On Tue, Dec 17, 2019 at 7:02 AM Denton Liu <liu.denton@xxxxxxxxx> wrote:
> Before, we were running `test_must_fail full_name`. However,
> `test_must_fail` should only be used on git commands. Teach full_name()
> to accept `!` as a potential first argument which will prepend
> `test_must_fail` to the enclosed git command. This increases the
> granularity of test_must_fail since it no longer runs on the `cd` as
> well.
>
> Signed-off-by: Denton Liu <liu.denton@xxxxxxxxx>
> ---
> diff --git a/t/t1507-rev-parse-upstream.sh b/t/t1507-rev-parse-upstream.sh
> @@ -29,8 +29,14 @@ test_expect_success 'setup' '
> full_name () {
> + fail= &&
> + if test "x$1" = 'x!'
> + then
> + fail=test_must_fail &&
> + shift
> + fi &&
> (cd clone &&
> - git rev-parse --symbolic-full-name "$@")
> + $fail git rev-parse --symbolic-full-name "$@")
> }
Yuck. These days this is entirely unnecessary. My suggestion is to
drop the full_name() function altogether and just invoke git-rev-parse
directly at the (few) call sites, taking advantage of the fact that we
now have -C. So...
> @@ -79,7 +85,7 @@ test_expect_success 'upstream of branch with @ at end' '
> test_expect_success 'refs/heads/my-side@{upstream} does not resolve to my-side{upstream}' '
> - test_must_fail full_name refs/heads/my-side@{upstream}
> + full_name ! refs/heads/my-side@{upstream}
> '
This just becomes:
test_must_fail git -C clone rev-parse --symbolic-full-name
refs/heads/my-side@{upstream}
Similarly for the other call sites.
> @@ -91,9 +97,9 @@ test_expect_success 'my-side@{u} resolves to correct commit' '
> test_expect_success 'not-tracking@{u} fails' '
> - test_must_fail full_name non-tracking@{u} &&
> + full_name ! non-tracking@{u} &&
> (cd clone && git checkout --no-track -b non-tracking) &&
> - test_must_fail full_name non-tracking@{u}
> + full_name ! non-tracking@{u}
> '
