Implement chrome.input.ime.setComposition/commitText API.

Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from ui::IMEInputContextHandlerInterface, thus we can implement functions CommitText() and UpdateCompositionText() once for Linux and Windows.

Make InputImeKeyEventHandledFunction::Run() calls the new added InputMethodBase::KeyEventHandler() in  to set |handling_key_event_|, which indicates whether the IME extension is handling key event. And when |handling_key_event_| == true, we should not setComposition and commitText until it's done to run the callback.

Add helper function UpdateCommitText() and CommitTextToInputContext() in InputMethodEngineBase, which will be override in chromeos::InputMehtodEngine and input_method::InputMethodEngine to implement the API on different platforms.

BUG=517773
TEST=None

Committed: https://crrev.com/db5505268890037e5af265049ab64f79c01ef162
Cr-Commit-Position: refs/heads/master@{#376095}

[Azure Wei, 2016-02-03 08:16:08]

Description was changed from

==========
Implement chrome.input.ime.setComposition API.

BUG=517773
TEST=None
==========

to

==========
Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs
on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from
ui::IMEInputContextHandlerInterface, thus we can implement functions
CommitText() and UpdateCompositionText() once for Linux and Windows.

BUG=517773
TEST=None
==========

[Azure Wei, 2016-02-03 08:16:08]

azurewei@chromium.org changed reviewers:
+ rdevlin.cronin@chromium.org, shuchen@chromium.org

[Azure Wei, 2016-02-03 08:16:48]

Shu and Devlin, can you please review this cl? Thank you!

[Shu Chen, 2016-02-03 09:22:29]

https://codereview.chromium.org/1657593007/diff/40001/ui/base/ime/input_metho...
File ui/base/ime/input_method_base.h (right):

https://codereview.chromium.org/1657593007/diff/40001/ui/base/ime/input_metho...
ui/base/ime/input_method_base.h:107: bool handling_key_event_;
Can this be moved to InputMethodEngineBase?

[Shu Chen, 2016-02-03 10:05:42]

https://codereview.chromium.org/1657593007/diff/40001/ui/base/ime/input_metho...
File ui/base/ime/input_method_base.h (right):

https://codereview.chromium.org/1657593007/diff/40001/ui/base/ime/input_metho...
ui/base/ime/input_method_base.h:107: bool handling_key_event_;
On 2016/02/03 09:22:29, Shu Chen wrote:
> Can this be moved to InputMethodEngineBase?

Per offline discussion, let improve the code structure a little bit:
InputImeKeyEventHandledFunction::Run() should NOT directly call
InputImeEventRouterBase::OnKeyEventHandled(), instead, it should call
InputMethodEngineBase::KeyEventHandled() method (newly created method), which
bridges the call to InputImeEventRouterBase::OnKeyEventHandled().
In this way, |handling_key_event_| can be moved to InputMethodEngineBase:
 - InputMethodEngineBase::ProcessKeyEvent(): |handling_key_event_| = true;
 - InputMethodEngineBase::KeyEventHandled(): |handling_key_event_| = false;
 - InputMethodEngineBase::CommitText()/etc.: check |handling_key_event_| as
appropriate.

This solution will minimize the impact to ui/base/ime, thus reduce the risk.

Thanks!

[Azure Wei, 2016-02-03 10:19:43]

