What's new in ICU 62.1:
  - Unicode 11 (new scripts, Emoji update, Text Boundary/Line break updates)
  - CLDR 33.1 (locale data; http://cldr.unicode.org/index/downloads/cldr-33-1 )
  - More Numberformatter improvements and its support in MsgFormat
  - See http://site.icu-project.org/download/62 for more details.

The update steps are recorded in
  https://chromium.googlesource.com/chromium/deps/icu/+log/172d331..c28591b

The list of local changes are:

  Cherry-pick Greek lowercasing fix
  Fix a DecimalFormat maxSignificantDigits issue.

* Rebuild the data files: size eduction of ~ 45 KiB

   6386176 => 6340576 android
   4929248 => 4883232 cast
  10264208 => 10218000 common
   6372480 => 6326896 ios

* Update README.chromium and remove a reference to
  icudt.dll for Windows because it's not used any more.

* Drop patches that became unnecessary.

* Apply the remaining patches
   Isvalidenum.patch, locid_map.patch,
   double_conversion.patch

* Apply data-build related patches
  {wpo,vscomp,data.build,data.build.win,data_symb}.patch

* Apply BreakIterator related patches + locale1 patch

* Apply Google's patch for locale data
  - region display names
  - likely subtags
  - Arabic locale: default number system => Latn
  - AM/PM markers for Hindi and a few Indic locales

* Check in ICU 62.1 to the tree (62staging branch) by
  running scripts/update.sh.
  Build files are updated by the script.

TBR=drott@chromium.org
Bug: 850334
Test: None
Change-Id: I7327363c90b9bccda20c1be8f4a65399f36c4e99
Reviewed-on: https://chromium-review.googlesource.com/1128503


Reviewed-by: Jungshik Shin <jshin@chromium.org>

ICU's Windows timezone detection is broken on Japanese/Korean
(and potentially other non-English) Windows.

Cherrypick the upstream fix.
See http://bugs.icu-project.org/trac/ticket/13826 .

TBR=gsathya@chromium.org
Bug: 849724
Test: Set the timezone to Tokyo on Japanese Windows
Test: In JS console, 'new Date()' outputs 'Japan Standard Time' in parens.
Change-Id: Ie76c1e97ee2e965825f2fe99442c55e797b45f41
Reviewed-on: https://chromium-review.googlesource.com/1094322


Reviewed-by: Jungshik Shin <jshin@chromium.org>

What's new in ICU 61.1:
  - CLDR 33 (locale data; http://cldr.unicode.org/index/downloads/cldr-33 )
  - A new number formatting API
  - Rewrote U8_NEXT
  - See http://site.icu-project.org/download/61 for more details.

The update steps are recorded in
  https://chromium.googlesource.com/chromium/deps/icu/+log/f7ab784..b0e0b84

On top of the above, IANA timezone db was updated once more to the
latest 2018e (released on May 4th).

The list of local changes are:

* Update the currency list. For desktop, include all the national
  currencies (see https://bugs.chromium.org/p/v8/issues/detail?id=7472 )

* Clean up disabled warning list in BUILD.gn/icu.gyp

* Fix clang on Windows and MSVC build failures
  https://github.com/google/double-conversion/issues/66
  http://bugs.icu-project.org/trac/ticket/13752
  http://bugs.icu-project.org/trac/ticket/13750

* Rebuild the data files
   Size is reduced by ~100kB for non-desktop.
   Desktop(common): increase by 90kB

    6,386,176 android/icudtl.dat  (6,502,048)
    4,929,248 cast/icudtl.dat     (5,032,368)
   10,264,208 common/icudtl.dat  (10,171,920)
    6,372,480 ios/icudtl.dat      (6,488,384)

* Update IANA tz DB to 2018e

* Update README.chromium

* Drop patches that became unnecessary.

* Apply the remaining patches

   Isvalidenum.patch, simplify_smallintformatter.patch,
   word_long.patch, locid_map.patch

* Apply data-build related patches

  {wpo,vscomp,data.build,data.build.win,data_symb}.patch

* Apply BreakIterator related patches + locale1 patch

* Apply Google's patch for locale data

* ICU 61.1 update step 1

  Check in ICU 61.1 to the tree (61staging branch) by
  running scripts/update.sh.
  Build files are updated by the script.

