C.UTF-8 locale

Jacek Konieczny jajcus at jajcus.net
Wed Dec 18 17:42:36 CET 2013


On 2013-12-18 17:35, Jakub Bogusz wrote:
> C is handled in code as special case (not included in localedb).
> C.UTF-8 would be larger, probably too large for builtin... see some
> RH/Fedora discussion:
> https://bugzilla.redhat.com/show_bug.cgi?id=902094
> 
> Debian has C.UTF-8 (as generated locale I think).

I guess C.UTF-8 could be efficiently handled as a special case in the
code (just process the encoding, use Unicode character order for
collation, some efficient bitmap/hash for character classes), but that
is a bigger change in glibc.

I think providing it as a regular 'built' locale should be good enough
for us.

> As can be googled, some other packages required changes to support it
> properly.

It won't be used by default (not at this point), so I guess this can be
solved later.

Greets,
	Jacek


More information about the pld-devel-en mailing list