Control: tags -1 fixed-upstream
Vincent Lefevre wrote on Mon, Dec 20, 2010 at 11:17:05 +0100:
> How the path is resolved should be clarified, in particular because
> one doesn't get the same result as realpath for the first test. The
> zshexpn(1) man page currently says:
>
> a Turn a file name into an absolute path: prepends the
> current directory, if necessary, and resolves any use of
> `..' and `.' in the path. Note that the transformation
> takes place even if the file or any intervening directo‐
> ries do not exist.
>
> A As `a', but also resolve use of symbolic links where
> possible. Note that resolution of `..' occurs before
> resolution of symbolic links. This call is equivalent
> to a unless your system has the realpath system call
> (modern systems do).
>
> The fact that it refers to the realpath call (BTW this is not a
> system call, but a GNU libc call under Linux) is strange because
> neither ":a" not ":A" give the same result as realpath.
Docs clarified upstream in users/21777, users/21779, and [soon]
users/21784.
> Moreover the error condition (no matches found) is inconsistent,
> as /tmp/foo/foo/symlink-to-dot doesn't exist either.
echo symlink-to-dot/../$d/does-not-exist(:A)
.
yields "no matches found" because the NULLGLOB option is unset (and the
part before the qualifiers does not exist, according to stat()). The
other case,
.
echo symlink-to-dot/../$d/symlink-to-dot(:A)
.
, behaves as documented. In particular, it performs the transformation
even though /tmp/foo/$d does not exist.
I note the upstream threads have proposed adding a :P modifier that
simply transforms the argument using realpath(3), however, no patch
implementing that has yet been submitted.
Cheers,
Daniel
