Reland "Replace ICU with CED for auto encoding detection"

Reland "Replace ICU with CED for auto encoding detection"

Another attempt to land crrev.com/2081653007.

The patch was reverted due accidentally having added new static
constructors (string), and causing a failure of the "sizes" step
of the build.

Link to broken build:
https://build.chromium.org/p/chromium/builders/Linux%20x64/builds/21618

Relevant log lines:
FAILED linux-release-64/sizes/chrome-si/initializers: actual 9, expected 7, better lower

The issue was fixed in the library on github.com:
https://github.com/google/compact_enc_det/commit/4abb8e48ca45e40bb6e4eb5767f20498e8df0075
https://github.com/google/compact_enc_det/commit/45c3d4a3d7ee40adf271748f3c4177fa293800e1

Tested:
-------
$ tools/linux/dump-static-initializers.py out/Release/chrome
...

Found 40 static initializers in 7 files.
-------

This reverts commit dbd9f219ef9b615d9e116964e46b1a390c9354a1.

BUG=

Committed: https://crrev.com/ec678d0e5493256ccf2b2b6a752de289b58b9d28
Cr-Commit-Position: refs/heads/master@{#405352}

[Jinsuk Kim, 2016-07-11 05:00:11]

jinsukkim@chromium.org changed reviewers:
+ tkent@chromium.org

[tkent, 2016-07-11 05:03:39]

lgtm

[Jinsuk Kim, 2016-07-13 22:01:47]

The CQ bit was checked by jinsukkim@chromium.org

[commit-bot: I haz the power, 2016-07-13 22:03:50]

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, 2016-07-13 23:55:41]

Committed patchset #1 (id:1)

[commit-bot: I haz the power, 2016-07-13 23:56:01]

CQ bit was unchecked.

[commit-bot: I haz the power, 2016-07-13 23:58:26]

Description was changed from

==========
Reland "Replace ICU with CED for auto encoding detection"

Another attempt to land crrev.com/2081653007.

The patch was reverted due accidentally having added new static
constructors (string), and causing a failure of the "sizes" step
of the build.

Link to broken build:
https://build.chromium.org/p/chromium/builders/Linux%20x64/builds/21618

Relevant log lines:
FAILED linux-release-64/sizes/chrome-si/initializers: actual 9, expected 7,
better lower

The issue was fixed in the library on github.com:
https://github.com/google/compact_enc_det/commit/4abb8e48ca45e40bb6e4eb5767f2...
https://github.com/google/compact_enc_det/commit/45c3d4a3d7ee40adf271748f3c41...

Tested:
-------
$ tools/linux/dump-static-initializers.py out/Release/chrome
...

Found 40 static initializers in 7 files.
-------

This reverts commit dbd9f219ef9b615d9e116964e46b1a390c9354a1.

BUG=
==========

to

==========
Reland "Replace ICU with CED for auto encoding detection"

Another attempt to land crrev.com/2081653007.

The patch was reverted due accidentally having added new static
constructors (string), and causing a failure of the "sizes" step
of the build.

Link to broken build:
https://build.chromium.org/p/chromium/builders/Linux%20x64/builds/21618

Relevant log lines:
FAILED linux-release-64/sizes/chrome-si/initializers: actual 9, expected 7,
better lower

The issue was fixed in the library on github.com:
https://github.com/google/compact_enc_det/commit/4abb8e48ca45e40bb6e4eb5767f2...
https://github.com/google/compact_enc_det/commit/45c3d4a3d7ee40adf271748f3c41...

Tested:
-------
$ tools/linux/dump-static-initializers.py out/Release/chrome
...

Found 40 static initializers in 7 files.
-------

This reverts commit dbd9f219ef9b615d9e116964e46b1a390c9354a1.

BUG=

Committed: https://crrev.com/ec678d0e5493256ccf2b2b6a752de289b58b9d28
Cr-Commit-Position: refs/heads/master@{#405352}
==========

[commit-bot: I haz the power, 2016-07-13 23:58:27]

Patchset 1 (id:??) landed as
https://crrev.com/ec678d0e5493256ccf2b2b6a752de289b58b9d28
Cr-Commit-Position: refs/heads/master@{#405352}

[jungshik at Google, 2016-07-19 21:18:29]

jshin@chromium.org changed reviewers:
+ jshin@chromium.org

[jungshik at Google, 2016-07-19 21:18:30]

Thank you !  

I'm adding comments below because I couldn't file a bug at github (somehow, I
don't see issues tab there). What's your plan for bug tracking of CED?  Are you
gonna use github or crbug.com?

https://codereview.chromium.org/2138643002/diff/1/third_party/WebKit/Source/p...
File third_party/WebKit/Source/platform/text/TextEncodingDetector.cpp (right):

https://codereview.chromium.org/2138643002/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/text/TextEncodingDetector.cpp:49: false, //
Include 7-bit encodings
The only 7-bit encoding Blink supports is ISO-2022-JP. We can introduce 'HTML5'
mode (#if defined(HTML5)) to suppress the detection of all the non-HTML5
encodings in CED.

https://codereview.chromium.org/2138643002/diff/1/third_party/WebKit/Source/p...
third_party/WebKit/Source/platform/text/TextEncodingDetector.cpp:53:
*detectedEncoding = WTF::TextEncoding(MimeEncodingName(encoding));
MimeEncodingName uses kEncodingTable (
https://cs.chromium.org/chromium/src/third_party/ced/src/util/encodings/encod...
) that has a lot of non-HTML5 encodings or non-HTML5 names. 

Perhaps, we have to introduce 'HTML5' mode (#if defined(HTML5) ...) in CED to
sanitize those names.