For example, "{{#switch:x|#defaultWTF? = This is crazy}}" should not
return "This is crazy".
Bug: 67871
Change-Id: Ie3f2e6b254b9f44028156268bd8d67aa3dd1cc62
This makes foo's equal to foo's. This is particularly important
since {{PAGENAME}} and friends html encode certain characters.
This change should be communicated to users before it hits wikis.
I don't think anyone is relying on the current behaviour, but
you never know.
Bug: 35628
Change-Id: I0035ac541d22491fcf15f55f2589aaab394ed0ed
This test is inspired by the Parsoid commit
https://gerrit.wikimedia.org/r/#/c/28623/, but applies to the PHP version too.
Change-Id: I4814b98597f2dd6a2279b8cdcc90a26fbbdecc71
Would be nice if it was possible to add a test for {{#time:d F Y|1988-02-28|user}} where user language is "nl", because that does not work. Ref bug 28655 comment 10 and comment 17.
This would make something like {{#time:H:i:s| 9:30 January 1, 2012 MST}} convert the time from 9:30 mountatin standard time to whatever it is in UTC.
I made one minor change from the patch on bugzilla in changing an @ to a wfSuppressWarnings. (The @ was already in the code, it wasn't introduced by the patch).
Bug 19093. Previously {{#switch:foo|bar|#default=baz}} would behave
differently from {{#switch:foo|#default|bar=baz}}, even though they'd
behave identically if "#default" was replaced by "foo".
I changed switchObj but not switchHook. In testing, switchObj was the
only method that seemed to be used, and I didn't want to make untested
changes to switchHook, so if that's used, the behavior may remain in
some cases. It seems like code is duplicated between them -- whatever
the difference is, that should probably be factored out.
This passes all existing parser tests for #switch, of which there are
none, so don't blame me if it breaks something. ;)
Uses DateTime class in PHP 5.2+ to support dates outside the 1970-2038 range. On earlier versions will still fall back to strtotime with the old 32-bit Unix timestamp range limitations.
Patch by rememberthedot -- https://bugzilla.wikimedia.org/attachment.cgi?id=5416
Added a couple quick parser test cases to confirm the new behavior. However I am seeing some annoyances with how input time zones are handled, so we'll want to clean that up at some point. :)