* Update prep

TBR=thakis@chromium.org
Bug: 838383
Test: None
Change-Id: Ib83ed624fe764abdb7fbfd567ae5e982497089bd
Reviewed-on: https://chromium-review.googlesource.com/1041578


Reviewed-by: Jungshik Shin <jshin@chromium.org>

1. Update grandfathered tag list and map
2. Update redundant tag list and map
3. Update deprecated lang list
4. Add deprecated region list and map them to modern region subtags.
5. Map ja-Latn-hepburn-heploc to ja-Latn-alalc97

The IANA language subtag regisry is used to update.

https://www.iana.org/assignments/language-subtag-registry/language-subtag-registry

Upstream bugs:

    https://ssl.icu-project.org/trac/ticket/13726
    https://ssl.icu-project.org/trac/ticket/13723
    https://ssl.icu-project.org/trac/ticket/13721
    https://ssl.icu-project.org/trac/ticket/13720
    https://ssl.icu-project.org/trac/ticket/13719

TBR=gsathya@chromium.org
Bug: v8:7669, v8:5693
Test: v8 test will be added and updated.
Change-Id: I773475ede43bb7a3048f176dde2cee6bfdf9a3e7
Reviewed-on: https://chromium-review.googlesource.com/1026797


Reviewed-by: Jungshik Shin <jshin@chromium.org>

* Fix the undefined behavior in decimal number parsing

      http://bugs.icu-project.org/trac/changeset/40950

* Fix the handling of non-BMP characters in CJK breakiterator

      http://www.icu-project.org/trac/changeset/40949

* Limit the recursion depth of UnicodeSet pattern

      http://bugs.icu-project.org/trac/changeset/40979

TBR=inferno@chromium.org
Bug: chromium:799850, chromium:796807, chromium:796752
Test: See the bugs.
Change-Id: I1a8909371b601f36faca911039b10d36c7a92c85
Reviewed-on: https://chromium-review.googlesource.com/1009001


Reviewed-by: Jungshik Shin <jshin@chromium.org>

The ICU word break rules are updated to deal with a performance issue
with selecting a very long word by double-click.

In addition, update the IANA tz database to 2018d. Palestine started
DST (UTC+3) on March 24 instead of March 31.

> new Date(Date.UTC(2018,2,24, 12)).toLocaleString("en", {'timeZone': 'Asia/Gaza'})
"3/24/2018, 3:00:00 PM"

Bug: chromium:829144, chromium:473288
Test: See the 1st bug description.
Test: See the above JS code snippet and result.
Change-Id: Ic8beeaaa4f0e5a16701b0d8b1563fc8537c6b0ab
TBR=mark@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/1006521


Reviewed-by: Jungshik Shin <jshin@chromium.org>

See http://mm.icann.org/pipermail/tz-announce/2018-January/000048.html
for details. Most notable is that Brazil will start DST (UTC-3 => UTC-2 in
São Paulo) on the 1st Sunday of November instead of the 2nd Sunday of
October.

In JS console:
> new Date(Date.UTC(2018, 9,31, 5)).toLocaleString("en", {timeZone: 'America/Sao_Paulo'})
"10/31/2018, 2:00:00 AM"
> new Date(Date.UTC(2017, 9,31, 5)).toLocaleString("en", {timeZone: 'America/Sao_Paulo'})
"10/31/2017, 3:00:00 AM"

TBR=mark@chromium.org
Bug: 473288
Test: See the above
Change-Id: I7cb1c48c461a8406cee1e34042c8eec0fb31f85f
Reviewed-on: https://chromium-review.googlesource.com/898245


Reviewed-by: Jungshik Shin <jshin@chromium.org>

This is a re-land of https://chromium-review.googlesource.com/c/836554
which was reverted because it includes a C++14 constexpr expression
that failed on the V8 build-bots (which are restricted to C++11
language features, unlike the Chromium ones).

The patch now uses a C++11-compatible constexpr function.

Drastically reduce the size of the SmallIntFormatter code.
This should make the code slightly slower, but reduce the
overall size of the generated library by about 16 kiB.

It is unlikely that the performance difference will be
noticeable on any existing system Chromium runs on.

