Move ICU headers from public/{common,i18n} to source/{common,i18n}

Move ICU headers from public/{common,i18n} to source/{common,i18n}

The upstream has headers in source/{common,i18n}/unicode.
With Blink's ICU headers removed, there's no need to keep them
in a location different than the upstream. Future ICU updates will
be a bit more streamlined. v8 also likes to have this change.
    
The Chrome-side of change is at http://codereview.chromium.org/18252003

This revision will be rolled in along with the global replacements
of include-references to ICU headers and DEPS changes.

See also http://crbug.com/232922 where ICU headers in Blink are removed.

Landed in 3 steps:

http://crrev.com/211838
http://crrev.com/211848
http://crrev.com/211851


BUG=251433
TEST=NONE for now.

[jungshik at Google, 2013-07-14 06:52:28]

Mark,  please see if there's anything wrong with icu.gyp changes. Other changes
are just moving headers.  

Xianzhu, do you have any concern for Android? 

This CL has to be reviewed together with http://codereview.chromium.org/18252003
(the chromium-side change). 

Thanks,

[Xianzhu, 2013-07-15 16:41:04]

On 2013/07/14 06:52:28, Jungshik Shin wrote:
> Xianzhu, do you have any concern for Android? 
> 
No.

+torne, for use_system_icu on Android.

[Mark Mentovai, 2013-07-15 16:42:25]

icu.gyp looks fine. LGTM.

[Torne, 2013-07-15 16:52:35]

use_system_icu stuff looks reasonable. LGTM.

[Torne, 2013-07-15 16:53:11]

On 2013/07/15 16:52:35, Torne wrote:
> use_system_icu stuff looks reasonable. LGTM.

Er, though actually, we now have an optional trybot for the android webview
config: run a try job on "android_aosp" for the CL and it should prove it works.

[jungshik at Google, 2013-07-15 19:05:12]

On 2013/07/15 16:53:11, Torne wrote:
> On 2013/07/15 16:52:35, Torne wrote:
> > use_system_icu stuff looks reasonable. LGTM.
> 
> Er, though actually, we now have an optional trybot for the android webview
> config: run a try job on "android_aosp" for the CL and it should prove it
works.

Thank you for the review, Mark and Torne. The next CL on th Chrome-side (
http://codereview.chromium.org/18252003 ) will bring in this change and I'll
make sure to run android_aosp bot.

[jungshik at Google, 2013-07-16 18:55:58]

CQ hasn't picked up this CL (because it's in svn but I used a git client?). I'll
try to land manually (git cl dcommit does not work, either). I'll have to copy
the CL to a svn client and land manually. 


On 2013/07/15 19:05:12, Jungshik Shin wrote:
> On 2013/07/15 16:53:11, Torne wrote:
> > On 2013/07/15 16:52:35, Torne wrote:
> > > use_system_icu stuff looks reasonable. LGTM.
> > 
> > Er, though actually, we now have an optional trybot for the android webview
> > config: run a try job on "android_aosp" for the CL and it should prove it
> works.
> 
> Thank you for the review, Mark and Torne. The next CL on th Chrome-side (
> http://codereview.chromium.org/18252003 ) will bring in this change and I'll
> make sure to run android_aosp bot.

[jungshik at Google, 2013-07-16 22:27:24]

Landed in 3 steps as noted in the CL description. 

On 2013/07/16 18:55:58, Jungshik Shin wrote:
> CQ hasn't picked up this CL (because it's in svn but I used a git client?).
I'll
> try to land manually (git cl dcommit does not work, either). I'll have to copy
> the CL to a svn client and land manually. 
> 
> 
> On 2013/07/15 19:05:12, Jungshik Shin wrote:
> > On 2013/07/15 16:53:11, Torne wrote:
> > > On 2013/07/15 16:52:35, Torne wrote:
> > > > use_system_icu stuff looks reasonable. LGTM.
> > > 
> > > Er, though actually, we now have an optional trybot for the android
webview
> > > config: run a try job on "android_aosp" for the CL and it should prove it
> > works.
> > 
> > Thank you for the review, Mark and Torne. The next CL on th Chrome-side (
> > http://codereview.chromium.org/18252003 ) will bring in this change and I'll
> > make sure to run android_aosp bot.