[Webkit-unassigned] [Bug 45959] String.prototype.replace passes undefined rather than empty string for groups that don't match

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=45959





--- Comment #5 from Oliver Hunt <oliver at apple.com>  2010-09-18 21:32:14 PST ---
(In reply to comment #4)
> Hmm, when I ran the js tests, there was a regression.  Upon investigating further into the file that contained the failing tests, I ran across this bug report from 2007: https://bugs.webkit.org/show_bug.cgi?id=14931.  If you look at the blog post it references, along with some of the proposed patches, it appears as if this bug were undoing what they that bug supposedly fixed.
> 
> I'm not really sure which way is the correct way, and the parts of the ECMA-262 spec quoted by the Mozilla bug report above seem to leave this particular detail undefined.  Thoughts?

Ah the hilarity.  Our normal approach in cases like this is to look at firefox and IE behaviour and see what they do, if they do the same thing we can consider changing to match their behaviour.

We should also contact the ecmascript mailing list to get them to clarify the expected behaviour.

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.