Roll ICU to icu52
The ICU update does the following among other things:
- Update EUC-JP (no more JIS X 212 round-trip coverage) per WHATWG
encoding spec.
- Add CP866
- Add far better/more extensive support for plural, format, etc
- Numerous locale data update
- Khmer/Lao line/word-breaking support
- More compact Chinese/Japanese dictionary (now a part of the upstream
instead of our local patch)
- Better plural support that we plan to use right away.
- Update icu/BUILD.gn : file list update, new defines, etc.
- Update tests
a. icu_string_conversions_unittests : drop iscii (nobody uses it) and
remove JIS X 212 code points from the round-trip test
b. l10n_util_unittests: Display names for zh-*.
c. ThaiSnippet's expected result. (due to a Unicode grapheme definition change).
d. Update Expected results in file_manager_browsertest/test_util.js
(see http://crbug.com/383930)
- Update SpellcheckerWordIterator
a. Add U+0027 back to MidNumLet
b. Add Lao and Khmer scripts to the list of scripts requiring dictionary-based
word-breaking.
- l10n_util : alias 'tl' to 'fil' (Tagalog to Filipino)
- Update IdentityInternalsSingleTokenWebUITest.verifyGetters (DateTimeFormat-Full
has 'at' before time, which is not handled by Date.parse()
- update license.py because we don't have a separate Chinese/Japanese
dictionary any more. It's now a part of ICU
BUG=132145,277023,136148,360433,364072
TEST=base_unittests:*ICU*, net_unittests:*IDN*, ui_unittests:*Lang*,
browser_tests (encoding related, Iden*verifyGet*),
unit_tests:Ext*L10*, unit_tests:*Lang*, component_tests:*Snipp*,
layout tests (encoding, date/calendar, line breaking, and others),
build 'base_i18n_nacl', GN build bots.
Review URL: https://codereview.chromium.org/265613002
git-svn-id: svn://svn.chromium.org/chrome/trunk/src@277111 0039d316-1c4b-4281-b951-d872f2087c98
12 files changed
