[PATCH] t0109: chain operations with &&

[PATCH] t0109: chain operations with &&

[john]

Without '&&' between operations, we will not detect if strace or cgit
exit with an error status, which would cause a false positive test
status in this case.

Signed-off-by: John Keeping <john at keeping.me.uk>
---
 tests/t0109-gitconfig.sh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/tests/t0109-gitconfig.sh b/tests/t0109-gitconfig.sh
index cdd570b..f64e691 100755
--- a/tests/t0109-gitconfig.sh
+++ b/tests/t0109-gitconfig.sh
@@ -18,7 +18,7 @@ test_expect_success 'no access to $HOME' '
 		-E HOME="$non_existant_path" \
 		-E CGIT_CONFIG="$PWD/cgitrc" \
 		-E QUERY_STRING="url=foo/commit" \
-		-e access -f -o strace.out cgit
+		-e access -f -o strace.out cgit &&
 	test_must_fail grep "$non_existant_path" strace.out
 '
 
-- 
1.8.2.694.ga76e9c3.dirty

[PATCH] t0109: chain operations with &&

[Jason]

Thanks. Merged.

[PATCH] t0109: chain operations with &&

[Jason]

On Wed, Apr 10, 2013 at 2:11 PM, John Keeping <john at keeping.me.uk> wrote:
> Without '&&' between operations, we will not detect if strace or cgit
> exit with an error status, which would cause a false positive test
> status in this case.

I wonder if for some of the larger && chains in other tests "set -e"
would be a bit cleaner. I don't know if setting that inside the test
string will bubble the setting back up to the test harness (which
would be bad) or if it would be local for the test (which would be
ideal). Something to investigate.

[PATCH] t0109: chain operations with &&

[john]

On Wed, Apr 10, 2013 at 02:16:44PM +0200, Jason A. Donenfeld wrote:
> On Wed, Apr 10, 2013 at 2:11 PM, John Keeping <john at keeping.me.uk> wrote:
> > Without '&&' between operations, we will not detect if strace or cgit
> > exit with an error status, which would cause a false positive test
> > status in this case.
> 
> I wonder if for some of the larger && chains in other tests "set -e"
> would be a bit cleaner. I don't know if setting that inside the test
> string will bubble the setting back up to the test harness (which
> would be bad) or if it would be local for the test (which would be
> ideal). Something to investigate.

I expect if it was that easy, the tests in Git would take an approach
more like that.  Personally, I have found in the past that using
"set -e" often leads to problems later on, so I would prefer to keep the
&&-chain.