Fwd: CLDR 1.2a+ in ICU format, tools
"Steven R. Loomis" <[email protected]> Wed, 6 Oct 2004 12:44:19 -0700
| Newsgroups | gmane.comp.lib.icu.general |
|---|---|
| Message-ID | <[email protected]> |
ICU users,
As you might be aware, the next release of ICU will have all locale
and collation data built from the CLDR project (
http://www.unicode.org/cldr ). CLDR has shipped an alpha of its
version 1.2, and to facilitate testing, I have created an interim drop
of CLDR data in ICU for C format. (Java data can be built from that
data using the normal process.)
Also listed below are instructions for performing this build of ICU
data yourself.
Please don't ask CLDR questions on the ICU list, if you are
interested in CLDR please see the CLDR web site (above). I'm forwarding
this mail as a service for those who might be interested in CLDR.
If you have questions or comments on this email, please contact me
directly.
___________________
Begin forwarded message:
Date: 6 ta' Ottubru 2004 10:54:18 AM GMT-07:00
To: cldr list
Subject: CLDR 1.2a+ in ICU format, tools
Since most of the tests on CLDR data involve ICU at this stage, I have
made available a drop of data in this form:
Here is a drop of the current CLDR 1.2a+ data:
* without draft data
* without alternates
Note, ICU tests do not pass yet with this data.
http://www.jtcsv.com/locales12a.tgz
( This URL is temporary and subject to go away after CLDR/ICU ship. )
-----------
GENERATING YOUR OWN ICU FORMAT DATA FROM CLDR (on UNIX platforms, not
yet Cygwin)
Paraphrasing from the CLDR wiki (
http://www.jtcsv.com/cgibin/cldrwiki.pl?CldrTools ) :
see http://oss.software.ibm.com/icu/ and
http://oss.software.ibm.com/icu4j/
* Checkout the latest ICU tree (including locale) from CVS so that the
'icu' directory (containing source) is parallel to the 'locale'
directory
* Checkout the latest ICU4J tree from CVS
* build ICU4J according to its instructions, run 'ant tools' to
build the tools. You will have a 'classes' directory.
* In ICU4C, create the file icu/source/data/locales/reslocal.mk
containing the single line:
include ../../../locale/tools/cldrres.mk
This will include the named file which is located in the CLDR tree.
* Then in the icu/source/data directory, run: make cldr-clean-old
; make -k
* You might see some errs for parts of ICU not built yet, but you
should end up with .txt files (after quite some processing time - 30
minutes to hours)