[Webkit-unassigned] [Bug 10827] SVGPathElement should only build SVGPathSeg* list on demand

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=10827





--- Comment #7 from Rob Buis <rwlbuis at gmail.com>  2010-06-13 11:04:43 PST ---
My answers below...

(In reply to comment #4)
 I like the general rebuildPathInfo() methot, but why don't you hand the content of the dAttr over to it?

That is indeed at least in once case more efficient, in the other case the attribute value still must
be fetched. But indeed it is more efficient overall.

 I guess m_ignoreAttributeChanges is intended as an interupt. Can you give some further comments on it?

In the patch it is needed since otherwise svgAttributeChanged -> rebuildPathInfo -> svgAttributeChanged ec. recursion. Seeing that SVGPolyElement removed a similar construction,
it is possible things are handled differently right now and it is not needed anymore.

> What is missing on the normalizedPathSegList-side?

Nothing I think, but I posed some possible problematic cases or cases where I didnt know what to do above.

I think what needs to be done with the patch is, first, to fix it against the bitrot, ie. make it compile and work as before. Then rebuildPathInfo could be changed as discussed. Then it could be tried to remove m_ignoreAttributeChanges. Finally the problematic cases should be analyzed/fixed.
Cheers,

Rob.

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