[webkit-reviews] review granted: [Bug 24150] Add virtual ScriptExecutionContext::encoding() : [Attachment 28345] Updated the test so it runs in FF too.

[bugzilla-daemon at webkit.org]

Alexey Proskuryakov <ap at webkit.org> has granted Dmitry Titov
<dimich at chromium.org>'s request for review:
Bug 24150: Add virtual ScriptExecutionContext::encoding()
https://bugs.webkit.org/show_bug.cgi?id=24150

Attachment 28345: Updated the test so it runs in FF too.
https://bugs.webkit.org/attachment.cgi?id=28345&action=review

------- Additional Comments from Alexey Proskuryakov <ap at webkit.org>
Thanks, that's very useful information! Much of the Gecko behavior looks like
it may be unintentional indeed - would you be willing to file bugs with
Mozilla? It could also be worth discussing this with WHATWG.

-    // FIXME: does this need to provide a charset, like Document::completeURL
does?
+    // Note Document::completeURL provides a charset here (to encode query
portion of url when
+    // submitting charset-encoded forms data). Workers always use UTF-8 which
is default for KURL.
     return KURL(m_location->url(), url);

I think that the word "Note" is extraneous here - it's a comment, after all.
Per your research, it is not clear that we'll want this behavior forever, given
that Firefox uses document encoding for Worker and importScripts() URL
completion. So, the FIXME still seems appropriate, perhaps amended with results
of your research.

r=me