[webkit-reviews] review denied: [Bug 8563] [DOMHTMLDocument doctype] always returning nil. : [Attachment 7940] Patch for this bug

[bugzilla-request-daemon at opendarwin.org]

Darin Adler <darin at apple.com> has denied Darin Adler <darin at apple.com>'s
request for review:
Bug 8563: [DOMHTMLDocument doctype] always returning nil.
http://bugzilla.opendarwin.org/show_bug.cgi?id=8563

Attachment 7940: Patch for this bug
http://bugzilla.opendarwin.org/attachment.cgi?id=7940&action=edit

------- Additional Comments from Darin Adler <darin at apple.com>
(In reply to comment #8)
> Do you have any objections to passing in a lowercase 'html' to new
DocumentType
> in HTMLDocument::determineParseMode instead:
>      setDocType(new DocumentType(this, "html", publicID, systemID));

Sounds fine. We should do this.

> An even better fix, but one which is far more trouble than it is worth, would

> be to pass the same version of the name as read in from the <!Doctype ...>,
but
> WebKit has already lost that information by the time the setDocType call is
> made.

I'm not sure why you say this is more trouble than it's worth, though.

We could easily add a fourth return value to the function
"parseDocTypeDeclaration" in HTMLDocument.cpp to complement the resultFlags,
publicID, and systemID return values.

I think the real issue is not the difficulty of implementation, but the
behavior of other browsers. Do those browsers always return the "html" in the
same case it was in the <!doctype> or do they always return lowercase "html"?
We should match them. If they always return lowercase "html", then this is not
a better fix after all.