Steve, Mark,
On 2005 Apr 11 , at 16.20, Rankin, SE (Stephen) wrote:
> I do not know if this is a bug but the tar.gz distribution file for SCB
> is missing size_test.sh but not scb_test.sh. The files are both
> specified in Makefile.am using TEST:
>
> TESTS = scb_test.sh size_test.sh
>
> So it would appear that automake is only getting the first file in the
> list?
Try this:
diff -u -r1.5 Makefile.am
--- Makefile.am 28 Feb 2005 20:23:51 -0000 1.5
+++ Makefile.am 11 Apr 2005 15:52:58 -0000
@@ -16,7 +16,8 @@
BUILT_SOURCES = Scb.pm YyTag.pm scb.csh scb.pl scbindex.pl scbcp.pl
dbmcat.pl \
ctag-y.h ftag-y.h ctag-l.c ftag-l.c
-TESTS = scb_test.sh size_test.sh
+TESTS = $(dist_check_SCRIPTS) scb_test.sh
+dist_check_SCRIPTS = size_test.sh
check_DATA = scb_test.tar
scb_test.tar:
TESTS simply names a set of programs/scripts to be run as checks, so
they're to be built at 'make check' time, but it doesn't imply anything
about whether they're to be distributed or not. The
'dist_check_SCRIPTS' variable indicates that size_test.sh is a script,
it's used for the check target, and it is to be distributed (which is
not the default for scripts, just as it's not the default for
*_PROGRAMS). You could name scb_test.sh as the value of a
check_SCRIPTS target, but there's really no need beyond symmetry.
scb_test.sh appeared after the distribution because scb_test.sh is
mentioned in AC_CONFIG_FILES in configure.ac, allowing automake to
deduce that scb_test.sh.in has to be distributed.
> I will add the file as an EXTRA_DIST for now.
That'll work OK, though the dist_check_SCRIPTS has the advantage of
being documentation, too.
Mark:
> Failing is the 'right' thing to do, since SCB is known to be somewhat
> (though not disastrously) broken on some systems in this respect,
> and as mentioned above I'm not inclined to work hard to fix it just
> now.
Can the script (be made to) tell when it's on a system it can't succeed
on? If so, and if it returns the magic value 77 in that case, then its
failure will be ignored.
Note, I do not claim that any of the above points are Obvious....
See you,
Norman
--
----------------------------------------------------------------------
Norman Gray : Physics & Astronomy, Glasgow University, UK
http://www.astro.gla.ac.uk/users/norman/ : www.starlink.ac.uk
|