On 2024-04-17 08:48, Eric Sunshine wrote:
On Tue, Apr 16, 2024 at 11:33 PM Dragan Simic <dsimic@xxxxxxxxxxx>
wrote:
Add a few new tests to the t4014 that cover the --resend command-line
option
for "git format-patch", which include the tests for its exclusivity
with the
already existing -k and --rfc command-line options.
I'd recommend squashing this patch into [3/4] which introduces the
--resend option since it's easier to review the tests when the code
which is being tested is still fresh in one's mind. (For the same
reason, reviewers like to see documentation added in the same patch
which changes the code since it's easier to verify that the
documentation matches the implementation while it's fresh in the
mind.)
I'm fine with that. Squashing these two patches together might also
be good for bisecting later, if need arises.
Signed-off-by: Dragan Simic <dsimic@xxxxxxxxxxx>
---
diff --git a/t/t4014-format-patch.sh b/t/t4014-format-patch.sh
@@ -1401,6 +1401,43 @@ test_expect_success '--rfc and -k cannot be
used together' '
+test_expect_success '--resend' '
+ cat >expect <<-\EOF &&
+ Subject: [PATCH RESEND 1/1] header with . in it
+ EOF
In all of the new tests, since it's just a single line body, it could
just as easily be created with `echo`:
echo "Subject: [PATCH RESEND 1/1] header with . in it" >expect &&
On the other hand, if you're following precedent in this script, then
using a here-doc may be just fine.
I agree that using "echo ..." would be nicer. Though, I just wanted
to follow the already existing tests for consistency, which may actually
outweigh a nicer approach.
At any rate it's somewhat subjective and not worth a reroll.
+ git format-patch -n -1 --stdout --resend >patch &&
+ grep "^Subject:" patch >actual &&
+ test_cmp expect actual
+'
