[CSS Exclusions] Properly position multiple stacked floats with non rectangular shape outside

[Bem Jones-Bey]

With a non-rectangular shape-outside, stacked floats should use the bounding box of previous floats to position.

[Bem Jones-Bey]

Created attachment 196877 [details]
Patch

Fix stacked float handling

[Dean Jackson]

Comment on attachment 196877 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=196877&action=review

> Source/WebCore/ChangeLog:22
> +        (WebCore::RenderBlock::logicalRightOffsetForLine): Add paramter for offset mode.

Typo: parameter

> Source/WebCore/rendering/RenderBlockLineLayout.cpp:202
> +    RenderBlock::FloatingObjectSetIterator it = floatingObjectSet.end();
> +    --it; // Go to last item.
> +    RenderBlock::FloatingObjectSetIterator begin = floatingObjectSet.begin();
> +    while (it != begin) {
> +        --it;

Doesn't this skip the very last item in the first pass? Is that ok?

[Bem Jones-Bey]

(In reply to comment #2)
> (From update of attachment 196877 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=196877&action=review
> 
> > Source/WebCore/ChangeLog:22
> > +        (WebCore::RenderBlock::logicalRightOffsetForLine): Add paramter for offset mode.
> 
> Typo: parameter

Thanks, I'll fix that.

> 
> > Source/WebCore/rendering/RenderBlockLineLayout.cpp:202
> > +    RenderBlock::FloatingObjectSetIterator it = floatingObjectSet.end();
> > +    --it; // Go to last item.
> > +    RenderBlock::FloatingObjectSetIterator begin = floatingObjectSet.begin();
> > +    while (it != begin) {
> > +        --it;
> 
> Doesn't this skip the very last item in the first pass? Is that ok?

I was thinking that would be ok, since our new float will always be the last one on the list the way that the code is currently used, but I should be paranoid and not skip the last one just in case I'm missing something.

[Bem Jones-Bey]

Created attachment 197167 [details]
Patch

Update for review comments

[WebKit Commit Bot]

Comment on attachment 197167 [details]
Patch

Rejecting attachment 197167 [details] from commit-queue.

Failed to run "['/Volumes/Data/EWS/WebKit/Tools/Scripts/webkit-patch', '--status-host=webkit-commit-queue.appspot.com', '--bot-id=webkit-cq-02', 'validate-changelog', '--non-interactive', 197167, '--port=mac']" exit_code: 1 cwd: /Volumes/Data/EWS/WebKit

/Volumes/Data/EWS/WebKit/LayoutTests/ChangeLog neither lists a valid reviewer nor contains the string "Unreviewed" or "Rubber stamp" (case insensitive).

Full output: http://webkit-commit-queue.appspot.com/results/17628013

[Bem Jones-Bey]

Created attachment 197169 [details]
Patch

Update for review comments

[Zoltan Horvath]

Comment on attachment 197169 [details]
Patch

Somebody should teach the commit-queue to figure out the reviewer's name at these straightforward cases.

[WebKit Commit Bot]

Comment on attachment 197169 [details]
Patch

Clearing flags on attachment: 197169

Committed r148056: <http://trac.webkit.org/changeset/148056>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.