Re: [cp-patches] Locale update and addition to java.util.Currency

[Michael Koch]

On Mon, Jan 31, 2005 at 01:35:08AM +0000, Andrew John Hughes wrote:
> I'm committing an update to the locales, along with a java.util.Currency
> fix (necessary, as the current code breaks with the new locales).
> I've not attached a patch as this is over 3mb; if anyone wants a copy,
> they are welcome to mail me for one ;)
> 
> The changes can be summarized as follows:
> 
>       * Fixed the ordering of the zone information, as suggested
>       by Ito.  The ID is now the first element, rather than the last.
>       * The locales all now contain \u00A6 (a broken bar) as the separator
>       rather than the | added by Mark.  Using the | as a separator breaks
>       when parsing the currency symbol for Indian rupees.  Hopefully, the
>       new value is unique enough (as well as being a pretty nifty pointer
>       to what our original implementation was).
>       * The data is now up-to-date with the latest CLDR data pulled from
>       CVS.  This also has the advantage of unifying our locale data, some
>       of which appears to have not been updated on the last run (root.xml
>       in particular, which broke down badly when updated).  The files
>       are now as generated.  If you find a regression due to missing data,
>       please report this as a bug.  Ideally, we can then fix it by adapting
>       the generator rather than having locale data as a hybrid of automatic
>       and manual updates.
>       * The changes broke our java.util.Currency, which gave me an incentive
>       to also finally add the update to this class, which allows our currency
>       symbols to be localized and correctly match the specification.  Before,
>       they were simply pulled from gnu.java.locale.LocaleInformation in
>       an unlocalized form.  As part of this, I also added a new class,
>       gnu.java.locale.LocaleHelper, which currently contains one static
>       method, getLocalizedString().  This is basically an adapted version
>       of a method I wrote for java.util.Locale, but which would seem to be
>       useful in other contexts (two at least so far ;) )
> 

You should at least the patch for the code changes but I agree that the
patch for the locale data itself is not needed.


Michael