R=ajwong@chromium.org, jshin@chromium.org

Bug: NONE
Change-Id: I425626a53e0a04af4110acd0410655e6a08d8568
Reviewed-on: https://chromium-review.googlesource.com/870039


Reviewed-by: Albert J. Wong <ajwong@chromium.org>
Reviewed-by: Jungshik Shin <jshin@chromium.org>

This reverts commit 2e0c90b4.

It is to make gcc with '-std=gnu+11' happy. v8 trybots with gcc(C++ 11)
failed to compile with this CL.

Bug: 802890
Test: None
Change-Id: I4c77d07281ebaefec9bb03fd4fba92e436009bd2
TBR=digit@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/869110


Reviewed-by: Jungshik Shin <jshin@chromium.org>

When the time is set to MAX_MILLIS, Islamic calendar
year and month calculation results in an integer overflow.

TBR=littledan@chromium.org
Bug: 801602
Test: See the bug (v8 test will be added)
Change-Id: Ifcb6907de81de83b690f76a536818054815a15a7
Reviewed-on: https://chromium-review.googlesource.com/865524


Reviewed-by: Jungshik Shin <jshin@chromium.org>

Drastically reduce the size of the SmallIntFormatter code.
This should make the code slightly slower, but reduce the
overall size of the generated library by about 48 kiB.

It is unlikely that the performance difference will be
noticeable on any existing system Chromium runs on.

Bug: NONE
Change-Id: I1faa6020c0e9a7cc61f0f274b9a2c4ab6701aece
Reviewed-on: https://chromium-review.googlesource.com/836554


Reviewed-by: Jungshik Shin <jshin@chromium.org>
Reviewed-by: Albert J. Wong <ajwong@chromium.org>

38e3cf74 has a typo.

Bug: None
Change-Id: Id16f30ab5f2938bef567ef898f481696122a1dd3
TBR=hablich@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/856237


Reviewed-by: Jungshik Shin <jshin@chromium.org>

The upstream also has this change to make AIX compiler happy.

Bug: None
Change-Id: I059ccc77a6b9aa0da9b4bab95103d2c10202f5a2
TBR=hablich@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/855583


Reviewed-by: Jungshik Shin <jshin@chromium.org>

In CLDR 32.0 (used in ICU 60), German AM/PM markers were changed to
'AM' and 'PM', but the change is controversial. For now, revert to
the previous values per the upstream (CLDR 32.0.1 / ICU 60.2). See
   https://unicode.org/cldr/trac/ticket/10735 .

Bug: 794737
Test: base_unittests --gtest_filter=TimeFormat*.*TimeOfDayDE
Change-Id: Ifb715af8cf6d41803eaf612cbb8940dbc7851806
Reviewed-on: https://chromium-review.googlesource.com/826250


Reviewed-by: Mark Mentovai <mark@chromium.org>

Bug: 794390
Change-Id: I31a1e622b38060c322be8086cc8aefdafb431a1b
Reviewed-on: https://chromium-review.googlesource.com/822188


Reviewed-by: Mark Mentovai <mark@chromium.org>

TBR=elwarence@chromium.org
Bug: 792537
Change-Id: Idfc4fa258d0f9ffb6fd236183275ec4580a00f73
Reviewed-on: https://chromium-review.googlesource.com/809729


Reviewed-by: Jungshik Shin <jshin@chromium.org>

After data trimming, data/curr/en_GB.txt has empty "Currencies" block.
This prevents a locale fallback chain from being triggered for currency
symbols (en-GB => en-001 => en), leading all the currency symbols in en-GB
to be the last resort value (ISO 3-letter currency code). To avoid that,
trim_data.sh is revised to drop empty "Currencies" block entirely ('Missing'
triggers a fallback chain, but 'empty' goes straight to the last resort.)

Additional currency data fixes:

1. Serbian: RON and ROL display names are swapped. RON is the current
currency and should be included with a shorter display name.

2. Add a few more en-* locale variants to the locale list for currency.

Moreover, make_data_all.sh is updated to work wherever the Chromium
source tree is.

Finally, update the list of resources to remove from cast's ICU data
and rebuild it with ICU 60.  Added entries to cast-removed-resources.txt
are data for locale-variants that are not useful in Cast.

Bug: 791318
Test: See the bug
Change-Id: Ib09ee36e3447557636d60d8aeabc7ca047718236
Reviewed-on: https://chromium-review.googlesource.com/810099


Reviewed-by: Mark Mentovai <mark@chromium.org>

Recent Clang versions started taking into account that enums are signed
on Windows when emitting these warnings.

