From: "Elijah Newren via GitGitGadget" <gitgitgadget@gmail.com>
To: git@vger.kernel.org
Cc: Derrick Stolee <stolee@gmail.com>,
Elijah Newren <newren@gmail.com>,
Elijah Newren <newren@gmail.com>
Subject: [PATCH v2 0/3] Fix bugs from interesting renaming pairs: one side renames A/file -> B/file, the other B/ -> A/
Date: Wed, 30 Jun 2021 17:29:57 +0000 [thread overview]
Message-ID: <pull.1039.v2.git.git.1625074200.gitgitgadget@gmail.com> (raw)
In-Reply-To: <pull.1039.git.git.1624727121.gitgitgadget@gmail.com>
Anders Kaseorg recently reported a few issues in an interesting rename
case[1]. I was able to duplicate and find multiple bugs from it; two in
merge-recursive, and one in merge-ort. This series has some fixes.
Changes since v1:
* Added a third testcase
[1]
https://lore.kernel.org/git/CABPp-BGDfucqae=HNES_QmmsjpDbdHrR6CG=H3gtiDygHzquVg@mail.gmail.com/
Elijah Newren (3):
t6423: test directory renames causing rename-to-self
merge-ort: ensure we consult df_conflict and path_conflicts
merge-recursive: handle rename-to-self case
merge-ort.c | 6 +-
merge-recursive.c | 19 ++-
t/t6423-merge-rename-directories.sh | 175 ++++++++++++++++++++++++++++
3 files changed, 193 insertions(+), 7 deletions(-)
base-commit: 670b81a890388c60b7032a4f5b879f2ece8c4558
Published-As: https://github.com/gitgitgadget/git/releases/tag/pr-git-1039%2Fnewren%2Frename-plus-dir-rename-cancel-v2
Fetch-It-Via: git fetch https://github.com/gitgitgadget/git pr-git-1039/newren/rename-plus-dir-rename-cancel-v2
Pull-Request: https://github.com/git/git/pull/1039
Range-diff vs v1:
1: d3572e8bc85 ! 1: 86b2eaea75e t6423: test directory renames causing rename-to-self
@@ t/t6423-merge-rename-directories.sh: test_expect_failure '12h: renaming a file w
+ test_cmp expect actual
+ )
+'
++
++# Testcase 12k, Directory rename with sibling causes rename-to-self
++# Commit O: dirB/foo, dirA/{bar, baz_1}
++# Commit A: dirA/{foo, bar, baz_1}
++# Commit B: dirB/{foo, bar}, dirA/baz_2
++# Expected: dirA/{foo, bar, baz_2}, with conflicts on dirA/bar vs. dirB/bar
++
++test_setup_12k () {
++ test_create_repo 12k &&
++ (
++ cd 12k &&
++
++ mkdir dirA dirB &&
++ echo foo >dirB/foo &&
++ echo bar >dirA/bar &&
++ echo baz >dirA/baz &&
++ git add . &&
++ git commit -m orig &&
++
++ git branch O &&
++ git branch A &&
++ git branch B &&
++
++ git switch A &&
++ git mv dirB/* dirA/ &&
++ git commit -m A &&
++
++ git switch B &&
++ git mv dirA/bar dirB/bar &&
++ echo more baz >>dirA/baz &&
++ git commit -m B
++ )
++}
++
++test_expect_merge_algorithm failure failure '12k: Directory rename with sibling causes rename-to-self' '
++ test_setup_12k &&
++ (
++ cd 12k &&
++
++ git checkout A^0 &&
++
++ test_must_fail git -c merge.directoryRenames=conflict merge -s recursive B^0 &&
++
++ test_path_is_missing dirB &&
++ test_path_is_file dirA/bar &&
++ test_path_is_file dirA/baz &&
++
++ git ls-files | uniq >tracked &&
++ test_line_count = 3 tracked &&
++
++ git status --porcelain -uno >actual &&
++ cat >expect <<-\EOF &&
++ UU dirA/bar
++ M dirA/baz
++ EOF
++ test_cmp expect actual
++ )
++'
+
###########################################################################
# SECTION 13: Checking informational and conflict messages
2: 052f40c3c1a ! 2: 6e7ef18bf53 merge-ort: ensure we consult df_conflict and path_conflicts
@@ t/t6423-merge-rename-directories.sh: test_setup_12j () {
test_setup_12j &&
(
cd 12j &&
+@@ t/t6423-merge-rename-directories.sh: test_setup_12k () {
+ )
+ }
+
+-test_expect_merge_algorithm failure failure '12k: Directory rename with sibling causes rename-to-self' '
++test_expect_merge_algorithm failure success '12k: Directory rename with sibling causes rename-to-self' '
+ test_setup_12k &&
+ (
+ cd 12k &&
3: dea97c25e52 ! 3: 7c96d6c7a0a merge-recursive: handle rename-to-self case
@@ t/t6423-merge-rename-directories.sh: test_setup_12j () {
test_setup_12j &&
(
cd 12j &&
+@@ t/t6423-merge-rename-directories.sh: test_setup_12k () {
+ )
+ }
+
+-test_expect_merge_algorithm failure success '12k: Directory rename with sibling causes rename-to-self' '
++test_expect_success '12k: Directory rename with sibling causes rename-to-self' '
+ test_setup_12k &&
+ (
+ cd 12k &&
--
gitgitgadget
next prev parent reply other threads:[~2021-06-30 17:30 UTC|newest]
Thread overview: 12+ messages / expand[flat|nested] mbox.gz Atom feed top
2021-06-26 17:05 [PATCH 0/3] Fix bugs from interesting renaming pairs: one side renames A/file -> B/file, the other B/ -> A/ Elijah Newren via GitGitGadget
2021-06-26 17:05 ` [PATCH 1/3] t6423: test directory renames causing rename-to-self Elijah Newren via GitGitGadget
2021-06-29 12:50 ` Derrick Stolee
2021-06-30 16:33 ` Elijah Newren
2021-06-26 17:05 ` [PATCH 2/3] merge-ort: ensure we consult df_conflict and path_conflicts Elijah Newren via GitGitGadget
2021-06-26 17:05 ` [PATCH 3/3] merge-recursive: handle rename-to-self case Elijah Newren via GitGitGadget
2021-06-29 4:02 ` Junio C Hamano
2021-06-29 12:55 ` Derrick Stolee
2021-06-30 17:29 ` Elijah Newren via GitGitGadget [this message]
2021-06-30 17:29 ` [PATCH v2 1/3] t6423: test directory renames causing rename-to-self Elijah Newren via GitGitGadget
2021-06-30 17:29 ` [PATCH v2 2/3] merge-ort: ensure we consult df_conflict and path_conflicts Elijah Newren via GitGitGadget
2021-06-30 17:30 ` [PATCH v2 3/3] merge-recursive: handle rename-to-self case Elijah Newren via GitGitGadget
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=pull.1039.v2.git.git.1625074200.gitgitgadget@gmail.com \
--to=gitgitgadget@gmail.com \
--cc=git@vger.kernel.org \
--cc=newren@gmail.com \
--cc=stolee@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).