "tooltiponly" key set to true.
X! [Fri, 31 Dec 2010 22:34:20 +0000 (22:34 +0000)]
Forgot to change the other API tests
X! [Fri, 31 Dec 2010 22:08:13 +0000 (22:08 +0000)]
Solve 1 test by changing static variables to non-static
X! [Fri, 31 Dec 2010 21:51:21 +0000 (21:51 +0000)]
For some reason tableExists() is being unreliable, which is resulting in tables not being dropped that _should_ be dropped. Drop them anyway and ignore the errors.
X! [Fri, 31 Dec 2010 21:40:26 +0000 (21:40 +0000)]
-If $this->db does not exist, skip destroying the DB
-Fix ApiTest errors due to DB rewrite
X! [Fri, 31 Dec 2010 21:10:36 +0000 (21:10 +0000)]
Add check to make sure only MySQL and SQLite are used at the moment. These two are the only ones with the listTables() function, and the only ones that are tested.
X! [Fri, 31 Dec 2010 21:05:51 +0000 (21:05 +0000)]
Only destroy DB if it needs it
X! [Fri, 31 Dec 2010 21:01:17 +0000 (21:01 +0000)]
I made MySQL work, and broke SQLite in the process. I fix SQLite, and made MySQL break. Fixing MySQL again, hoping that SQLite works too. :)
X! [Fri, 31 Dec 2010 20:47:48 +0000 (20:47 +0000)]
Solved problem with SQLite: Globals were getting stored, and when a PDO reference is transferred, as is the case with SQLite, it serializes the globals.
Aryeh Gregor [Fri, 31 Dec 2010 20:43:27 +0000 (20:43 +0000)]
RELEASE-NOTES update for r79363
X! [Fri, 31 Dec 2010 20:42:39 +0000 (20:42 +0000)]
More work on getting SQLite to work with unit tests. DB Prefix changing is now static to allow for external classes to change it.
Aryeh Gregor [Fri, 31 Dec 2010 19:53:54 +0000 (19:53 +0000)]
Update dependencies for r79363 per Roan's advice
Aryeh Gregor [Fri, 31 Dec 2010 19:30:29 +0000 (19:30 +0000)]
Enable mwsuggest datalist implementation
Tested in Firefox 4.0b8, which is the only browser it will trigger for
due to the feature test. Opera is explicitly blacklisted for the reason
given in the comment. This feature was added in r61348 but remained
disabled by default because no browser actually supported it properly
(due to Opera's bug). Of course, this is somewhat obsolescent given the
fancy new Vector search, but it's a two-line change, so why not.
I was reminded of this by Jonas Sicking, a Firefox developer, remarking
on how <datalist> should indeed be usable for this purpose in Firefox 4:
http://lists.whatwg.org/htdig.cgi/whatwg-whatwg.org/2010-December/029577.html
Niklas Laxström [Fri, 31 Dec 2010 18:18:57 +0000 (18:18 +0000)]
New method for making it easier for developers to get localised language names.
For now the core itself doesn't provide those, but extension can use the hook
to provide them. Callers should not except full coverage in every language.
Daniel Friesen [Fri, 31 Dec 2010 17:48:07 +0000 (17:48 +0000)]
Add SkinTemplateNavigation::SpecialPage and SkinTemplateNavigation::Universal hooks to match the content_actions based hooks that vector based hooks are missing.
Daniel Friesen [Fri, 31 Dec 2010 17:26:31 +0000 (17:26 +0000)]
Might as well fix this doc file... even though it's completely out of date...
Roan Kattouw [Fri, 31 Dec 2010 16:53:01 +0000 (16:53 +0000)]
r79347 didn't update other callers to mungeQuery(), so let's just take the easy way out and obtain $dbr inside the function so we don't have to change the parameter list
Roan Kattouw [Fri, 31 Dec 2010 16:47:57 +0000 (16:47 +0000)]
Followup r78786: pass $fname in the raw SQL case too
Roan Kattouw [Fri, 31 Dec 2010 16:44:31 +0000 (16:44 +0000)]
Follow-up r78786: the querypage branch merge threw SpecialLinkSearch back to the pre-buildLike() era, so bring back makeLikeArray() and buildLike(). Also, fix an undefined $dbr in the old version.
Roan Kattouw [Fri, 31 Dec 2010 16:31:41 +0000 (16:31 +0000)]
Fix r78970 per CR: params were the wrong way around
Roan Kattouw [Fri, 31 Dec 2010 16:29:44 +0000 (16:29 +0000)]
Remove now-unused SQL timestamp conversion functions added in r77231. They were made obsolete by the querypage-work2 branch merge in r78786
Raimond Spekking [Fri, 31 Dec 2010 14:32:24 +0000 (14:32 +0000)]
Localisation updates for core and extension messages from translatewiki.net (2010-12-31 14:10:00 UTC)
Platonides [Fri, 31 Dec 2010 14:08:10 +0000 (14:08 +0000)]
Follow up r79314.
The table to drop is the NEW one, the one that is going to be recreated.
Sure, the prefix was changed, which means that we need just the table base.
The r79314 behavior was to perform the following:
DROP /* CloneDatabase::cloneTableStructure */ TABLE IF EXISTS `archive`
CREATE /* DatabaseMysql::duplicateTableStructure */ TEMPORARY TABLE `unittest_archive` (LIKE `archive`)
Luckily, it failed there as it didn't find the table it had just dropped.
nadeesha [Fri, 31 Dec 2010 10:05:15 +0000 (10:05 +0000)]
Change the values taken from the SeleniumTestConstants.php
Sam Reed [Fri, 31 Dec 2010 07:25:51 +0000 (07:25 +0000)]
Followup r74519, fixup code style issues
Sam Reed [Fri, 31 Dec 2010 07:13:40 +0000 (07:13 +0000)]
Make BenchmarkPurge subclass benchmark, not really reusing most of the code yet...
