Am 08.05.2011 23:30, schrieb Junio C Hamano:
Jens Lehmann [off-list ref] writes:
quoted
diff --git a/t/t1000-read-tree-m-3way.sh b/t/t1000-read-tree-m-3way.sh
index ca8a409..bcfb5e6 100755
--- a/t/t1000-read-tree-m-3way.sh
+++ b/t/t1000-read-tree-m-3way.sh
@@ -259,6 +259,8 @@ test_expect_success \
"rm -f .git/index AA &&
cp .orig-A/AA AA &&
git update-index --add AA &&
+ git read-tree -n -m $tree_O $tree_A $tree_B &&
+ test_must_fail check_result &&
That's a rather sloppy way to test that the command did not do anything to
compare one possible outcome, instead of verifying that the result is
identical to the original condition, no?
How about
...
git update_index --add AA &&
git ls-files -s >pre-dry-run &&
git read-tree -n -m $tree_O $tree_A $tree_B &&
git ls-files -s >post-dry-run &&
test_cmp pre-dry-run post-dry-run &&
...
Yeah, that is much better than my first sketch.
We also should be checking that the command reports it is going to fail
when it should as well. Always remember to check both sides of the coin.
quoted
git read-tree -m $tree_O $tree_A $tree_B &&
check_result"
I suspect that it would make sense to replace
git read-tree $args && check_result
with
read_tree_must_succeed $args
and
test_must_fail git read-tree $args
with
read_tree_must_fail $args
and implement two shell wrappers, perhaps like this.
read_tree_must_succeed () {
git ls-files -s >pre-dry-run &&
git read-tree -n "$@" &&
git ls-files -s >post-dry-run &&
test_cmp pre-dry-run post-dry-run &&
git read-tree "$@"
}
read_tree_must_fail () {
git ls-files -s >pre-dry-run &&
test_must_fail git read-tree -n "$@" &&
git ls-files -s >post-dry-run &&
test_must_fail git read-tree "$@"
}
Thanks, will send an updated patch with better test cases soon ...