[Bug localedata/10502] sorting between Indic Languages should be as per unicode code point

[fweimer at redhat dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=10502

Florian Weimer <fweimer at redhat dot com> changed:

What |Removed |Added
----------------------------------------------------------------------------
Flags| |security-

[vapier at gentoo dot org]

https://sourceware.org/bugzilla/show_bug.cgi?id=10502

--- Comment #1 from Mike Frysinger <vapier at gentoo dot org> ---
can you post some example unicode data that shows incorrect sorting ? that'll
make it easier for us to integrate into tests to prevent future regressions.

[pravin.d.s at gmail dot com]

https://sourceware.org/bugzilla/show_bug.cgi?id=10502

--- Comment #2 from Pravin S <pravin.d.s at gmail dot com> ---
Hi Mike,

This is not about incorrect sorting. But what should be order when different
scripts come together. Example:

I think order for these codepoints should be as follows:
u+0915, u+0995, u+0A15, u+0A95, u+0B15, u+0B95

I dont remember any reference as of now, but when we decide sorting order
between different Unicode script, what should we follow? And IMO answer is
http://www.unicode.org/Public/UCA/latest/allkeys.txt