t: replace incorrect test_must_fail usage (part 1)
Get series with b4
b4 am [email protected]Versions newest first · linked by author and title
- v3 Latest15 patches 0 repliesNew
- v216 patches 3 repliesSuperseded
- v115 patches 12 repliesSuperseded
Changes in v3 from the cover letter
* Drop the inclusion of test_non_git_must_fail() and open code it instead
Patches in v3 No applies recorded yet
| # | Subject and files | Size | Replies | Applied |
|---|---|---|---|---|
| 1/15 | t/lib-git-p4: use test_path_is_missing()t/ | +1 | 0 replies | |
| 2/15 | t0000: replace test_must_fail with run_sub_test_lib_test_err()t/ | +7 | 0 replies | |
| 3/15 | t0003: use named parameters in attr_check()t/ | +4 | 0 replies | |
| 4/15 | t0003: use test_must_be_empty()t/ | +3 | 0 replies | |
| 5/15 | t0003: don't use `test_must_fail attr_check`t/ | +15 | 0 replies | |
| 6/15 | t0020: don't use `test_must_fail has_cr`t/ | +9 | 0 replies | |
| 7/15 | t0020: use ! check_packed_refs_markedt/ | +2 | 0 replies | |
| 8/15 | t1306: convert `test_might_fail rm` to `rm -f`t/ | +4 | 0 replies | |
| 9/15 | t1307: reorder `nongit test_must_fail`t/ | +1 | 0 replies | |
| 10/15 | t1409: let sed open its own input filet/ | +1 | 0 replies | |
| 11/15 | t1409: use test_path_is_missing()t/ | +5 | 0 replies | |
| 12/15 | t1501: remove use of `test_might_fail cp`t/ | +1 | 0 replies | |
| 13/15 | t1507: stop losing return codes of git commandst/ | +32 | 0 replies | |
| 14/15 | t1507: run commands within test_expect_successt/ | +20 | 0 replies | |
| 15/15 | t1507: inline full_name()t/ | +15 | 0 replies |
Reviews trailers per patch
No Reviewed-by, Acked-by or Tested-by trailers for v3 yet.
Cover letter [PATCH v3 00/15] t: replace incorrect test_must_fail usage (part 1)
The overall scope of these patches is to replace inappropriate uses of test_must_fail. IOW, we should only allow test_must_fail to run on git and test-tool. Ultimately, we will conclude by making test_must_fail error out on non-git commands. An advance view of the final series can be found here[1].
This is the first part. It focuses on t[01]*.sh and also t/lib-git-p4.
Changes since v2:
* Drop the inclusion of test_non_git_must_fail() and open code it instead
Changes since v1:
* Incorporate review comments by Junio, Eric and J6t
* Further cleanup of t1507 before inlining full_name() (since we didn't want to inline full_name() into a bad command substitution)
[1]: (may be rebased at any time) https://github.com/Denton-L/git/tree/ready/cleanup-test-must-fail2
Denton Liu (15):
t/lib-git-p4: use test_path_is_missing()
t0000: replace test_must_fail with run_sub_test_lib_test_err()
t0003: use named parameters in attr_check()
t0003: use test_must_be_empty()
t0003: don't use test_must_fail attr_check
t0020: don't use test_must_fail has_cr
t0020: use ! check_packed_refs_marked
t1306: convert test_might_fail rm to rm -f
t1307: reorder nongit test_must_fail
t1409: let sed open its own input file
t1409: use test_path_is_missing()
t1501: remove use of test_might_fail cp
t1507: stop losing return codes of git commands
t1507: run commands within test_expect_success
t1507: inline full_name()
t/lib-git-p4.sh | 2 +- t/t0000-basic.sh | 14 ++--- t/t0003-attributes.sh | 47 +++++++-------- t/t0020-crlf.sh | 18 +++--- t/t1306-xdg-files.sh | 8 +-- t/t1307-config-blob.sh | 2 +- t/t1409-avoid-packing-refs.sh | 16 +++--- t/t1501-work-tree.sh | 2 +- t/t1507-rev-parse-upstream.sh | 104 +++++++++++++++++++--------------- 9 files changed, 112 insertions(+), 101 deletions(-)
Range-diff against v2:
1: 85cee92765 < -: ---------- test-lib-functions: introduce test_non_git_might_fail() 2: 3d64837af1 = 1: cf3dd04b8a t/lib-git-p4: use test_path_is_missing() 3: 778ae9052b = 2: 51a2226726 t0000: replace test_must_fail with run_sub_test_lib_test_err() 4: dbc82d45c6 = 3: 9374fcd8db t0003: use named parameters in attr_check() 5: e06a06cff5 = 4: 7f8808a850 t0003: use test_must_be_empty() 6: 219011f983 = 5: b725d53dbe t0003: don't use `test_must_fail attr_check` 7: 8da6c96b39 = 6: f6ef6d245c t0020: don't use `test_must_fail has_cr` 8: 27550eaae6 = 7: fc32b8d584 t0020: use ! check_packed_refs_marked 9: c19f6344a4 = 8: 7e29b0154e t1306: convert `test_might_fail rm` to `rm -f` 10: d6ea8a6df0 = 9: cf43579d65 t1307: reorder `nongit test_must_fail` 11: d57dfe95e2 = 10: e2fe278bfa t1409: let sed open its own input file 12: eacf4e0fb4 = 11: 1eef3f4bc5 t1409: use test_path_is_missing() 13: 83e47748bc ! 12: fddd224225 t1501: remove use of `test_might_fail cp`
@@ Commit message
The test_must_fail() family of functions (including test_might_fail()) should only be used on git commands. Replace test_might_fail() with
- test_non_git_might_fail().
+ a compound command wrapping the old cp invocation that always returns 0.
The test_might_fail cp line was introduced in 466e8d5d66 (t1501: fix test with split index, 2015-03-24). It is necessary because there might exist some index files in repo.git/sharedindex.* and, if they exist, we want to copy them over. However, if they don't exist, we don't want to error out because we expect that possibility. As a result, we want to
- keep the "might fail" semantics so we use test_non_git_might_fail().
+ keep the "might fail" semantics so we always return 0, even if the
+ underlying cp errors out.
## t/t1501-work-tree.sh ##
@@ t/t1501-work-tree.sh: test_expect_success 'Multi-worktree setup' '
@@ t/t1501-work-tree.sh: test_expect_success 'Multi-worktree setup' '
mkdir -p repo.git/repos/foo &&
cp repo.git/HEAD repo.git/index repo.git/repos/foo &&
- test_might_fail cp repo.git/sharedindex.* repo.git/repos/foo &&
-+ test_non_git_might_fail cp repo.git/sharedindex.* repo.git/repos/foo &&
++ { cp repo.git/sharedindex.* repo.git/repos/foo || :; } &&
sane_unset GIT_DIR GIT_CONFIG GIT_WORK_TREE
'
14: 9e20865f94 = 13: 63ca18207d t1507: stop losing return codes of git commands 15: 7c61ac6b67 = 14: 44a410d57a t1507: run commands within test_expect_success 16: d09370455f = 15: 4fe445279b t1507: inline full_name()
Signature · 1 line
2.24.1.703.g2f499f1283