[PATCH] stash drops the stash even if creating the branch fails because it already exists

This bug was disovered by someone on IRC when he tried to 'git stash branch <branch> <stash>'
while <branch> already existed. In that case the stash is dropped even though it isn't
applied on any branch, so the stash is effectively lost. I think that shouldn't happen,
so here is a test.

---
 t/t3903-stash.sh |   11 +++++++++++
 1 files changed, 11 insertions(+), 0 deletions(-)

diff --git a/t/t3903-stash.sh b/t/t3903-stash.sh
index 9ed2396..0f6b2e4 100755
--- a/t/t3903-stash.sh
+++ b/t/t3903-stash.sh

@@ -545,4 +545,15 @@ test_expect_success 'invalid ref of the form stash@{n}, n >= N' '
 	git stash drop
 '
 
+test_expect_failure 'stash branch should not drop the stash if the branch exists' '
+	git stash clear &&
+	echo foo > file &&
+	git add file &&
+	git commit -m initial &&
+	echo bar > file &&
+	git stash &&
+	test_must_fail git stash branch master stash@{0} &&
+	git rev-parse stash@{0} --
+'
+
 test_done

-- 
1.7.3.3.gd2416

This series fixes another (sorry!) regression I introduced with my detached-stash series.

I have included Tomas' test, so that my tweak to the test script applies cleanly.

Jon Seymour (1):
  stash: fix git stash branch regression when branch creation fails

Tomas Carnecky (1):
  stash drops the stash even if creating the branch fails because it
    already exists

 git-stash.sh     |    6 +++---
 t/t3903-stash.sh |   11 +++++++++++
 2 files changed, 14 insertions(+), 3 deletions(-)

-- 
1.7.3.4.g787b.dirty

From: Tomas Carnecky <redacted>

This bug was disovered by someone on IRC when he tried to 'git stash branch <branch> <stash>'
while <branch> already existed. In that case the stash is dropped even though it isn't
applied on any branch, so the stash is effectively lost. I think that shouldn't happen,
so here is a test.
---
 t/t3903-stash.sh |   11 +++++++++++
 1 files changed, 11 insertions(+), 0 deletions(-)

diff --git a/t/t3903-stash.sh b/t/t3903-stash.sh
index 9ed2396..0f6b2e4 100755
--- a/t/t3903-stash.sh
+++ b/t/t3903-stash.sh

@@ -545,4 +545,15 @@ test_expect_success 'invalid ref of the form stash@{n}, n >= N' '
 	git stash drop
 '
 
+test_expect_failure 'stash branch should not drop the stash if the branch exists' '
+	git stash clear &&
+	echo foo > file &&
+	git add file &&
+	git commit -m initial &&
+	echo bar > file &&
+	git stash &&
+	test_must_fail git stash branch master stash@{0} &&
+	git rev-parse stash@{0} --
+'
+
 test_done

-- 
1.7.3.4.g787b.dirty

Tomas Carnecky reported a regression in the behaviour
of git stash branch, when the branch creation fails.

This patch fixes that regression by restoring the
pre-condition for dropping that previously existed.

This patch assumes Tomas Carnecky's patch has already been applied.

Signed-off-by: Jon Seymour <redacted>
---
 git-stash.sh     |    6 +++---
 t/t3903-stash.sh |    2 +-
 2 files changed, 4 insertions(+), 4 deletions(-)

diff --git a/git-stash.sh b/git-stash.sh
index 23a9ab5..5fb1245 100755
--- a/git-stash.sh
+++ b/git-stash.sh

@@ -432,9 +432,9 @@ apply_to_branch () {
 	assert_stash_like "$@"
 
 	git checkout -b $branch $REV^ &&
-	apply_stash "$@"
-
-	test -z "$IS_STASH_REF" || drop_stash "$@"
+	apply_stash "$@" && {
+		test -z "$IS_STASH_REF" || drop_stash "$@"
+	}
 }
 
 PARSE_CACHE='--not-parsed'

diff --git a/t/t3903-stash.sh b/t/t3903-stash.sh
index 0f6b2e4..336e244 100755
--- a/t/t3903-stash.sh
+++ b/t/t3903-stash.sh

@@ -545,7 +545,7 @@ test_expect_success 'invalid ref of the form stash@{n}, n >= N' '
 	git stash drop
 '
 
-test_expect_failure 'stash branch should not drop the stash if the branch exists' '
+test_expect_success 'stash branch should not drop the stash if the branch exists' '
 	git stash clear &&
 	echo foo > file &&
 	git add file &&

-- 
1.7.3.4.g787b.dirty

On 9/28/10 3:19 PM, Jon Seymour wrote:

From: Tomas Carnecky <redacted>

This bug was disovered by someone on IRC when he tried to 'git stash branch <branch> <stash>'
while <branch> already existed. In that case the stash is dropped even though it isn't
applied on any branch, so the stash is effectively lost. I think that shouldn't happen,
so here is a test.

This line was missing from my original patch, sorry about that:
Signed-off-by: Tomas Carnecky <redacted>

Junio,

Let me know if you want me to re-roll this series with Tomas'
sign-off. I can also add my simplification of the Brian's fix and
another fix I have made to have git stash save/create fail early in
case the index contains merge conflicts.

jon.

On Tue, Sep 28, 2010 at 11:21 PM, Tomas Carnecky [off-list ref] wrote:

On 9/28/10 3:19 PM, Jon Seymour wrote:

quoted

From: Tomas Carnecky <redacted>

This bug was disovered by someone on IRC when he tried to 'git stash branch <branch> <stash>'
while <branch> already existed. In that case the stash is dropped even though it isn't
applied on any branch, so the stash is effectively lost. I think that shouldn't happen,
so here is a test.

This line was missing from my original patch, sorry about that:
Signed-off-by: Tomas Carnecky <redacted>

quoted

---
 t/t3903-stash.sh |   11 +++++++++++
 1 files changed, 11 insertions(+), 0 deletions(-)

diff --git a/t/t3903-stash.sh b/t/t3903-stash.sh
index 9ed2396..0f6b2e4 100755
--- a/t/t3903-stash.sh
+++ b/t/t3903-stash.sh

@@ -545,4 +545,15 @@ test_expect_success 'invalid ref of the form stash@{n}, n >= N' '

      git stash drop
 '

+test_expect_failure 'stash branch should not drop the stash if the branch exists' '
+     git stash clear &&
+     echo foo > file &&
+     git add file &&
+     git commit -m initial &&
+     echo bar > file &&
+     git stash &&
+     test_must_fail git stash branch master stash@{0} &&
+     git rev-parse stash@{0} --
+'
+
 test_done

--
To unsubscribe from this list: send the line "unsubscribe git" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html