Bug: 792519
Change-Id: I32782dc34fc0480942038e4950266448e38d2eec
Reviewed-on: https://chromium-review.googlesource.com/813495


Reviewed-by: Jungshik Shin <jshin@chromium.org>

Our Cast Audio devices use a stripped down version of the android
icudtl.dat. Essentially the patch_locale.sh has been modified to keep in
things provided by Java, as cast doesn't have Java and needs there
resources.

On the other hand, a bunch of resources have been stripped from the
icudtl.dat to reduce its size because we are very constrained on that.

We have been using this icudtl.dat process for a long time now, and it's
throughly validated. However, we used to generate this in our internal
repo, which meant that we have to manually update it whenever chromium
updates their icudtl.dat.

This CL aims to intergrate the generation of the icudtl.dat in chromium,
so we no longer have to manually update in our internal repo.

Bug: internal b/28053340
Test: On cast audio device

Change-Id: I881eb2ddf046fdd16fd7efbfb0d56877abfaf947
Reviewed-on: https://chromium-review.googlesource.com/773591


Reviewed-by: Jungshik Shin <jshin@chromium.org>

Cherry pick upstream fix for
  http://bugs.icu-project.org/trac/ticket/13462
which is
  http://bugs.icu-project.org/trac/changeset/40667

TBR=mathias@chromium.org

Bug: v8:4743
Test: v8: test262/built-ins/RegExp/property-escapes/generated/*
Change-Id: I3f94b6bf519bd395bec1f73e5f8d602f93910032
Reviewed-on: https://chromium-review.googlesource.com/760120


Reviewed-by: Jungshik Shin <jshin@chromium.org>

What's new in ICU 60.1:
  - Unicode 10.0 (including Emoji 5.0)
  - CLDR 32 (locale data; http://cldr.unicode.org/index/downloads/cldr-32 )
  - Smaller rule file size for BreakIterator (no more reverse rules)
  - A new number formatting API
  - Malformed UTF-8 sequences are handled per W3C Encoding Standard
  - See http://site.icu-project.org/download/60 for more details.

In addition, the following changes are made:

* Add more locale variants for en,fr,de,it,pt,nl,ru,ar

* Add the minimum locale data for Osage and Chakma

* Apply locale data patches (Google and Chromium)

* Apply build-related and break iterator patches:
  wpo.patch
  vscomp.patch
  data.build.win.patch
  data_symb.patch
  cjdict.patch
  khmer-dictbe.patch
  wordbrk.patch
  patches/data.build.patch

* Update BUILD.gn and gyp files.

* Data file size (25 ~ 130kB smaller than the latest 59.1)
  android/icudtl.dat 6499904
  common/icudtl.dat 10170896
  ios/icudtl.dat 6486240

The update steps are recorded in
https://chromium.googlesource.com/chromium/deps/icu/+log/f88520f..1034c7e5 .


Bug: 766816
Test: None (DEPS roll CL will have test specified).
Change-Id: I9144bb37464eec1f282f40f9f2d28b6306511083
Reviewed-on: https://chromium-review.googlesource.com/756237


Reviewed-by: Mark Mentovai <mark@chromium.org>

Bug: chromium:778300
Test: See bug commment 4
Change-Id: I1ae525323cb3354d3936a92e66cff266ebca2fb4
Reviewed-on: https://chromium-review.googlesource.com/738833


Reviewed-by: Mark Mentovai <mark@chromium.org>

The location of zoneinfo directory has changed on macOS 10.13,
Ubuntu 16, RHEL 7 and SuSe Linux 12. It results in the misdetection
of the OS timezone by ICU. Cherry-picking the CLs for the following
upstream bug fixes it.

https://ssl.icu-project.org/trac/ticket/12770

Bug:754053,766916
Test: In Javascript console, the following should give the correct
   timezone.
    (new Intl.DateTimeFormat()).resolvedOptions().timeZone

Change-Id: I711f4b27e73dc6855951055a601e80f711d34423
Reviewed-on: https://chromium-review.googlesource.com/710496


Reviewed-by: Mark Mentovai <mark@chromium.org>

Bug:770452
Test: See the bug

Change-Id: I5e19012a57507393670d68a7228bdb0831918997
Reviewed-on: https://chromium-review.googlesource.com/705920


Reviewed-by: Mark Mentovai <mark@chromium.org>

Because the Fuchsia system bootdata.bin puts libicuuc.so and
libicui18n.so in /system/lib already (and they don't match our build),
and we have same-named .sos in component build, rename our copy.

