Optimize some code in RenderTextHarfbuzz.

Optimize some code in RenderTextHarfbuzz.

This was showing up in the UMA callstacks profiler as taking
significant time on start up. Turns out, creating and Init'ing
the BreakIterator does some icu initialization which is expensive,
e.g. ~50ms at the mean from dev UMA users.

However, we only ever need the BreakIterator if there's a grapheme
run of over 2 characters - which might be uncommon for many users.
So this change makes that initialization lazy - only done when we
actually need it.

BUG=651848
Committed: https://crrev.com/49cf5df2724445f3160b4fdf13a187295abe14fb
Cr-Commit-Position: refs/heads/master@{#422564}

[Alexei Svitkine (slow), 2016-09-30 17:58:18]

Description was changed from

==========
Optimize some code in RenderTextHarfbuzz.

This was showing up in the UMA callstacks profiler as taking
significant time on start up. Turns out, creating and Init'ing
the BreakIterator does some icu initialization which is expensive.

However, we only ever need the BreakIterator if there's a grapheme
run of over 2 characters - which might be uncommon for many users.
So this change makes that initialization lazy - only done when we
actually need it.

BUG=651848
==========

to

==========
Optimize some code in RenderTextHarfbuzz.

This was showing up in the UMA callstacks profiler as taking
significant time on start up. Turns out, creating and Init'ing
the BreakIterator does some icu initialization which is expensive,
e.g. ~50ms at the mean from dev UMA users.

However, we only ever need the BreakIterator if there's a grapheme
run of over 2 characters - which might be uncommon for many users.
So this change makes that initialization lazy - only done when we
actually need it.

BUG=651848
==========

[Alexei Svitkine (slow), 2016-09-30 18:29:03]

Patchset #1 (id:1) has been deleted

[Alexei Svitkine (slow), 2016-09-30 18:29:40]

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

[Alexei Svitkine (slow), 2016-09-30 18:29:51]

asvitkine@chromium.org changed reviewers:
+ msw@chromium.org

[commit-bot: I haz the power, 2016-09-30 18:30:25]

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

[Alexei Svitkine (slow), 2016-09-30 18:31:42]

Example data showing this taking ~50ms mean on startup:
https://uma.googleplex.com/callstacks?sid=2cdebb69b518e116060e39850c14e38f

[msw, 2016-09-30 19:10:18]

lgtm with nits; thanks!

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_harf...
File ui/gfx/render_text_harfbuzz.cc (right):

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_harf...
ui/gfx/render_text_harfbuzz.cc:649: RenderTextHarfBuzz* render_text,
nit: fits on line above (git cl format)

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_unit...
File ui/gfx/render_text_unittest.cc (right):

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_unit...
ui/gfx/render_text_unittest.cc:3261: const base::string16 kString =
ASCIIToUTF16("abcd");
optional nit: inline |render_text->SetText(ASCIIToUTF16("abcd"));|

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_unit...
ui/gfx/render_text_unittest.cc:3404: const base::string16 kString =
ASCIIToUTF16("abcdefgh");
ditto optional nit: inline |render_text->SetText(ASCIIToUTF16("abcdefgh"));|

[Alexei Svitkine (slow), 2016-09-30 19:14:17]

Thanks!

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_harf...
File ui/gfx/render_text_harfbuzz.cc (right):

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_harf...
ui/gfx/render_text_harfbuzz.cc:649: RenderTextHarfBuzz* render_text,
On 2016/09/30 19:10:17, msw wrote:
> nit: fits on line above (git cl format)

Done.

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_unit...
File ui/gfx/render_text_unittest.cc (right):

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_unit...
ui/gfx/render_text_unittest.cc:3261: const base::string16 kString =
ASCIIToUTF16("abcd");
On 2016/09/30 19:10:17, msw wrote:
> optional nit: inline |render_text->SetText(ASCIIToUTF16("abcd"));|

Done.

https://codereview.chromium.org/2379033004/diff/20001/ui/gfx/render_text_unit...
ui/gfx/render_text_unittest.cc:3404: const base::string16 kString =
ASCIIToUTF16("abcdefgh");
On 2016/09/30 19:10:17, msw wrote:
> ditto optional nit: inline |render_text->SetText(ASCIIToUTF16("abcdefgh"));|

Done.

[Alexei Svitkine (slow), 2016-09-30 19:14:50]

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

[commit-bot: I haz the power, 2016-09-30 19:15:57]

Dry run: 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, 2016-09-30 20:25:37]

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

[commit-bot: I haz the power, 2016-09-30 20:25:39]

Dry run: This issue passed the CQ dry run.

[Alexei Svitkine (slow), 2016-10-03 21:08:21]

The CQ bit was checked by asvitkine@chromium.org

[Alexei Svitkine (slow), 2016-10-03 21:08:22]

The patchset sent to the CQ was uploaded after l-g-t-m from msw@chromium.org
Link to the patchset: https://codereview.chromium.org/2379033004/#ps40001
(title: "Nits & git cl format")

[commit-bot: I haz the power, 2016-10-03 21:09:20]

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, 2016-10-03 22:25:35]

Description was changed from

==========
Optimize some code in RenderTextHarfbuzz.

This was showing up in the UMA callstacks profiler as taking
significant time on start up. Turns out, creating and Init'ing
the BreakIterator does some icu initialization which is expensive,
e.g. ~50ms at the mean from dev UMA users.

However, we only ever need the BreakIterator if there's a grapheme
run of over 2 characters - which might be uncommon for many users.
So this change makes that initialization lazy - only done when we
actually need it.

BUG=651848
==========

to

==========
Optimize some code in RenderTextHarfbuzz.

This was showing up in the UMA callstacks profiler as taking
significant time on start up. Turns out, creating and Init'ing
the BreakIterator does some icu initialization which is expensive,
e.g. ~50ms at the mean from dev UMA users.

However, we only ever need the BreakIterator if there's a grapheme
run of over 2 characters - which might be uncommon for many users.
So this change makes that initialization lazy - only done when we
actually need it.

BUG=651848
==========

[commit-bot: I haz the power, 2016-10-03 22:25:36]

Committed patchset #2 (id:40001)

[commit-bot: I haz the power, 2016-10-03 22:27:53]

Description was changed from

==========
Optimize some code in RenderTextHarfbuzz.

This was showing up in the UMA callstacks profiler as taking
significant time on start up. Turns out, creating and Init'ing
the BreakIterator does some icu initialization which is expensive,
e.g. ~50ms at the mean from dev UMA users.

However, we only ever need the BreakIterator if there's a grapheme
run of over 2 characters - which might be uncommon for many users.
So this change makes that initialization lazy - only done when we
actually need it.

BUG=651848
==========

to

==========
Optimize some code in RenderTextHarfbuzz.

This was showing up in the UMA callstacks profiler as taking
significant time on start up. Turns out, creating and Init'ing
the BreakIterator does some icu initialization which is expensive,
e.g. ~50ms at the mean from dev UMA users.

However, we only ever need the BreakIterator if there's a grapheme
run of over 2 characters - which might be uncommon for many users.
So this change makes that initialization lazy - only done when we
actually need it.

BUG=651848

Committed: https://crrev.com/49cf5df2724445f3160b4fdf13a187295abe14fb
Cr-Commit-Position: refs/heads/master@{#422564}
==========

[commit-bot: I haz the power, 2016-10-03 22:27:54]

Patchset 2 (id:??) landed as
https://crrev.com/49cf5df2724445f3160b4fdf13a187295abe14fb
Cr-Commit-Position: refs/heads/master@{#422564}