On 2016/02/03 10:05:42, Shu Chen wrote:
>
https://codereview.chromium.org/1657593007/diff/40001/ui/base/ime/input_metho...
> File ui/base/ime/input_method_base.h (right):
> 
>
https://codereview.chromium.org/1657593007/diff/40001/ui/base/ime/input_metho...
> ui/base/ime/input_method_base.h:107: bool handling_key_event_;
> On 2016/02/03 09:22:29, Shu Chen wrote:
> > Can this be moved to InputMethodEngineBase?
> 
> Per offline discussion, let improve the code structure a little bit:
> InputImeKeyEventHandledFunction::Run() should NOT directly call
> InputImeEventRouterBase::OnKeyEventHandled(), instead, it should call
> InputMethodEngineBase::KeyEventHandled() method (newly created method), which
> bridges the call to InputImeEventRouterBase::OnKeyEventHandled().
> In this way, |handling_key_event_| can be moved to InputMethodEngineBase:
>  - InputMethodEngineBase::ProcessKeyEvent(): |handling_key_event_| = true;
>  - InputMethodEngineBase::KeyEventHandled(): |handling_key_event_| = false;
>  - InputMethodEngineBase::CommitText()/etc.: check |handling_key_event_| as
> appropriate.
> 
> This solution will minimize the impact to ui/base/ime, thus reduce the risk.
> 
> Thanks!

I'm working on it.

[Azure Wei, 2016-02-03 12:08:32]

On 2016/02/03 10:05:42, Shu Chen wrote:
>
https://codereview.chromium.org/1657593007/diff/40001/ui/base/ime/input_metho...
> File ui/base/ime/input_method_base.h (right):
> 
>
https://codereview.chromium.org/1657593007/diff/40001/ui/base/ime/input_metho...
> ui/base/ime/input_method_base.h:107: bool handling_key_event_;
> On 2016/02/03 09:22:29, Shu Chen wrote:
> > Can this be moved to InputMethodEngineBase?
> 
> Per offline discussion, let improve the code structure a little bit:
> InputImeKeyEventHandledFunction::Run() should NOT directly call
> InputImeEventRouterBase::OnKeyEventHandled(), instead, it should call
> InputMethodEngineBase::KeyEventHandled() method (newly created method), which
> bridges the call to InputImeEventRouterBase::OnKeyEventHandled().
> In this way, |handling_key_event_| can be moved to InputMethodEngineBase:
>  - InputMethodEngineBase::ProcessKeyEvent(): |handling_key_event_| = true;
>  - InputMethodEngineBase::KeyEventHandled(): |handling_key_event_| = false;
>  - InputMethodEngineBase::CommitText()/etc.: check |handling_key_event_| as
> appropriate.
> 
> This solution will minimize the impact to ui/base/ime, thus reduce the risk.
> 
> Thanks!

Please review the latest patch set! Thank you!

[Azure Wei, 2016-02-03 12:51:13]

Description was changed from

==========
Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs
on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from
ui::IMEInputContextHandlerInterface, thus we can implement functions
CommitText() and UpdateCompositionText() once for Linux and Windows.

BUG=517773
TEST=None
==========

to

==========
Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs
on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from
ui::IMEInputContextHandlerInterface, thus we can implement functions
CommitText() and UpdateCompositionText() once for Linux and Windows.

Call the new added InputMethodBase::KeyEventHandler() in
InputImeKeyEventHandledFunction::Run() to set |handling_key_event_|, which
indicates whether the IME extension is handling key event. And when
|handling_key_event_| == true, we should not setComposition and commitText until
it's done to run the callback.

Add helper function UpdateCommitText() and CommitTextToInputContext() in
InputMethodEngineBase, which will be override in chromeos::InputMehtodEngine and
input_method::InputMethodEngine to implement the API on different platforms.


BUG=517773
TEST=None
==========

[Azure Wei, 2016-02-03 12:52:11]

Description was changed from

==========
Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs
on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from
ui::IMEInputContextHandlerInterface, thus we can implement functions
CommitText() and UpdateCompositionText() once for Linux and Windows.

Call the new added InputMethodBase::KeyEventHandler() in
InputImeKeyEventHandledFunction::Run() to set |handling_key_event_|, which
indicates whether the IME extension is handling key event. And when
|handling_key_event_| == true, we should not setComposition and commitText until
it's done to run the callback.

Add helper function UpdateCommitText() and CommitTextToInputContext() in
InputMethodEngineBase, which will be override in chromeos::InputMehtodEngine and
input_method::InputMethodEngine to implement the API on different platforms.


