Re^7: Sorting Vietnamese text

[Atacama]

Any other ideas?

A quick search shows that vietnamese words are sorted by letters first, then tone-marks. Sometimes tone-marks are even ignored. So that's probably correct:

ỳ :
ỷ :
ỳ ạch :

http://vietunicode.sourceforge.net/charset/quytacABC_en.html ...looks like Unicode::Collate does it right, but additional first-character ordering is required to get a dictionary order (happens to be different from a simple sorted order in vietnamese).