See the result below from the test prgram:
import java.text.*;
import java.util.*;
class dttest {
static final int[] format = {
DateFormat.FULL,
DateFormat.LONG,
DateFormat.MEDIUM,
DateFormat.SHORT,
};
public static void main(String[] arg) {
Date date = new Date();
for (int i = 0; i<format.length; i++) {
DateFormat fmt = DateFormat.getDateTimeInstance(format[i],
format[i]);
DateFormat fmt_en = DateFormat.getDateTimeInstance(format[i],
format[i], Locale.ENGLISH);
System.out.println(fmt.format(date));
System.out.println(fmt_en.format(date) + "\n");
}
}
}
Test was done on Solaris2.5.1 under ja locale.
1.1.5 fcs
1998 03 03 15 17 21 JST
Monday, March 2, 1998 10:17:21 o'clock PM PST
1998/03/03 15:17:21:JST
March 2, 1998 10:17:21 PM PST
98/03/03 15:17:21
02-Mar-98 10:17:21 PM
03/03 15:17
3/2/98 10:17 PM
1.1.6H
1998 3 3 13 24 53 GMT+09:00
Tuesday, March 3, 1998 1:24:53 o'clock PM JST
1998/03/03 13:24:53:GMT+09:00
March 3, 1998 1:24:53 PM JST
1998/03/03 13:24:53
03-Mar-98 1:24:53 PM
98/03/03 13:24
3/3/98 1:24 PM
Problems:
Time zone is always GMT+09:00 under ja locale.
because kZoneStrings in DateFormatZoneData_ja containes incorrect timezone
string.
This is 116 regression.
Formatted date and time is inconsistent between Japanese and English locale.
The full format on 1.1.5 has the day of the week, but 1.1.6 doesn't.
English has AM/PM but not for Japanese on both 1.1.5 and 1.1.6.
This is probably RFE.
Era definition is missing for Japanese locale.
This is regression.
There is no equivarent entry in LocaleElement
{ "Eras",
new String[] { // era strings
"BC",
"AD"
}
for Japanese locale. It should be translated to Kigenzen and Kigengo.