BUG=517773
TEST=None
==========

to

==========
Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs
on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from
ui::IMEInputContextHandlerInterface, thus we can implement functions
CommitText() and UpdateCompositionText() once for Linux and Windows.

Make InputImeKeyEventHandledFunction::Run() calls the new added
InputMethodBase::KeyEventHandler() in  to set |handling_key_event_|, which
indicates whether the IME extension is handling key event. And when
|handling_key_event_| == true, we should not setComposition and commitText until
it's done to run the callback.

Add helper function UpdateCommitText() and CommitTextToInputContext() in
InputMethodEngineBase, which will be override in chromeos::InputMehtodEngine and
input_method::InputMethodEngine to implement the API on different platforms.


BUG=517773
TEST=None
==========

[Devlin, 2016-02-04 00:09:17]

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/chromeos...
File chrome/browser/chromeos/input_method/input_method_engine.cc (right):

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/chromeos...
chrome/browser/chromeos/input_method/input_method_engine.cc:59: size_t
GetUtf8StringLength(const char* s) {
Is this *really* that much faster than std::string(s).length()?  Is it in
extremely performance-sensitive code?  If not, let's avoid doing this (note that
it'd also be avoided if we can pass in a string to CommitTextToInputContext)

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/chromeos...
File chrome/browser/chromeos/input_method/input_method_engine.h (right):

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/chromeos...
chrome/browser/chromeos/input_method/input_method_engine.h:158: void
CommitTextToInputContext(int context_id, const char* text) override;
Is there a reason this has to be a const char* instead of a const std::string&?

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/extensio...
File chrome/browser/extensions/api/input_ime/input_ime_api.cc (right):

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/extensio...
chrome/browser/extensions/api/input_ime/input_ime_api.cc:266:
base::DictionaryValue* output = new base::DictionaryValue();
prefer scoped ptr generally, but also we should just be using
SetComposition::Results here.

[Azure Wei, 2016-02-04 03:21:19]

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/chromeos...
File chrome/browser/chromeos/input_method/input_method_engine.cc (right):

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/chromeos...
chrome/browser/chromeos/input_method/input_method_engine.cc:59: size_t
GetUtf8StringLength(const char* s) {
On 2016/02/04 00:09:17, Devlin wrote:
> Is this *really* that much faster than std::string(s).length()?  Is it in
> extremely performance-sensitive code?  If not, let's avoid doing this (note
that
> it'd also be avoided if we can pass in a string to CommitTextToInputContext)

Abandoned these codes.

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/chromeos...
File chrome/browser/chromeos/input_method/input_method_engine.h (right):

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/chromeos...
chrome/browser/chromeos/input_method/input_method_engine.h:158: void
CommitTextToInputContext(int context_id, const char* text) override;
On 2016/02/04 00:09:17, Devlin wrote:
> Is there a reason this has to be a const char* instead of a const
std::string&?

Yeah, it's more convenient to use const std::string&. Thanks.

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/extensio...
File chrome/browser/extensions/api/input_ime/input_ime_api.cc (right):

https://codereview.chromium.org/1657593007/diff/80001/chrome/browser/extensio...
chrome/browser/extensions/api/input_ime/input_ime_api.cc:266:
base::DictionaryValue* output = new base::DictionaryValue();
On 2016/02/04 00:09:17, Devlin wrote:
> prefer scoped ptr generally, but also we should just be using
> SetComposition::Results here.

Done.

[Shu Chen, 2016-02-04 08:00:47]

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/extensi...
File chrome/browser/extensions/api/input_ime/input_ime_api.cc (right):

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/extensi...
chrome/browser/extensions/api/input_ime/input_ime_api.cc:253:
engine->KeyEventHandled();
Please merge the below OnKeyEventHandled() call into this KeyEventHandled()
method.

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/ui/inpu...
File chrome/browser/ui/input_method/input_method_engine_base.cc (right):

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/ui/inpu...
chrome/browser/ui/input_method/input_method_engine_base.cc:159: profile_(NULL),
s/NULL/nullptr/g

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/ui/inpu...
chrome/browser/ui/input_method/input_method_engine_base.cc:386: if
(input_context)
bracket.

