You could use regex to trim off all y
's and any non-alphabetic characters before and after, if any. Here's a kickoff example:
public static void main(String[] args) throws Exception {
for (Locale locale : Locale.getAvailableLocales()) {
DateFormat df = getShortDateInstanceWithoutYears(locale);
System.out.println(locale + ": " + df.format(new Date()));
}
}
public static DateFormat getShortDateInstanceWithoutYears(Locale locale) {
SimpleDateFormat sdf = (SimpleDateFormat) DateFormat.getDateInstance(DateFormat.SHORT, locale);
sdf.applyPattern(sdf.toPattern().replaceAll("[^\\p{Alpha}]*y+[^\\p{Alpha}]*", ""));
return sdf;
}
You see that this snippet tests it for all locales as well. It looks to work fine for all locales here.