ICU4C Proposal: uloc_canonicalize

I'd like to add the following function

U_CAPI int32_t U_EXPORT2
uloc_canonicalize(const char*    localeID,
         char* name,
         int32_t nameCapacity,
         UErrorCode* err);

It will work similar to uloc_getName(), but it will convert currently 
registered IANA language tags 
<http://www.iana.org/assignments/language-tags>, Microsoft .NET and POSIX 
style locale names to the current ICU format for locale identifiers.

Here are some examples:

de-1901 ->      de__1901        (registered IANA language tag)
uz-UZ-Cyrl      ->      uz_Cyrl_UZ      (.Net name)
C               ->      en_US_POSIX     (The C locale is a common POSIX 
convention)
qz-qz@EURO      ->      qz_QZ@currency=EUR (generic conversion of a POSIX 
Euro locale name)

This function won't be as fast as uloc_getName, but it will allow you to 
get a proper ICU locale ID based on a locale identifier coming from other 
sources of localized data.

deadline for comments: November 5th, 2003

George Rhoten
IBM Globalization Center of Competency/ICU  San José, CA, USA