[Azure Wei, 2016-02-04 11:57:21]

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/extensi...
File chrome/browser/extensions/api/input_ime/input_ime_api.cc (right):

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/extensi...
chrome/browser/extensions/api/input_ime/input_ime_api.cc:253:
engine->KeyEventHandled();
On 2016/02/04 08:00:46, Shu Chen (OOO till Feb 14) wrote:
> Please merge the below OnKeyEventHandled() call into this KeyEventHandled()
> method.

Done. Issue 1664513008 was created to move all these keyEventHandled related
things to InputMethodEngineBase::KeyEventHandled(). The new patch set was
rebased on that cl.

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/ui/inpu...
File chrome/browser/ui/input_method/input_method_engine_base.cc (right):

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/ui/inpu...
chrome/browser/ui/input_method/input_method_engine_base.cc:159: profile_(NULL),
On 2016/02/04 08:00:46, Shu Chen (OOO till Feb 14) wrote:
> s/NULL/nullptr/g

Done.

https://codereview.chromium.org/1657593007/diff/100001/chrome/browser/ui/inpu...
chrome/browser/ui/input_method/input_method_engine_base.cc:386: if
(input_context)
On 2016/02/04 08:00:46, Shu Chen (OOO till Feb 14) wrote:
> bracket.

Done.

[Shu Chen, 2016-02-04 12:54:23]

lgtm

[Devlin, 2016-02-10 18:11:21]

lgtm

[Azure Wei, 2016-02-15 01:36:42]

The CQ bit was checked by azurewei@chromium.org

[Azure Wei, 2016-02-15 01:36:42]

The patchset sent to the CQ was uploaded after l-g-t-m from
shuchen@chromium.org, rdevlin.cronin@chromium.org
Link to the patchset: https://codereview.chromium.org/1657593007/#ps140001
(title: "Sync up to date.")

[commit-bot: I haz the power, 2016-02-15 01:36:50]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1657593007/140001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1657593007/140001

[commit-bot: I haz the power, 2016-02-15 02:11:48]

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

[commit-bot: I haz the power, 2016-02-15 02:11:49]

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

[Azure Wei, 2016-02-15 05:16:15]

The CQ bit was checked by azurewei@chromium.org

[Azure Wei, 2016-02-15 05:16:16]

The patchset sent to the CQ was uploaded after l-g-t-m from
shuchen@chromium.org, rdevlin.cronin@chromium.org
Link to the patchset: https://codereview.chromium.org/1657593007/#ps160001
(title: "Fix test error.")

[commit-bot: I haz the power, 2016-02-15 05:16:37]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1657593007/160001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1657593007/160001

[commit-bot: I haz the power, 2016-02-15 05:24:54]

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

[commit-bot: I haz the power, 2016-02-15 05:24:55]

Try jobs failed on following builders:
  chromium_presubmit on tryserver.chromium.linux (JOB_FAILED,
http://build.chromium.org/p/tryserver.chromium.linux/builders/chromium_presub...)

[Azure Wei, 2016-02-15 05:47:46]

The CQ bit was checked by azurewei@chromium.org

[Azure Wei, 2016-02-15 05:47:47]

The patchset sent to the CQ was uploaded after l-g-t-m from
shuchen@chromium.org, rdevlin.cronin@chromium.org
Link to the patchset: https://codereview.chromium.org/1657593007/#ps180001
(title: "Fix presubmit warning.")

