Re: posix me harder

2002-03-14 Thread Martin Pool
> This does not demonstrate that test(1) is broken. I should have left off the '(1)'. Apparently different systems support different options. I suppose the best portable test for symlinks will be if test -L "$f" || test -h "$f" || /bin/test -L "$f" || /bin/test -h "$f" then echo is a

Re: posix me harder

2002-03-14 Thread Mike Gerdts
On Thu, 2002-03-14 at 08:11, David Collier-Brown wrote: > Fixed, in Solaris 8, but broken in > at least 2.6 and possibly 7. Not fixed in Solaris 8: $ ps -fp $$ UID PID PPID CSTIME TTY TIME CMD mgerdts 7104 12228 0 11:34:30 pts/10:00 /bin/sh $

Re: posix me harder

2002-03-14 Thread David Collier-Brown
Mike Gerdts wrote: > $ what /bin/sh > /bin/sh: > SunOS 5.8 Generic 109324-04 October 20 Aha! Fixed in /bin/test, and in at least ksh but not fixed in sh, which seems to use a builtin! Try if /bin/test -L foo ; then echo "foo is a symlink" ;f

Re: posix me harder

2002-03-14 Thread Jos Backus
On Thu, Mar 14, 2002 at 07:40:53AM -0500, Mike Gerdts wrote: > This does not demonstrate that test(1) is broken. First of all, pretty > much every shell these days has test as a builtin, and as such you > should be accusing sh(1), ksh(1), or similar. I know, I should have dropped the ``(1)'' fro

Re: posix me harder

2002-03-14 Thread David Collier-Brown
Martin Pool wrote: > is that Sun's test(1) is breathtakingly broken when passed > a dangling symlink: > > $ ln -s /nowhere bad > $ ls -l bad > lrwxrwxrwx 1 josb user8 Mar 13 20:10 bad -> /nowhere > $ if test -L bad; then echo bad; fi > test: argument expected > $ if test -h bad

Re: posix me harder

2002-03-14 Thread Mike Gerdts
On Thu, 2002-03-14 at 01:37, Martin Pool wrote: > People might find this entertaining and/or useful: > > http://www.gnu.org/manual/autoconf-2.52/html_chapter/autoconf_10.html > > With the help of Jos Backus I just discovered the answer to > > http://lists.samba.org/pipermail/rsync-cvs/2002-

posix me harder

2002-03-13 Thread Martin Pool
People might find this entertaining and/or useful: http://www.gnu.org/manual/autoconf-2.52/html_chapter/autoconf_10.html With the help of Jos Backus I just discovered the answer to http://lists.samba.org/pipermail/rsync-cvs/2002-January/001271.html is that Sun's test(1) is breathtakingly b