Reduce infobar lag on KitKat from 1000ms to 100ms.

Reduce infobar lag on KitKat from 1000ms to 100ms.

This fixes four compounding inefficiencies that caused infobar layouts
to take waaaay longer than they should have (over 1 second in the direst
cases!). The worst manifestation of this problem could be seen by
opening the change-language panel on the translate infobar on a pre-L
device. This resulted in 3 300ms measures as well as 250ms spent
measuring text.

The fixes are:
 - InfoBarLayout was measuring each InfoBarControlLayout twice: once
   with UNSPECIFIED width and once with an exact width. Since
   InfoBarControlLayouts handle layout logic internally, they should
   only be measured once, with an exact width.
 - The infobar view's width was WRAP_CONTENT, which caused its parent,
   the wrapper view, to measure it twice. This is fixed by setting
   the width to MATCH_PARENT.
 - TranslateLanguagePanel.onItemSelected() is called twice when the
   translate panel is first created to set the initial values of the
   spinners. This resulted in two superfluous measure/layout/draw
   passes. As a fix, TranslateLanguagePanel.onItemSelected() only calls
   reloadSpinners() if the selected language has actually changed.
 - TranslateLanguagePanel.setUpSpinners() took 250ms to determine the
   width of the longest language. This CL makes this measuring logic
   smarter and avoids redundant work in two different ways: it only
   measures each language once (not once per spinner), and it only
   measures "Page language: " and "Translation language: " once.
   (instead of once per language per spinner = ~140 times each)

BUG=564406
Committed: https://crrev.com/6e4b88e29b0968427fdaa5e70307a10845409452
Cr-Commit-Position: refs/heads/master@{#367464}

[newt (away), 2016-01-05 00:20:04]

newt@chromium.org changed reviewers:
+ dfalcantara@chromium.org

[newt (away), 2016-01-05 00:20:05]

PTAL

[gone, 2016-01-05 00:41:53]

lgtm

[newt (away), 2016-01-05 00:42:20]

The CQ bit was checked by newt@chromium.org

[commit-bot: I haz the power, 2016-01-05 00:42:51]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1560573002/1
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1560573002/1

[commit-bot: I haz the power, 2016-01-05 01:46:51]

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

[commit-bot: I haz the power, 2016-01-05 01:46:51]

Try jobs failed on following builders:
  android_clang_dbg_recipe on tryserver.chromium.android (JOB_FAILED,
https://build.chromium.org/p/tryserver.chromium.android/builders/android_clan...)

[newt (away), 2016-01-05 02:05:21]

The CQ bit was checked by newt@chromium.org

[newt (away), 2016-01-05 02:05:22]

The patchset sent to the CQ was uploaded after l-g-t-m from
dfalcantara@chromium.org
Link to the patchset: https://codereview.chromium.org/1560573002/#ps20001
(title: "appease the FindBugs gods")

[commit-bot: I haz the power, 2016-01-05 02:06:27]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1560573002/20001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1560573002/20001

[commit-bot: I haz the power, 2016-01-05 02:44:01]

Committed patchset #2 (id:20001)

[commit-bot: I haz the power, 2016-01-05 02:45:39]

Description was changed from

==========
Reduce infobar lag on KitKat from 1000ms to 100ms.

This fixes four compounding inefficiencies that caused infobar layouts
to take waaaay longer than they should have (over 1 second in the direst
cases!). The worst manifestation of this problem could be seen by
opening the change-language panel on the translate infobar on a pre-L
device. This resulted in 3 300ms measures as well as 250ms spent
measuring text.

The fixes are:
 - InfoBarLayout was measuring each InfoBarControlLayout twice: once
   with UNSPECIFIED width and once with an exact width. Since
   InfoBarControlLayouts handle layout logic internally, they should
   only be measured once, with an exact width.
 - The infobar view's width was WRAP_CONTENT, which caused its parent,
   the wrapper view, to measure it twice. This is fixed by setting
   the width to MATCH_PARENT.
 - TranslateLanguagePanel.onItemSelected() is called twice when the
   translate panel is first created to set the initial values of the
   spinners. This resulted in two superfluous measure/layout/draw
   passes. As a fix, TranslateLanguagePanel.onItemSelected() only calls
   reloadSpinners() if the selected language has actually changed.
 - TranslateLanguagePanel.setUpSpinners() took 250ms to determine the
   width of the longest language. This CL makes this measuring logic
   smarter and avoids redundant work in two different ways: it only
   measures each language once (not once per spinner), and it only
   measures "Page language: " and "Translation language: " once.
   (instead of once per language per spinner = ~140 times each)

BUG=564406
==========

to

==========
Reduce infobar lag on KitKat from 1000ms to 100ms.

This fixes four compounding inefficiencies that caused infobar layouts
to take waaaay longer than they should have (over 1 second in the direst
cases!). The worst manifestation of this problem could be seen by
opening the change-language panel on the translate infobar on a pre-L
device. This resulted in 3 300ms measures as well as 250ms spent
measuring text.

The fixes are:
 - InfoBarLayout was measuring each InfoBarControlLayout twice: once
   with UNSPECIFIED width and once with an exact width. Since
   InfoBarControlLayouts handle layout logic internally, they should
   only be measured once, with an exact width.
 - The infobar view's width was WRAP_CONTENT, which caused its parent,
   the wrapper view, to measure it twice. This is fixed by setting
   the width to MATCH_PARENT.
 - TranslateLanguagePanel.onItemSelected() is called twice when the
   translate panel is first created to set the initial values of the
   spinners. This resulted in two superfluous measure/layout/draw
   passes. As a fix, TranslateLanguagePanel.onItemSelected() only calls
   reloadSpinners() if the selected language has actually changed.
 - TranslateLanguagePanel.setUpSpinners() took 250ms to determine the
   width of the longest language. This CL makes this measuring logic
   smarter and avoids redundant work in two different ways: it only
   measures each language once (not once per spinner), and it only
   measures "Page language: " and "Translation language: " once.
   (instead of once per language per spinner = ~140 times each)

BUG=564406

Committed: https://crrev.com/6e4b88e29b0968427fdaa5e70307a10845409452
Cr-Commit-Position: refs/heads/master@{#367464}
==========

[commit-bot: I haz the power, 2016-01-05 02:45:39]

Patchset 2 (id:??) landed as
https://crrev.com/6e4b88e29b0968427fdaa5e70307a10845409452
Cr-Commit-Position: refs/heads/master@{#367464}