Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(52)

Issue 201133002: Add the minimal Tajik locale data and update tz data (Closed)

Created:
6 years, 9 months ago by jungshik at Google
Modified:
6 years, 9 months ago
Reviewers:
Mark Mentovai, Torne
CC:
chromium-reviews, please use gerrit instead
Visibility:
Public.

Description

Add the minimal Tajik locale data and update tz data 1. Add the minimal Tajik locale (exemplar character sets and script) for for Tajik spellchecking 2. Update the timezone data to 2014a (the latest as of today) 3. Drop tranlsiteration files for case changes except for el_Upper.txt because all others (Turkic and Lithuanian) are covered by ICU's case changing API without relying on the transliterator. Blink was already changed to use the regular ICU case changing API except for Greek uppercasing for which transliteration (el_Upper) is used. 4. Rebuild icudtl.dat for non-Android and Android (for build icu_use_data_file_flag=1) as well as the assembly source files for Linux, Mac and Android (icu_use_data_file_flag=0). icudt.dll for Windows (icu_use_data_file_flag=0) is not updated in this CL. We don't need it because icu_use_data_file_flag=1 on all platforms now. Just in case, I'll update that in a separate CL. For the record, icudtl.dat got smaller by ~ 11kB. The difference mainly comes from a smaller zoneinfo64.res : icudtold/translit/root.res 2256 icudtnew/translit/root.res 832 icudtold/windowsZones.res 20288 icudtnew/windowsZones.res 20256 icudtold/zoneinfo64.res 160400 icudtnew/zoneinfo64.res 149520 icudtold/metaZones.res 39824 icudtnew/metaZones.res 39968 TBR=mark BUG=349194 TEST=None until this version of ICU is rolled in. Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=257461

Patch Set 1 #

Patch Set 2 : #

Unified diffs Side-by-side diffs Delta from patch set Stats (+901 lines, -976 lines) Patch
M README.chromium View 1 2 chunks +6 lines, -5 lines 0 comments Download
M android/icudtl.dat View Binary file 0 comments Download
M source/data/in/icudtl.dat View Binary file 0 comments Download
M source/data/locales/reslocal.mk View 1 1 chunk +1 line, -0 lines 0 comments Download
A source/data/locales/tg.txt View 1 1 chunk +17 lines, -0 lines 0 comments Download
M source/data/misc/metaZones.txt View 1 11 chunks +14 lines, -24 lines 0 comments Download
M source/data/misc/windowsZones.txt View 1 9 chunks +10 lines, -15 lines 0 comments Download
M source/data/misc/zoneinfo64.txt View 1 117 chunks +853 lines, -876 lines 0 comments Download
M source/data/translit/css3transform.txt View 1 1 chunk +0 lines, -56 lines 0 comments Download

Messages

Total messages: 6 (0 generated)
jungshik at Google
6 years, 9 months ago (2014-03-15 06:55:42 UTC) #1
jungshik at Google
Committed patchset #2 manually as r257461 (presubmit successful).
6 years, 9 months ago (2014-03-17 17:34:58 UTC) #2
Torne
Hi Jungshik, The change that you landed here seems to have inadvertently changed the name ...
6 years, 9 months ago (2014-03-18 15:06:19 UTC) #3
Mark Mentovai
LGTM
6 years, 9 months ago (2014-03-18 15:23:15 UTC) #4
Mark Mentovai
(LGTM to the change, you should still look at Torne’s concerns.)
6 years, 9 months ago (2014-03-18 15:23:46 UTC) #5
Torne
6 years, 9 months ago (2014-03-19 18:12:34 UTC) #6
Message was sent while issue was closed.
On 2014/03/18 15:06:19, Torne wrote:
> Hi Jungshik,
> 
> The change that you landed here seems to have inadvertently changed the name
of
> the ICU data symbol in the mac assembly source file to no longer be prefixed
by
> an underscore (the file in question isn't in this review but was part of the
> committed CL). This didn't break any mac bots on the chromium waterfall
because
> the normal bots all use the separate data file now and not the assembly source
> file, but it broke the Android WebView downstream because we support building
> using mac host platforms.
> 
> Presumably the generation tool got run with the wrong arguments or something?
> Would you be able to double check the generation process and fix this? Just
> re-adding the underscore to the assembly file should fix the immediate issue,
> but if the generator has changed or the instructions are wrong that may be
worth
> addressing too..

Alternatively, we can stop using the assembly file here entirely by landing
https://codereview.chromium.org/204783003/ - and then it would be possible to
just remove the icu_use_data_file flag and *always* use it, and then not
generate or check in the assembly files at all :)

Powered by Google App Engine
This is Rietveld 408576698