Patch: Store correct collation sequence values.

[Isamu Hasegawa]

Hello,

Attached program shows the collation sequnce values for 'a' and "a-a"
in da_DK.ISO-8859-1 locale.

$ ./tst-collseq
current locale : da_DK.ISO-8859-1
multibyte collation sequence for 'a': 141
wide char collation sequence for 'a': 0
multibyte collation sequence for 'a-a': 0
wide char collation sequence for 'a-a': 0

These values except multibyte value for 'a' are incorrect.
Attached patch (and rebuilding the locale) seems to fix the several
problems.  Is this patch appropriate?

However it isn't trivial to fix the multibyte collation sequence value
for "a-a".

In addition, I'd like to know how to lookup collation sequence value
for multi character collating elements by *byte sequence* (not by name).
I guess that there are no methods.

If no one works on these problem, please let me know.
We are ready to contribute.

Thanks,
--
Isamu Hasegawa
IBM Japan, Ltd.

2002-01-18  Isamu Hasegawa  <isamu@yamato.ibm.com>

	* locale/program/ld-collate.c (collate_finish): assign a wide
	char collation sequence value to multi character collating
	elements, and avoid over writing by non-character elements.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: patch
Type: application/octet-stream
Size: 954 bytes
Desc: not available
URL: <http://sourceware.org/pipermail/libc-alpha/attachments/20020118/177cb918/attachment.obj>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: tst-collseq.c
Type: application/octet-stream
Size: 2310 bytes
Desc: not available
URL: <http://sourceware.org/pipermail/libc-alpha/attachments/20020118/177cb918/attachment-0001.obj>