[commit-bot: I haz the power, 2016-02-15 05:48:12]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1657593007/180001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1657593007/180001

[commit-bot: I haz the power, 2016-02-15 06:50:30]

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

[commit-bot: I haz the power, 2016-02-15 06:50:31]

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

[Azure Wei, 2016-02-15 08:12:49]

The CQ bit was checked by azurewei@chromium.org

[Azure Wei, 2016-02-15 08:12:49]

The patchset sent to the CQ was uploaded after l-g-t-m from
shuchen@chromium.org, rdevlin.cronin@chromium.org
Link to the patchset: https://codereview.chromium.org/1657593007/#ps200001
(title: " ")

[commit-bot: I haz the power, 2016-02-15 08:13:36]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1657593007/200001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1657593007/200001

[commit-bot: I haz the power, 2016-02-15 09:30:04]

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

[commit-bot: I haz the power, 2016-02-15 09:30:05]

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

[Azure Wei, 2016-02-15 09:42:02]

Hi Devlin, I removed the function GetUtf8StringLength() and used
std::string(s).length() as you suggested since patch 6 (in function
CommitTextToInputClient in
chrome/browser/chromeos/input_method/input_method_engine.cc).
While, std::string(s).length() returns the length of bytes rather than
characters. Like '入力' (\xE5\x85\xA5\xE5\x8A\x9B) in unit_tests with
gtest_filter=InputMethodEngineTest.TestHistograms, std::string(s).length() gets
6 rather than 2.
I tried to use functions base::UTF8ToWide() and std::wstring(s).length(), it
failed pre_submit checking with warning 'New code should not use wstrings.' and
browser_tests (see patch 9 & 10). 
So, is it OK to add the function GetUtf8StringLength() back, or are there any
other solutions?

[Devlin, 2016-02-16 17:47:25]

On 2016/02/15 09:42:02, Azure Wei wrote:
> Hi Devlin, I removed the function GetUtf8StringLength() and used
> std::string(s).length() as you suggested since patch 6 (in function
> CommitTextToInputClient in
> chrome/browser/chromeos/input_method/input_method_engine.cc).
> While, std::string(s).length() returns the length of bytes rather than
> characters. Like '入力' (\xE5\x85\xA5\xE5\x8A\x9B) in unit_tests with
> gtest_filter=InputMethodEngineTest.TestHistograms, std::string(s).length()
gets
> 6 rather than 2.
> I tried to use functions base::UTF8ToWide() and std::wstring(s).length(), it
> failed pre_submit checking with warning 'New code should not use wstrings.'
and
> browser_tests (see patch 9 & 10). 
> So, is it OK to add the function GetUtf8StringLength() back, or are there any
> other solutions?

What about using a base::string16?  base::UTF8[or ASCII if using a
char*]ToUTF16(s).length().

[Azure Wei, 2016-02-17 01:25:57]

The CQ bit was checked by azurewei@chromium.org

[Azure Wei, 2016-02-17 01:25:57]

The patchset sent to the CQ was uploaded after l-g-t-m from
shuchen@chromium.org, rdevlin.cronin@chromium.org
Link to the patchset: https://codereview.chromium.org/1657593007/#ps220001
(title: "Use string16 rather than wstring")

[commit-bot: I haz the power, 2016-02-17 01:28:51]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1657593007/220001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1657593007/220001

[commit-bot: I haz the power, 2016-02-17 03:28:19]

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

[commit-bot: I haz the power, 2016-02-17 03:28:20]

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

[Azure Wei, 2016-02-18 02:58:38]

The CQ bit was checked by azurewei@chromium.org

[Azure Wei, 2016-02-18 02:58:38]

The patchset sent to the CQ was uploaded after l-g-t-m from
shuchen@chromium.org, rdevlin.cronin@chromium.org
Link to the patchset: https://codereview.chromium.org/1657593007/#ps240001
(title: "Fix test error.")

[commit-bot: I haz the power, 2016-02-18 02:58:58]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1657593007/240001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1657593007/240001

[Azure Wei, 2016-02-18 03:16:17]

The CQ bit was unchecked by azurewei@chromium.org

[Azure Wei, 2016-02-18 03:35:49]

The CQ bit was checked by azurewei@chromium.org

[Azure Wei, 2016-02-18 03:35:50]

The patchset sent to the CQ was uploaded after l-g-t-m from
shuchen@chromium.org, rdevlin.cronin@chromium.org
Link to the patchset: https://codereview.chromium.org/1657593007/#ps260001
(title: "Use Xxx::Results::Create().")

