Yes. Well, no. Cygpath isn't "chok[ing]." It's operating correctly.
It accepts one input name per line when used in filter mode. You can use tr or sed to insert newlines where appropriate, but it's not feasible to have cygpath interpret spaces as separators between input names. Since spaces are valid in file names, that's the only reasonable (reasonably general) way for it to work. (Yes, under Unix / Linux / etc., newlines are acceptable file name characters, too, but are rarely used as such.)
Randall Schulz
At 00:47 2003-03-10, Ralf Hauser wrote:
in http://sources.redhat.com/ml/cygwin/2003-01/msg01707.html I thought to see how I should do this. The example works fine. But if I create a script that does the same
echo "$@" | cygpath -w -f -
and call it with
testscript `/bin/ls /bin/ch*`
I get
C:\cygwin\bin\checkgid.exe \bin\chgrp.exe \bin\chmod.exe \bin\chown.exe \bin\chroot.exe
is it possible that it chokes upon missing newlines (CR/LF)? Or even worse if I put in the long path:
testscript `/bin/ls /cygdrive/c/cygwin/bin/ch*`
cygpath: error converting "/cygdrive/c/cygwin/bin/checkgid.exe /cygdrive/c/cygwin/bin/chgrp.exe /cygdrive/c/cygwin/bin/chmod.exe /cygdrive/c/cygwin/bin/chown.exe /cygdrive/c/cygwin/bin/chroot.exe"
or yet worse when I do that on my desktop with three test files a.txt, aa.txt and aaa.txt: [EMAIL PROTECTED]:/<3>rhauser/Desktop> testscript `ls /cygdrive/c/Docume~1/rhauser/Desktop/a*.txt` c:\Docume~1\rhauser\Desktop\a.txt \cygdrive\c\Docume~1\rhauser\Desktop\aa.txt \cygdrive\c\Docume~1\rhauser\Desktop\aaa.txt
so the first path is converted properly while number 2 and 3 are wrong?
Any thoughts?
Ralf
-- Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple Bug reporting: http://cygwin.com/bugs.html Documentation: http://cygwin.com/docs.html FAQ: http://cygwin.com/faq/
