Roll HarfBuzz to 1.4.1

Roll HarfBuzz to 1.4.1

Includes support for specifying OpenType 1.8 font variations axis
parameters.

Baseline updates due to expected advance width differences for some
Khmer combining marks, as well as combining marks placement in
combination with ligatures, due to HarfBuzz commit 1a0f4aa6dfee69236b4.

BUG=679305
Review-Url: https://codereview.chromium.org/2622553002
Cr-Commit-Position: refs/heads/master@{#444707}
Committed: https://chromium.googlesource.com/chromium/src/+/f2dc38825cce54c0b87841a420afbd71453d506b

[drott, 2017-01-09 12:00:22]

The CQ bit was checked by drott@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-09 12:00:37]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[drott, 2017-01-09 12:01:11]

drott@chromium.org changed reviewers:
+ behdad@chromium.org, eae@chromium.org, kojii@chromium.org

[drott, 2017-01-09 13:28:59]

Behdad, are the Khmer clustering differences expected?
https://storage.googleapis.com/chromium-layout-test-archives/linux_chromium_r...

[commit-bot: I haz the power, 2017-01-09 14:23:32]

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

[commit-bot: I haz the power, 2017-01-09 14:23:33]

Dry run: Try jobs failed on following builders:
  linux_chromium_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[drott, 2017-01-09 14:32:42]

There is also a small width difference with the "composed chars" test:
https://storage.googleapis.com/chromium-layout-test-archives/linux_chromium_r...
Is that explainable / expected?

Thanks for taking a look.

[eae, 2017-01-09 17:17:27]

LGTM

[behdad_google, 2017-01-09 23:25:53]

Dominik,

I'm curious what causes those two changes, but none look like a blocker.

What was the previous version of HarfBuzz we had?  If you can reproduce those
with hb-view / hb-shape, I can bisect and make sure the changes are expected. 
Khmer one is with tofu boxes, so that should be easy to reproduce.  The other
one depends on the exact version of the font used.

[eae, 2017-01-10 23:21:52]

On 2017/01/09 23:25:53, behdad_google wrote:
> Dominik,
> 
> I'm curious what causes those two changes, but none look like a blocker.
> 
> What was the previous version of HarfBuzz we had? 
1.3.4

[drott, 2017-01-18 14:05:42]

The CQ bit was checked by drott@chromium.org to run a CQ dry run

[commit-bot: I haz the power, 2017-01-18 14:06:06]

Dry run: CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[drott, 2017-01-18 14:38:37]

The first Khmer difference from 
fast/text/international/khmer-selection.html is that KHMER SIGN MUUSIKATOAN has
a non-zero advance with as a .notdef from DejaVu Sans after, but does not have
an advance with before.

Bisecting this leads to commit:

commit 1a0f4aa6dfee69236b422f7c8825eff940f45c19
Author: Behdad Esfahbod <behdad@behdad.org>
Date:   Thu Dec 22 13:33:54 2016 -0600

    Revert "Synthesize GDEF glyph class for any glyph that does not have one in
GDEF"

Before this commit, a HarfBuzz reduction look like this:

util/hb-shape /usr/share/fonts/truetype/ttf-dejavu/DejaVuSans.ttf
"`test/shaping/hb-unicode-encode U+179C,U+17C9,U+17C2,U+1794,U+0020`"
[.notdef=0+1229|.notdef=0+0|.notdef=0+1229|.notdef=3+1229|space=4+651]

Afterwards:

util/hb-shape /usr/share/fonts/truetype/ttf-dejavu/DejaVuSans.ttf
"`test/shaping/hb-unicode-encode U+179C,U+17C9,U+17C2,U+1794,U+0020`"
[.notdef=0+1229|.notdef=0+1229|.notdef=0+1229|.notdef=3+1229|space=4+651]

The second character gets an advance width of 1229, while it did not have one
before.

[drott, 2017-01-18 14:53:43]

Bisecting the other test difference for "Af̀fiZ" identifies the exact same
commit 1a0f4aa6dfee692:

Before (looking wider between the two f's):
util/hb-shape /usr/share/fonts/truetype/msttcorefonts/Times_New_Roman.ttf
"Af̀fiZ"
[A=0+1479|f=1+682|combininggraveaccent=1@130,57+0|f=3+682|i=4+569|Z=5+1251]

After (looking narrower - or too narrow? - between the two f's):
util/hb-shape /usr/share/fonts/truetype/msttcorefonts/Times_New_Roman.ttf
"Af̀fiZ"
[A=0+1479|f=1+663|combininggraveaccent=1@130,57+0|f=3@-18,0+664|i=4+569|Z=5+1251]

[commit-bot: I haz the power, 2017-01-18 16:30:30]

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

[commit-bot: I haz the power, 2017-01-18 16:30:31]

Dry run: Try jobs failed on following builders:
  linux_chromium_rel_ng on master.tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/linux_chromium_...)

