On 12/18/2013 6:43 PM, Aleksej Efimov wrote:
Hi,
Please help to review a fix [1] for 8025051 bug [2]. The following fix
includes:
Common to all modified files:
- All year ranges in the copyright header should be modified accordingly.
- The translation of time zone generic names were added to all locales.
I haven't fully reviewed translations, especially all \uxxxx strings.
But I noticed the following.
Common to all TimeZoneNames_*.java:
- The same generic abbreviation is used for the long name in MET. I
thought this was fixed in the root properties...
- Some generic names don't match the style of their standard and/or
daylight time names.
src/share/classes/sun/util/resources/de/TimeZoneNames_de.java:
- Some generic names use "Normalzeit". Is that OK?
src/share/classes/sun/util/resources/ja/TimeZoneNames_ja.java:
- "Chuuk Time" isn't translated.
- Time Zone names were updated according to the latest translations.
- Added tz names regression test
(test/sun/util/resources/TimeZone/TimeZoneNames) for a timezone names
across all locales. This test compares short/long
standard/daylight/generic names with translations stored in
.properties files.
test/sun/util/resources/TimeZone/TimeZoneNames/TimeZoneNamesTest.java:
- lines 33, 34: unused imports?
- line 75: Should it be detected as an error?
- line 108: IOException should be used as a cause. (OK to assume "not
found"?)
- lines 118 -121: Locale.getDefault() has to be replaced with
Locale.ROOT. Regression tests are run in different locales.
- Tests updates to address current changes (
GenericTimeZoneNamesTest.sh and Bug6317929.java).
The TODO comment in GenericTimeZoneNamesTest.sh should fully been removed.
The following fix was tested with JPRT build and the 'jdk_util' test
set succeeded (new test is included in this test set).
Have you executed all date-time related regression tests (including
java.time and java.text)?
Thanks,
Masayoshi
[1] http://cr.openjdk.java.net/~aefimov/8025051/8/webrev.00/
[2] https://bugs.openjdk.java.net/browse/JDK-8025051
