Re: [PATCH v4 0/4] Introduce diff.submodule

From: Ramkumar Ramachandra <artagnon@gmail.com>
Date: Nov 16, 2012, 10:45 UTC
Message-ID: <CALkWK0k65oyz1FWdyjjHLjj-DS75cdfk_8Vd3NV4OwzS+--VaA@mail.gmail.com>
In-Reply-To: <20121115163356.GF6157@sigill.intra.peff.net>

Jeff King wrote:

> It may be worth squashing this test into patch 3:

Looks good.  Thanks.

Show 5 quoted lines

> BTW, while writing the test, I noticed two minor nits with your tests:
>
>   1. They can use test_config, which is simpler (you do not need to
>      unset yourself after the test) and safer (the unset happens via
>      test_when_finished, so it works even if the test fails).

I see.  Can you squash this in?

-- 8< --

diff --git a/t/t4041-diff-submodule-option.sh b/t/t4041-diff-submodule-option.sh
index e401814..876800f 100755
--- a/t/t4041-diff-submodule-option.sh
+++ b/t/t4041-diff-submodule-option.sh
@@ -56,7 +56,7 @@ EOF
 "

 test_expect_success '--submodule=short overrides diff.submodule' "
-	git config diff.submodule log &&
+	test_config diff.submodule log &&
 	git add sm1 &&
 	git diff --submodule=short --cached >actual &&
 	cat >expected <<-EOF &&
@@ -68,7 +68,6 @@ index 0000000..a2c4dab
 @@ -0,0 +1 @@
 +Subproject commit $fullhead1
 EOF
-	git config --unset diff.submodule &&
 	test_cmp expected actual
 "

>   2. You can still indent expected output when using <<-.

I know;  however, I wanted to be consistent with the surrounding code.

Ram