[illumos-Developer] webrev: some locale cleanups/improvements

Garrett D'Amore garrett at damore.org
Thu Jan 13 12:30:24 PST 2011


http://mexico.purplecow.org/gdamore/webrev/noascii/

This removes support for 7-bit ASCII, kind of.  Basically, we can always 
treat that as 8-bit clean data, the same way we treat the C/POSIX locale 
and all 8859-x locales.  An extra enforcement that complains when you 
try to use a byte with the high order bit offers no value.  (This 
encoding scheme was never used, and was an addition from FreeBSD...  
Solaris and illumos don't have any locales that use it.)

At the same I've made some very minor changes to UTF-8 which eliminate 
some extra tests during conversion between wide and multi-byte string 
formats.  This will have a very small performance benefit on anything 
that does these operations -- which is most everything that does sorting 
or display, and a bunch of other things -- the path is actually pretty 
hot especially for those running graphical desktops.

These changes are obvious enough that they have a very short timeout.

    - Garrett




More information about the Developer mailing list