On Wed, Dec 30, 2009 at 01:30:19PM -0500, Christopher Faylor wrote:
>On Wed, Dec 30, 2009 at 05:36:05PM +0100, Bengt Larsson wrote:
>>I seem to have a problem with wildcards from the Windows command line
>>when there are high-bit characters in a filename.
>>
>>A directory contains only the two files "user" and "anv?ndare"
>>("anv?ndare" being user in Swedish):
>>
>> C:\Documents and Settings\Bengt2\Desktop\test\ttt>ls -l
>> total 0
>> -rw-r--r-- 1 Bengt2 Users 0 2009-12-30 02:23 anv?ndare
>> -rw-r--r-- 1 Bengt2 Users 0 2009-12-30 02:23 user
>>
>> C:\Documents and Settings\Bengt2\Desktop\test\ttt>ls u*
>> user
>>
>> C:\Documents and Settings\Bengt2\Desktop\test\ttt>ls a*
>> ls: cannot access a*: No such file or directory
>
>I get a different (and slightly more understandable) error
>(not sure how this will translate to my email client from vnc):
>
>c:\tmp> ls a*
>ls: cannot access anvA?ndare: No such file or directory
>autoruns.chm autoruns.exe autorunsc.exe
>
>I'm not really familiar with all of the changes that Corinna made
>for this but I'll take a look to see if this is easy to fix.
It looks like things should work better if you set LANG appropriately.
For instance, this works for me:
c:\>set LANG=en_US.UTF-8
c:\>ls a*
What's happening is that cygwin's readdir is converting the on-disk
filename to utf-8 but cygwin's open/stat apparently aren't converting it
back. As I said, I don't know enough about the labyrinthian multi-byte
character set code that is now in cygwin but I really can't see how this
could be anything other than a bug. If readdir() returns a filename
then open should be able to open it.
cgf
--
Problem reports: http://cygwin.com/problems.html
FAQ: http://cygwin.com/faq/
Documentation: http://cygwin.com/docs.html
Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
