[Webkit-unassigned] [Bug 68866] CompositeEditCommand::prune should remove subtree at once

[bugzilla-daemon at webkit.org]

https://bugs.webkit.org/show_bug.cgi?id=68866


Darin Adler <darin at apple.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
 Attachment #108786|review?                     |review-
               Flag|                            |




--- Comment #2 from Darin Adler <darin at apple.com>  2011-09-27 09:36:59 PST ---
(From update of attachment 108786)
View in context: https://bugs.webkit.org/attachment.cgi?id=108786&action=review

review- because of the traverseNextSibling issue, and in part due to lack of test

> Source/WebCore/ChangeLog:9
> +        This reduces the number of node removals from O(n) to O(1) where n is the depth of the tree.

Can we make a performance test demonstrating this fix? Ojan’s performance harness typically makes that easy.

> Source/WebCore/editing/CompositeEditCommand.cpp:254
> +            n = n->nextSibling();

I think you want:

    n = n->traverseNextSibling(node);

here, not just nextSibling.

> Source/WebCore/editing/CompositeEditCommand.cpp:277
> +void CompositeEditCommand::prune(PassRefPtr<Node> prpNode)

Normally we only use the "prp" naming if we are transferring ownership to a local variable with a simpler name. Since we’re not doing that here, a prp prefix is not necessary.

> Source/WebCore/editing/CompositeEditCommand.cpp:280
> +        removeNode(highestNodeToRemove);

This should probably have a “release” in it.

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