The current glibc localedata does not define a collation order for Sinhala
(si), U+0D80 to U+0DFF. The collation order is defined by SCCII Part 1 :
Collation Sequence (SLS1134).
I have a patch against localedata's iso14651_t1_common file, which I'll attach
to this bug.
Created attachment 3007 [details]
Define Sinhala (si) collation order
Created attachment 3008 [details]
Created attachment 3009 [details]
Correct output with LC_ALL=si_LK.utf8
Added to cvs.