Bug: 732897, 743296
Change-Id: Ia195230c14d515ed14f9d90eb35aa87301eab601
Reviewed-on: https://chromium-review.googlesource.com/613506


Reviewed-by: Nico Weber <thakis@chromium.org>

1. Add DST abbreviation for Lord Howe timezone

2. Drop a few aliases for GBK/GB18030
  These aliases are not listed in https://encoding.spec.whatwg.org/

3. Fix Russian names for Hans/Hant (non-standalone). They're too long.

4. Add the abbreviated zone names for Asia/Pyongyang

ICU data files get a bit smaller as below:

  Desktop: 10196720 -> 10196592 bytes
  Android:  6635312 -> 6635168 bytes
  iOS:      6623568 -> 6623360 bytes

BUG=chromium:544228,chromium:339862,v8:6031
TEST=LayoutTests: fast/encoding/supported-encodings.html
TBR=littledan@chromium.org

Change-Id: I93881545a09485ff383a3630c85782c8754a9bd0
Reviewed-on: https://chromium-review.googlesource.com/599587


Reviewed-by: Jungshik Shin <jshin@chromium.org>

Bug: 751810
Change-Id: I377431074c25fbc8669cee311be304635953cc2a
Reviewed-on: https://chromium-review.googlesource.com/600488


Reviewed-by: Jungshik Shin <jshin@chromium.org>

R=jshin@chromium.org

Change-Id: Ia650d3b9699e0cf074265c61a8656f65bc4c7def
Reviewed-on: https://chromium-review.googlesource.com/587912


Reviewed-by: Jungshik Shin <jshin@chromium.org>

Blink does not need it any more. Drop UTF-32 converters and aliases.

Save 320 bytes in ICU data. icuuc.so will be slightly smaller, too.

BUG=chromium:417850
TEST=None

Change-Id: Ib4fecc3fcc905c974e68a652e7f4722bfd271129
Reviewed-on: https://chromium-review.googlesource.com/576872


Reviewed-by: Joshua Bell <jsbell@chromium.org>

The following locale variants are added to the ICU. Only source/data/locales
are added and other locale categories are left alone.

- es-* variants
- en-SG and en-HK
- Arabic variants whose NumberElements are different from ar.

Data file size changes:

- desktop:
   10192200 - 10178576 = 13624  (adding es variants)
   10197040 - 10192200 =  4840  (adding ar variants)
- Android: 6635632 - 6616864 = 18768
- iOS: 6623888 - 6605152 = 18736


Bug:chromium:732762
Test:See the bug report and comment 13
TBR=littledan@chromium.org
Change-Id: I3e87be3413e22aa49f24f81b0e25bd6e4187c0d4
Reviewed-on: https://chromium-review.googlesource.com/576113


Reviewed-by: Jungshik Shin <jshin@chromium.org>

They're supported by Google Translate and their names need to be shown
in the Translate UI.

Hmong (hmn)
Javanese (jv)
Luxembourgish (lb)
Samoan (sm)

The size impact for desktop Chrome is about 3kB.
( 10,175,056 => 10,178,576)

BUG=733398
TEST=See the bug
TBR=yyushkina@chromium.org

Change-Id: I9ca232343847a9a95c1dabacb6336bd43ba39849
Reviewed-on: https://chromium-review.googlesource.com/574771


Reviewed-by: Jungshik Shin <jshin@chromium.org>

http://bugs.icu-project.org/trac/ticket/13277

Also fix README.chromium file to refer to 59.1 instead of 58.1.

BUG=chromium:734093
TEST=None (upstream has a test but not included in the Chrome's copy of ICU)
TBR=avi@chromium.org

Change-Id: I0177d80d7c94199fdea107f891a922b67f0d8481
Reviewed-on: https://chromium-review.googlesource.com/574935


Reviewed-by: Jungshik Shin <jshin@chromium.org>

BUG=729032
R=jshin@chromium.org

Review-Url: https://codereview.chromium.org/2921843003 .

