Bug 35981 – Account for margin after when laying out <legend> element

• Component: Layout and Rendering
• Status: RESOLVED FIXED

[Patrick Paul-Hus]

Created attachment 50425 [details]
Demonstration of the bug

I defined a form using semantic markup, but it seems that I can't get a bottom-margin to render under the legend element. Using the web inspector, it does say that there's a bottom-margin applied, but the content that comes after is pushed right underneath and seems to ignore the bottom-margin. I've attached a small html document which shows the issue. It renders fine in IE and FF but there's no bottom margin in Chrome Beta for  and Safari 4.

I am attaching a small HTML document as proof of concept.... for now I fixed the problem by using padding instead of margin.

[Pip]

Created attachment 106589 [details]
Screenshot of margins not being applied in Chromium 13 ubuntu 10.04

Exactly the same code in Firefox 6 makes the second legend float 50px above the fieldset. From http://www.w3schools.com/cssref/tryit.asp?filename=trycss_margin-bottom

[WebKit Review Bot]

Attachment 106589 [details] did not pass style-queue:

Failed to run "['Tools/Scripts/check-webkit-style', '--diff-files']" exit_code: 1

Total errors found: 0 in 0 files


If any of these errors are false positives, please file a bug against check-webkit-style.

[Octavian Damiean]

I can confirm this bug.

Here is a jsfiddle[1] demonstrating the bug. This fiddle[2] shows that the legend element gets rendered with a margin when not setting it inside of a fieldset which of course is an importer use of the legend element.

[1]: http://jsfiddle.net/mainerror/XJYgL/
[2]: http://jsfiddle.net/mainerror/XJYgL/1/

[Patrick Paul-Hus]

2 years soon! Happy bday Bug 35981!

[Julien Chaffraix]

The issue is in RenderFieldset logic that needs to account for the <legend>'s margins in the block flow direction. Note that using margin-top doesn't work either.

[Julien Chaffraix]

Created attachment 146736 [details]
Proposed fix, take the margin after into account.

[Abhishek Arya]

Comment on attachment 146736 [details]
Proposed fix, take the margin after into account.

View in context: https://bugs.webkit.org/attachment.cgi?id=146736&action=review

r=me

> Source/WebCore/ChangeLog:3
> +        Can't apply a bottom-margin to the legend element

nit: probably like "Add bottom-margin support to legend element".

> Source/WebCore/rendering/RenderFieldset.cpp:108
> +        LayoutUnit collapsedLegendExtend;

nit: probably s/Extend/Extent

[Julien Chaffraix]

Created attachment 146923 [details]
Patch for landing

[Abhishek Arya]

Just a very friendly reminder on this cleanup in the testcase :)

-moz-writing-mode: vertical-lr;
writing-mode: vertical-lr;

[Julien Chaffraix]

Comment on attachment 146923 [details]
Patch for landing

> Just a very friendly reminder on this cleanup in the testcase :)

Right, let me remove the -moz* ones.

[Julien Chaffraix]

Created attachment 146936 [details]
Patch for landing

[WebKit Review Bot]

Comment on attachment 146936 [details]
Patch for landing

Clearing flags on attachment: 146936

Committed r120017: <http://trac.webkit.org/changeset/120017>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

(In reply to comment #12)
> (From update of attachment 146936 [details])
> Clearing flags on attachment: 146936
> 
> Committed r120017: <http://trac.webkit.org/changeset/120017>

This caused fast/block/basic/fieldset-stretch-to-legend.html to fail. See for example <http://build.webkit.org/results/Lion%20Release%20(WebKit2%20Tests)/r120021%20(8106)/results.html>

(In reply to comment #14)
> (In reply to comment #12)
> > (From update of attachment 146936 [details] [details])
> > Clearing flags on attachment: 146936
> > 
> > Committed r120017: <http://trac.webkit.org/changeset/120017>
> 
> This caused fast/block/basic/fieldset-stretch-to-legend.html to fail. See for example <http://build.webkit.org/results/Lion%20Release%20(WebKit2%20Tests)/r120021%20(8106)/results.html>

This test needs its expected results updated.

[Julien Chaffraix]

(In reply to comment #15)
> (In reply to comment #14)
> > (In reply to comment #12)
> > > (From update of attachment 146936 [details] [details] [details])
> > > Clearing flags on attachment: 146936
> > > 
> > > Committed r120017: <http://trac.webkit.org/changeset/120017>
> > 
> > This caused fast/block/basic/fieldset-stretch-to-legend.html to fail. See for example <http://build.webkit.org/results/Lion%20Release%20(WebKit2%20Tests)/r120021%20(8106)/results.html>
> 
> This test needs its expected results updated.

Yes, it looks like I missed it during testing but the new results make sense. I will do the update as EFL needs the same treatment.

(In reply to comment #16)
> (In reply to comment #15)
> > (In reply to comment #14)
> > > (In reply to comment #12)
> > > > (From update of attachment 146936 [details] [details] [details] [details])
> > > > Clearing flags on attachment: 146936
> > > > 
> > > > Committed r120017: <http://trac.webkit.org/changeset/120017>
> > > 
> > > This caused fast/block/basic/fieldset-stretch-to-legend.html to fail. See for example <http://build.webkit.org/results/Lion%20Release%20(WebKit2%20Tests)/r120021%20(8106)/results.html>
> > 
> > This test needs its expected results updated.
> 
> Yes, it looks like I missed it during testing but the new results make sense. I will do the update as EFL needs the same treatment.

I updated the Mac expected results in <http://trac.webkit.org/r120026>.

[Julien Chaffraix]

> I updated the Mac expected results in <http://trac.webkit.org/r120026>.

And updated / skipped the rest of the platforms in http://trac.webkit.org/changeset/120027.