Don't pixel snap for ellipsis

Don't pixel snap for ellipsis

This is a rough port of Zalan Bujtas' patch on https://bugs.webkit.org/show_bug.cgi?id=112227

Due to pixel snapping/truncating, we prematurely truncate some rtl content when
text-overflow: ellipsis is enabled. One piece of this puzzle is a hack to fix old
multi-column code, which didn't work well with sub-pixel layout. The other
is unnecessary pixel snapping in ellipsis line layout. Both of these issues need to
be fixed to address the bug in rtl ellipsis content.

This changes old-multicolumn layout by one pixel when compared to the (better)
region-based multicolumn on 2 ref tests.

BUG=378489
Committed: https://src.chromium.org/viewvc/blink?view=rev&revision=175434

[leviw_travelin_and_unemployed, 2014-05-30 18:35:40]

This breaks a pair of ref tests that compare old-multicolumn vs
region-multicolumn. A box moves over by 1 pixel in the old-multicolumn case. I
propose disabling them, but am open to other options.

[eae, 2014-05-30 21:08:51]

On 2014/05/30 18:35:40, leviw wrote:
> This breaks a pair of ref tests that compare old-multicolumn vs
> region-multicolumn. A box moves over by 1 pixel in the old-multicolumn case. I
> propose disabling them, but am open to other options.

I'd argue we should skip those tests as long as the new multicolum case works
fine.

LGTM

[mstensho (USE GERRIT), 2014-05-30 21:11:56]

On 2014/05/30 18:35:40, leviw wrote:
> This breaks a pair of ref tests that compare old-multicolumn vs
> region-multicolumn. A box moves over by 1 pixel in the old-multicolumn case. I
> propose disabling them, but am open to other options.

Sounds good to me. Just disable them. The multicol reftests that just compare
the old implementation against the new one will have to be replaced with
something better soon anyway, once we make it so that
--enable-experimental-web-platform-features enables the new implementation.

[leviw_travelin_and_unemployed, 2014-05-30 21:14:11]

On 2014/05/30 21:11:56, mstensho - back 2014-08-19 wrote:
> On 2014/05/30 18:35:40, leviw wrote:
> > This breaks a pair of ref tests that compare old-multicolumn vs
> > region-multicolumn. A box moves over by 1 pixel in the old-multicolumn case.
I
> > propose disabling them, but am open to other options.
> 
> Sounds good to me. Just disable them. The multicol reftests that just compare
> the old implementation against the new one will have to be replaced with
> something better soon anyway, once we make it so that
> --enable-experimental-web-platform-features enables the new implementation.

Thanks! I'll disable them :)

[leviw_travelin_and_unemployed, 2014-06-02 20:25:17]

This breaks a shape-outside ref test (by a pixel) on Linux and Windows, too...
Any advice from the shape outside guys?
https://storage.googleapis.com/chromium-layout-test-archives/linux_blink_dbg/...

[Zoltan, 2014-06-02 21:25:55]

On 2014/06/02 20:25:17, leviw wrote:
> This breaks a shape-outside ref test (by a pixel) on Linux and Windows, too...
> Any advice from the shape outside guys?
>
https://storage.googleapis.com/chromium-layout-test-archives/linux_blink_dbg/...

You'll need to update some shapes tests too:
http://trac.webkit.org/changeset/169048/trunk/LayoutTests/ChangeLog

fast/shapes/shape-outside-floats/shape-outside-floats-ellipse-margin-left.html:
fast/shapes/shape-outside-floats/shape-outside-floats-ellipse-margin-right.html:
fast/shapes/shape-outside-floats/shape-outside-floats-inset-rounded-different-writing-modes-left.html:
fast/shapes/shape-outside-floats/shape-outside-floats-inset-rounded-different-writing-modes-right.html:
fast/shapes/shape-outside-floats/shape-outside-rounded-boxes-001.html:
fast/shapes/shape-outside-floats/shape-outside-rounded-boxes-002.html:
fast/shapes/shape-outside-floats/shape-outside-rounded-inset-expected.html:

[leviw_travelin_and_unemployed, 2014-06-02 21:27:30]

Let me be more specific: How can I fix the ref tests?

[leviw_travelin_and_unemployed, 2014-06-03 17:40:11]

Looks like everything is peachy if I skip the 2 ref tests. I filed a bug about
the test issues. Anyone object to me moving forward on this approach?

[eae, 2014-06-03 17:41:55]

On 2014/06/03 17:40:11, leviw wrote:
> Looks like everything is peachy if I skip the 2 ref tests. I filed a bug about
> the test issues. Anyone object to me moving forward on this approach?

No objections from me.

[Zoltan, 2014-06-03 17:44:29]

On 2014/06/03 17:41:55, eae wrote:
> On 2014/06/03 17:40:11, leviw wrote:
> > Looks like everything is peachy if I skip the 2 ref tests. I filed a bug
about
> > the test issues. Anyone object to me moving forward on this approach?
> 
> No objections from me.

It's not clear at this point what causes the difference on Linux/Win, so let's
keep the patch rolling and we can investigate the skipped cases separately.

[leviw_travelin_and_unemployed, 2014-06-03 17:47:27]

The CQ bit was checked by leviw@chromium.org

[commit-bot: I haz the power, 2014-06-03 17:47:55]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/leviw@chromium.org/303673007/120001

[leviw_travelin_and_unemployed, 2014-06-03 18:07:19]

The CQ bit was checked by leviw@chromium.org

[commit-bot: I haz the power, 2014-06-03 18:07:45]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/leviw@chromium.org/303673007/140001

[commit-bot: I haz the power, 2014-06-03 19:20:38]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  win_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/win_blink_rel/builds/10424)

[commit-bot: I haz the power, 2014-06-03 19:48:37]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-06-03 19:48:37]

Try jobs failed on following builders:
  win_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/win_blink_rel/builds/10438)

[leviw_travelin_and_unemployed, 2014-06-03 23:05:12]

The CQ bit was checked by leviw@chromium.org

[commit-bot: I haz the power, 2014-06-03 23:06:02]

CQ is trying da patch. Follow status at
 https://chromium-status.appspot.com/cq/leviw@chromium.org/303673007/160001

[commit-bot: I haz the power, 2014-06-04 02:29:42]

FYI, CQ is re-trying this CL (attempt #1).
The failing builders are:
  linux_blink_rel on tryserver.blink
(http://build.chromium.org/p/tryserver.blink/builders/linux_blink_rel/builds/1...)

[commit-bot: I haz the power, 2014-06-04 03:10:15]

Change committed as 175434