I tested init.sh on Solaris 10 by doing this locally,
./gnulib-tool --create-testdir --with-tests --dir=tt xstrtoll
Running (in tt) ./configure && make && make dist, and copying the tarball
to the Solaris system. There, I untarred and ran the usual
./configure && make && make check and got this:
./test-xstrtoll.sh: syntax error at line 132: `base_names_=$' unexpected
FAIL: test-xstrtoll.sh
While eventually we will be guaranteed to have a shell with
sufficient functionality to handle $(...) and more, this avoids one
pointless failure right now.
>From 0ce0394a2834b9edb76e2f70af71ab8a5738e68c Mon Sep 17 00:00:00 2001
From: Jim Meyering <meyer...@redhat.com>
Date: Thu, 18 Feb 2010 10:28:24 +0100
Subject: [PATCH] init.sh: don't use $(...) just yet
* tests/init.sh (create_exe_shim_functions_): Use `...`, not $(...),
to accommodate e.g., Solaris' /bin/sh.
---
ChangeLog | 6 ++++++
tests/init.sh | 2 +-
2 files changed, 7 insertions(+), 1 deletions(-)
diff --git a/ChangeLog b/ChangeLog
index a0f0b94..d73841a 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,9 @@
+2010-02-18 Jim Meyering <meyer...@redhat.com>
+
+ init.sh: don't use $(...) just yet
+ * tests/init.sh (create_exe_shim_functions_): Use `...`, not $(...),
+ to accommodate e.g., Solaris' /bin/sh.
+
2010-02-17 Bruno Haible <br...@clisp.org>
* doc/posix-headers/netdb.texi: Mention NetBSD 5.0 problem.
diff --git a/tests/init.sh b/tests/init.sh
index 02f53c8..e7664b8 100644
--- a/tests/init.sh
+++ b/tests/init.sh
@@ -129,7 +129,7 @@ create_exe_shim_functions_()
*) echo "$0: unexpected \$EXEEXT value: $EXEEXT" 1>&2; return 1 ;;
esac
- base_names_=$(find_exe_basenames_ $1) \
+ base_names_=`find_exe_basenames_ $1` \
|| { echo "$0 (exe_shim): skipping directory: $1" 1>&2; return 1; }
if test -n "$base_names_"; then
--
1.7.0.233.g05e1a