Sam Reed [Fri, 31 Dec 2010 07:09:49 +0000 (07:09 +0000)]
Add descriptions
Sam Reed [Fri, 31 Dec 2010 07:06:50 +0000 (07:06 +0000)]
Move benchmarkPurge to maintenance/benchmarks folder
X! [Fri, 31 Dec 2010 02:18:25 +0000 (02:18 +0000)]
Use $oldTableName in CloneDatabase. The dropTable() appends the prefix, so $oldTableName becomes $newTableName
Raimond Spekking [Thu, 30 Dec 2010 21:45:57 +0000 (21:45 +0000)]
Localisation updates for core and extension messages from translatewiki.net (2010-12-20 21:02:00 UTC)
Platonides [Thu, 30 Dec 2010 19:18:09 +0000 (19:18 +0000)]
Follow up r77654
These two files use constants from SeleniumTestConstants.php
I would be more confortable if all these defines were class variables, though.
Removed an empty statement.
Greg Sabino Mullane [Thu, 30 Dec 2010 18:53:48 +0000 (18:53 +0000)]
Don't put multiple tables in the same file.
Sam Reed [Thu, 30 Dec 2010 17:41:19 +0000 (17:41 +0000)]
Fix private class member variable access
Some method documentation
X! [Thu, 30 Dec 2010 17:30:35 +0000 (17:30 +0000)]
-Destroy the DB automatically when initting the DB
-Add $force option to wfSetVar
-More work on getting SQLite to work
Sam Reed [Thu, 30 Dec 2010 17:06:09 +0000 (17:06 +0000)]
Styleli[zs]e the API up to date (Been a while since)
Platonides [Thu, 30 Dec 2010 16:57:44 +0000 (16:57 +0000)]
Spacing, esthetical fixes.
Platonides [Thu, 30 Dec 2010 16:24:29 +0000 (16:24 +0000)]
Follow up r79109. Do not use create_function.
Platonides [Thu, 30 Dec 2010 16:19:10 +0000 (16:19 +0000)]
This global declaration is now unused.
Platonides [Thu, 30 Dec 2010 16:16:16 +0000 (16:16 +0000)]
ResourceLoader::register() works differently since r77011.
This test isn't useful anymore.
Platonides [Thu, 30 Dec 2010 16:10:42 +0000 (16:10 +0000)]
Fix r71751 problems with textual parameters.
Add a test for checking the problems which appeared in r71751
Roan Kattouw [Thu, 30 Dec 2010 12:22:39 +0000 (12:22 +0000)]
(bug 26507) Wrap all ResourceLoader-served JS in (function($, mw) { code here })(jQuery, mediaWiki) so modules can always use $ as an alias for jQuery even if it's been unset in the global scope with jQuery.noConflict()
Roan Kattouw [Thu, 30 Dec 2010 12:16:16 +0000 (12:16 +0000)]
Fix JS errors in prefs.js caused by the (unintentional?) renaming of mw-input-timecorrection to mw-input-wptimecorrection. This was caused by one of the HTMLForm changes but those are all very mysterious to me, so I haven't tried to track down which change is to blame or figure out how to rename the inputs back. Instead, I updated the JS file to use the new IDs
Daniel Friesen [Thu, 30 Dec 2010 10:03:40 +0000 (10:03 +0000)]
Followup r79055 c12495 fix undefined index notice.
nadeesha [Thu, 30 Dec 2010 09:10:46 +0000 (09:10 +0000)]
Replace the common xpaths with the 'LINK_DIV', 'LINK_FORM', 'LINK_RIGHT_FRAMEWORK' variables
nadeesha [Thu, 30 Dec 2010 09:07:23 +0000 (09:07 +0000)]
Added 'LINK_DIV' variable
nadeesha [Thu, 30 Dec 2010 08:37:01 +0000 (08:37 +0000)]
Added two variables 'LINK_FORM' and 'LINK_RIGHT_FRAMEWORK'
Alexandre Emsenhuber [Thu, 30 Dec 2010 07:24:41 +0000 (07:24 +0000)]
Fix for r79231: $dbname was missing
Sam Reed [Thu, 30 Dec 2010 03:00:45 +0000 (03:00 +0000)]
Remove another usage of extract()
Sam Reed [Thu, 30 Dec 2010 02:56:36 +0000 (02:56 +0000)]
No magic variables are used after this extract call. All done $defaults['blah']. Hence remove extract call
Sam Reed [Thu, 30 Dec 2010 02:48:27 +0000 (02:48 +0000)]
Extract sucks
Wrap long line
Sam Reed [Thu, 30 Dec 2010 02:45:28 +0000 (02:45 +0000)]
Followup r79229, fix varible type fail
Sam Reed [Thu, 30 Dec 2010 02:44:26 +0000 (02:44 +0000)]
Add function return type hint to wfGetLBFactory
More parameter documentation to CloneDatabase
Sam Reed [Thu, 30 Dec 2010 02:41:20 +0000 (02:41 +0000)]
Array to array
Add some function documentation
Sam Reed [Thu, 30 Dec 2010 00:56:30 +0000 (00:56 +0000)]
Per CR on r68482, fix adss to add
Also remove trailing whitespace in files