Bug:chromium:726654
Test: gn gen works without an error
Review-Url: https://codereview.chromium.org/2913423002 .

This is necessary for Chrome OS to calculate sunset time.

Bug:chromium:726654
Test: With https://codereview.chromium.org/2887913004 patched in, 'gn
Change-Id: Icf6a8cd833f8afcd1a2d0ce874cf1c8c92fcba44
gen <out_dir> --check' does not complain.
TBR=afakhry@chromium.org
Reviewed-on: https://chromium-review.googlesource.com/520323


Reviewed-by: Jungshik Shin <jshin@chromium.org>

Without a Linux-ish define for Fuchsia, neither U_WCHAR_IS_UTF16 nor
U_WCHAR_IS_UTF32 are defined, resulting in warnings in ICU code when
compiled on Fuchsia. Filed upstream at
http://www.icu-project.org/trac/ticket/13202.

BUG=706592
R=jshin@chromium.org

Review-Url: https://codereview.chromium.org/2885143008 .

   http://www.icu-project.org/trac/ticket/12333
   http://www.icu-project.org/trac/ticket/13189
   http://www.icu-project.org/trac/ticket/12635

   - patches/ucase_utf8.patch
   - patches/ucurr_locale.patch
   - patches/collator_range.patch

Bug=chromium:722124
TEST=See the test
TBR=inferno@chromium.org

Change-Id: I3e754734a1c1121c8059a386ca1b5d0f004adb79
Reviewed-on: https://chromium-review.googlesource.com/506592


Reviewed-by: Abhishek Arya <inferno@chromium.org>
Reviewed-by: Jungshik Shin <jshin@chromium.org>

Also add support for host_byteorder logic which is introduced in -
https://codereview.chromium.org/2815453004/

R=adamk@chromium.org, dpranke@chromium.org, machenbach@chromium.org
BUG=706728

Review-Url: https://codereview.chromium.org/2812173002 .

* Highlights:
  - Emoji 5.0 data (partial; Emoji_Component property not included)
  - CLDR 31.0.1 (http://blog.unicode.org/2017/03/cldr-version-31-released.html)
    UTC and GMT are treated as distinct)
  - New case mapping API for styled text
  - C++ 11 is required
  - char16_t for UChar (UTF-16)
  - Source code is in UTF-8

* Size changes

  common: 10,130,560 => 10,175,056
  android: 6,573,872 => 6,616,864
  iOS: 6,562,352 => 6,605,152

On top of ICU 59.1 from the upstream, the following changes were applied.
See https://chromium.googlesource.com/chromium/deps/icu/+log/chromium/59staging

  - Fix C++ 11 string literal assignment issue (upstream bug: 13192)
  - Fix C4229 warning by MSVC
  - Apply utf32.patch and include unistr.h in fuzzer_util
  - Update ICU data files
  - Fix wpo.patch
  - Apply Google locale patch and locale1.patch
  - update readme
  - Apply breakiterator related patches
  - Apply and update wpo.patch
  - Drop unused patch, apply data.build.win.patch, update README.chromium
  - Add /utf-8 flag for Windows/Visual Studio
  - Update BUILD.gn for UChar, stubdata and apply data_sym.patch
  - use stubdata.cpp instead of stubdata.c in icu.gyp
  - Update icu.gyp* files for v8
  - Update BUILD.gn, apply data.build.patch and vscomp.patch
  - Add new files in ICU 59.1
  - Get a fresh copy of ICU 59.1 from the upstream
  - Update update.sh script

TBR=drott@chromium.org, yangguo@chromium.org
Bug:699469
TEST: layout tests, all unittests, browser tests
Change-Id: Ie1e77323aa0c7f872153680c4deca6471a771a5c
Reviewed-on: https://chromium-review.googlesource.com/505173


Reviewed-by: Jungshik Shin <jshin@chromium.org>

Add ios/icudtl.dat and ios/patch_locale.sh.

Update README.chromium and BUILD.gn accordingly.
Update scripts/copy_data.sh to take (ios|common|android).

At the moment, iOS data is almost identical to that of Android, but in the future
more cuts may be made (e.g. dictionary data for breakiterator).

Bug: 718955
TEST: iOS Chrome works as before.
Review-Url: https://codereview.chromium.org/2743123002 .