[commit-bot: I haz the power, 2016-02-18 03:37:18]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1657593007/260001
View timeline at
 https://chromium-cq-status.appspot.com/patch-timeline/1657593007/260001

[commit-bot: I haz the power, 2016-02-18 04:41:14]

Description was changed from

==========
Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs
on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from
ui::IMEInputContextHandlerInterface, thus we can implement functions
CommitText() and UpdateCompositionText() once for Linux and Windows.

Make InputImeKeyEventHandledFunction::Run() calls the new added
InputMethodBase::KeyEventHandler() in  to set |handling_key_event_|, which
indicates whether the IME extension is handling key event. And when
|handling_key_event_| == true, we should not setComposition and commitText until
it's done to run the callback.

Add helper function UpdateCommitText() and CommitTextToInputContext() in
InputMethodEngineBase, which will be override in chromeos::InputMehtodEngine and
input_method::InputMethodEngine to implement the API on different platforms.


BUG=517773
TEST=None
==========

to

==========
Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs
on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from
ui::IMEInputContextHandlerInterface, thus we can implement functions
CommitText() and UpdateCompositionText() once for Linux and Windows.

Make InputImeKeyEventHandledFunction::Run() calls the new added
InputMethodBase::KeyEventHandler() in  to set |handling_key_event_|, which
indicates whether the IME extension is handling key event. And when
|handling_key_event_| == true, we should not setComposition and commitText until
it's done to run the callback.

Add helper function UpdateCommitText() and CommitTextToInputContext() in
InputMethodEngineBase, which will be override in chromeos::InputMehtodEngine and
input_method::InputMethodEngine to implement the API on different platforms.


BUG=517773
TEST=None
==========

[commit-bot: I haz the power, 2016-02-18 04:41:16]

Committed patchset #14 (id:260001)

[commit-bot: I haz the power, 2016-02-18 04:42:24]

Description was changed from

==========
Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs
on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from
ui::IMEInputContextHandlerInterface, thus we can implement functions
CommitText() and UpdateCompositionText() once for Linux and Windows.

Make InputImeKeyEventHandledFunction::Run() calls the new added
InputMethodBase::KeyEventHandler() in  to set |handling_key_event_|, which
indicates whether the IME extension is handling key event. And when
|handling_key_event_| == true, we should not setComposition and commitText until
it's done to run the callback.

Add helper function UpdateCommitText() and CommitTextToInputContext() in
InputMethodEngineBase, which will be override in chromeos::InputMehtodEngine and
input_method::InputMethodEngine to implement the API on different platforms.


BUG=517773
TEST=None
==========

to

==========
Implement chrome.input.ime.setComposition and chrome.input.ime.commitText APIs
on Linux and Windows platform.

This cl makes ui::InputMethodBase inherits from
ui::IMEInputContextHandlerInterface, thus we can implement functions
CommitText() and UpdateCompositionText() once for Linux and Windows.

Make InputImeKeyEventHandledFunction::Run() calls the new added
InputMethodBase::KeyEventHandler() in  to set |handling_key_event_|, which
indicates whether the IME extension is handling key event. And when
|handling_key_event_| == true, we should not setComposition and commitText until
it's done to run the callback.

Add helper function UpdateCommitText() and CommitTextToInputContext() in
InputMethodEngineBase, which will be override in chromeos::InputMehtodEngine and
input_method::InputMethodEngine to implement the API on different platforms.

BUG=517773
TEST=None

Committed: https://crrev.com/db5505268890037e5af265049ab64f79c01ef162
Cr-Commit-Position: refs/heads/master@{#376095}
==========

[commit-bot: I haz the power, 2016-02-18 04:42:26]

Patchset 14 (id:??) landed as
https://crrev.com/db5505268890037e5af265049ab64f79c01ef162
Cr-Commit-Position: refs/heads/master@{#376095}