[chromium-reviews, 2017-01-18 20:04:47]

Ok, thanks.  That's expected.

On Wed, Jan 18, 2017 at 6:53 AM, <drott@chromium.org> wrote:

> Bisecting the other test difference for "Af̀fiZ" identifies the exact same
> commit 1a0f4aa6dfee692:
>
> Before (looking wider between the two f's):
> util/hb-shape /usr/share/fonts/truetype/msttcorefonts/Times_New_Roman.ttf
> "Af̀fiZ"
> [A=0+1479|f=1+682|combininggraveaccent=1@130,57+
> 0|f=3+682|i=4+569|Z=5+1251]
>
> After (looking narrower - or too narrow? - between the two f's):
> util/hb-shape /usr/share/fonts/truetype/msttcorefonts/Times_New_Roman.ttf
> "Af̀fiZ"
> [A=0+1479|f=1+663|combininggraveaccent=1@130,57+0|f=3@
> -18,0+664|i=4+569|Z=5+1251]
>
>
>
> https://codereview.chromium.org/2622553002/
>

-- 
You received this message because you are subscribed to the Google Groups
"Chromium-reviews" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to chromium-reviews+unsubscribe@chromium.org.

[drott, 2017-01-19 09:12:48]

Linux rebaselines

[drott, 2017-01-19 09:14:40]

Description was changed from

==========
Roll HarfBuzz to 1.4.1

Includes support for specifying OpenType 1.8 font variations axis
parameters.

BUG=679305
==========

to

==========
Roll HarfBuzz to 1.4.1

Includes support for specifying OpenType 1.8 font variations axis
parameters.

Baseline updates due to expected advance width differences for some
Khmer combining marks, as well as combining marks placement in
combination with ligatures, due to HarfBuzz commit 1a0f4aa6dfee69236b4.

BUG=679305
==========

[drott, 2017-01-19 09:14:43]

The CQ bit was checked by drott@chromium.org

[drott, 2017-01-19 09:14:43]

The patchset sent to the CQ was uploaded after l-g-t-m from eae@chromium.org
Link to the patchset: https://codereview.chromium.org/2622553002/#ps20001
(title: "Linux rebaselines")

[commit-bot: I haz the power, 2017-01-19 09:14:56]

CQ is trying da patch. Follow status at
 
https://chromium-cq-status.appspot.com/v2/patch-status/codereview.chromium.or...

[commit-bot: I haz the power, 2017-01-19 11:28:14]

CQ is committing da patch.

Bot data: {"patchset_id": 20001, "attempt_start_ts": 1484817283251640,
"parent_rev": "cb466e8dc98de4c73e7010187f98283dceacd17f", "commit_rev":
"f2dc38825cce54c0b87841a420afbd71453d506b"}

[commit-bot: I haz the power, 2017-01-19 11:28:44]

Description was changed from

==========
Roll HarfBuzz to 1.4.1

Includes support for specifying OpenType 1.8 font variations axis
parameters.

Baseline updates due to expected advance width differences for some
Khmer combining marks, as well as combining marks placement in
combination with ligatures, due to HarfBuzz commit 1a0f4aa6dfee69236b4.

BUG=679305
==========

to

==========
Roll HarfBuzz to 1.4.1

Includes support for specifying OpenType 1.8 font variations axis
parameters.

Baseline updates due to expected advance width differences for some
Khmer combining marks, as well as combining marks placement in
combination with ligatures, due to HarfBuzz commit 1a0f4aa6dfee69236b4.

BUG=679305

Review-Url: https://codereview.chromium.org/2622553002
Cr-Commit-Position: refs/heads/master@{#444707}
Committed:
https://chromium.googlesource.com/chromium/src/+/f2dc38825cce54c0b87841a420af...
==========

[commit-bot: I haz the power, 2017-01-19 11:28:46]

Committed patchset #2 (id:20001) as
https://chromium.googlesource.com/chromium/src/+/f2dc38825cce54c0b87841a420af...