Phillip Wood <phillip.wood123@xxxxxxxxx> writes:
> From: Phillip Wood <phillip.wood@xxxxxxxxxxxxx>
>
> As the tests are all run in separate repositories, set the branch name
> to "master" when creating the repository for the tests where the result
> depends on the branch name. This reduceds the number of tests that
> depend on the default branch name being "master" and removes the last
> instance of a test file using "GIT_TEST_DEFAULT_INITIAL_BRANCH_NAME=
> master".
>
> Signed-off-by: Phillip Wood <phillip.wood@xxxxxxxxxxxxx>
> ---
> t/t0613-reftable-write-options.sh | 20 +++++++++-----------
> 1 file changed, 9 insertions(+), 11 deletions(-)
>
> diff --git a/t/t0613-reftable-write-options.sh b/t/t0613-reftable-write-options.sh
> index d77e601111f..6d00432219f 100755
> --- a/t/t0613-reftable-write-options.sh
> +++ b/t/t0613-reftable-write-options.sh
> @@ -11,16 +11,14 @@ export GIT_TEST_REFTABLE_AUTOCOMPACTION
> # Block sizes depend on the hash function, so we force SHA1 here.
> GIT_TEST_DEFAULT_HASH=sha1
> export GIT_TEST_DEFAULT_HASH
> -# Block sizes also depend on the actual refs we write, so we force "master" to
> -# be the default initial branch name.
> -GIT_TEST_DEFAULT_INITIAL_BRANCH_NAME=master
> -export GIT_TEST_DEFAULT_INITIAL_BRANCH_NAME
>
> . ./test-lib.sh
>
> test_expect_success 'default write options' '
> test_when_finished "rm -rf repo" &&
> - git init repo &&
> + # Block sizes also depend on the actual refs we write, so we force
> + # "master" to be the default initial branch name.
A new developer who would want to add a new test later may or may
not pay attention to this first test piece that would look an equal
among many others. I suspect that keeping the comment before these
test sequences begin would have been better. Or perhaps append "all
the following test use the --initial-branch option for the same
reason" after it?
> + git init --initial-branch master repo &&
> + git init --initial-branch master repo &&
> + git init --initial-branch master repo &&
> + git init --initial-branch master repo &&
> + git init --initial-branch master repo &&
> + git init --initial-branch master repo &&
> + git init --initial-branch master repo &&
Having to change this many "git init" invocations feels awkward,
especially when we envision what needs to be done if somebody needs
to update this file later. I wonder if a new helper function
upfront
+init_repo () {
+ git init --initial-branch master repo
+}
+
with many changes like this to the rest of file
- git init repo &&
+ init_repo &&
may be easier to maintain over time? I dunno.
