Jim Meyering [Fri, 12 Apr 2002 20:50:53 +0000 (20:50 +0000)]
(checking for getmntinfo function...): Remove now-bogus
check for f_type in sys/mount.h. Instead, just test for the existence
of the getmntinfo function. Needed for Darwin 5.3.
Jim Meyering [Fri, 12 Apr 2002 11:02:59 +0000 (11:02 +0000)]
Don't AC_REPLACE(strnlen), now that we use
AC_FUNC_STRNLEN. That would end up putting two copies of strnlen.o
in the library, and that makes some versions of ranlib object.
Jim Meyering [Fri, 12 Apr 2002 10:41:54 +0000 (10:41 +0000)]
Include long-options.h.
[long_opts]: Remove.
(usage): Tweak --help output; use *_OPTION_DESCRIPTION macros.
(main): Don't use getopt directly. Use parse_long_options instead.
Tweak a diagnostic.
Use EXIT_FAILURE rather than a literal `1'.
(main): If POSIXLY_CORRECT is set, don't recognize --help or --version,
so the program can operate on a file with one of those names.
Jim Meyering [Fri, 12 Apr 2002 10:41:22 +0000 (10:41 +0000)]
Include long-options.h.
[long_opts]: Remove.
(usage): Tweak --help output; use *_OPTION_DESCRIPTION macros.
(main): Don't use getopt directly. Use parse_long_options instead.
Tweak a diagnostic.
Use EXIT_FAILURE rather than a literal `1'.
Jim Meyering [Tue, 9 Apr 2002 17:55:36 +0000 (17:55 +0000)]
(jm_FUNC_REALLOC): Change the `checking ...' message
to be more precise. Rather than saying we're checking whether the
function `works', say what we're testing.
Jim Meyering [Tue, 9 Apr 2002 17:55:23 +0000 (17:55 +0000)]
(jm_FUNC_MALLOC): Change the `checking ...' message
to be more precise. Rather than saying we're checking whether the
function `works', say what we're testing.
Jim Meyering [Sat, 30 Mar 2002 07:10:57 +0000 (07:10 +0000)]
(copy_internal): Move the block that sets `earlier_file'
down to just before the first use of that variable. Otherwise, it was
possible to make mv (and probably cp, too) malfunction when copying
hard-linked files into a directory containing at least one of the
source file names. Call forget_created everywhere thereafter where
this function returns without creating a destination file that might
subsequently be linked. Reported by Iida Yosiaki.
Jim Meyering [Mon, 25 Mar 2002 09:53:07 +0000 (09:53 +0000)]
(age_of): Return -1 and 0 rather than 0 and 1.
Might as well keep it simple, and like bash.
(binary_operator): Fix bug with -nt and -ot, when one of the
files did not exist. We want to be compatible with the ksh93
documentation, and with Bash.
Jim Meyering [Mon, 25 Mar 2002 09:52:25 +0000 (09:52 +0000)]
(File characteristic tests): Document the
behavior of test -nt and -ot when one of the files does not exist,
using the same behavior that is documented in ksh93.
Jim Meyering [Sun, 17 Mar 2002 19:21:16 +0000 (19:21 +0000)]
(usage): Mention that --format=FORMAT must be
a *floating-point* format, also in description of that option.
(usage): Also add the `=' signs here: --format=FORMAT,
--separator=STRING.
Jim Meyering [Sun, 17 Mar 2002 16:00:54 +0000 (16:00 +0000)]
(copy_internal) [move_mode]: Give a better diagnostic
(using errno from the failed unlink) when a cross-device `mv'
fails, e.g., because the destination cannot be unlinked.
Prompted by a report from Karl Berry.
Jim Meyering [Sun, 10 Mar 2002 08:17:20 +0000 (08:17 +0000)]
* src/copy.c (copy_reg): Use a more concise diagnostic for
reporting replaced files. This avoids a bug in the code,
which mishandled ino_t wider than long.
* src/remove.c (remove_dir): Likewise, twice.