Re: canonicalize-lgpl bug

[Eric Blake]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

According to Jim Meyering on 9/18/2009 1:40 AM:
> + mkdir subdir removed
> + touch regfile
> + ln -s regfile link1
> + ln -s subdir link2
> + ln -s missing link3
> + ln -s subdir/missing link4
...
> ++ readlink -f subdir/more
> + v=/sh/j/w/co/cu/tests/cu-can-f.Y2RpVABkn2/d/subdir/more
> + test /sh/j/w/co/cu/tests/cu-can-f.Y2RpVABkn2/d/subdir/more = 
> /sh/j/w/co/cu/tests/cu-can-f.Y2RpVABkn2/d/subdir/more
> ++ readlink -f ./subdir/more/
> + v=/sh/j/w/co/cu/tests/cu-can-f.Y2RpVABkn2/d/subdir/more
> + fail=1

Yep - back to the argument of whether we special case dangling symlinks
with a trailing slash.  Part of me argues that this behavior change is
intentional, since:

mkdir subdir/more/

will succeed.  But whether we change the test depends on whether we decide
that I need to fix this regression in gnulib.

> ++ readlink -f missing
> + v=/sh/j/w/co/cu/tests/cu-can-f.Y2RpVABkn2/d/missing
> + test /sh/j/w/co/cu/tests/cu-can-f.Y2RpVABkn2/d/missing = 
> /sh/j/w/co/cu/tests/cu-can-f.Y2RpVABkn2/d/missing
> ++ readlink -f ./missing/
> + v=/sh/j/w/co/cu/tests/cu-can-f.Y2RpVABkn2/d/missing
> + fail=1

Here, the test is just wrong; we really did intentionally change the
semantics.

It looks like all the other fail=1 listings in the log are variants of the
above two scenarios (missing/ is an intentional change, and dangling/ is a
questionable one).

- --
Don't work too hard, make some time for fun as well!

Eric Blake             address@hidden
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (Cygwin)
Comment: Public key at home.comcast.net/~ericblake/eblake.gpg
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/

iEYEARECAAYFAkqzcecACgkQ84KuGfSFAYCBRwCgku78Xmnjp0nol7v3kOPTrvW8
gYEAn3GXLcKkJQeZpp+szTGB5qyIgbh5
=rgUj
-----END PGP SIGNATURE-----