threads / patch / 28396

patcht6019: avoid refname collision on case-insensitive systems

Subject: [PATCH] t6019: avoid refname collision on case-insensitive systems

## tl;dr

5 messages between Sep 15, 2011 and Sep 15, 2011. Diffs are folded; open one to read it.

replies: 4people: 4as markdown or json

Thomas Rast· Sep 15, 2011, 08:34 UTC · lore

The criss-cross tests kept failing for me because of collisions of 'a' with 'A' etc. Prefix the lowercase refnames with an extra letter to disambiguate.

Signed-off-by: Thomas Rast <trast@student.ethz.ch>
---

This fixes the tests on OS X. Together with Peff's fix to the poll issue, it now tests clean again.

 t/t6019-rev-list-ancestry-path.sh |   19 +++++++++++--------
 1 files changed, 11 insertions(+), 8 deletions(-)
Show changes to t/t6019-rev-list-ancestry-path.sh +11 −8
diff --git a/t/t6019-rev-list-ancestry-path.sh b/t/t6019-rev-list-ancestry-path.sh
index 738af73..39b4cb0 100755
--- a/t/t6019-rev-list-ancestry-path.sh
+++ b/t/t6019-rev-list-ancestry-path.sh
@@ -75,33 +75,36 @@ test_expect_success 'rev-list --ancestry-patch D..M -- M.t' '
 # a   X
 #  \ / \
 #   c---cb
+#
+# All refnames prefixed with 'x' to avoid confusion with the tags
+# generated by test_commit on case-insensitive systems.
 test_expect_success 'setup criss-cross' '
 	mkdir criss-cross &&
 	(cd criss-cross &&
 	 git init &&
 	 test_commit A &&
-	 git checkout -b b master &&
+	 git checkout -b xb master &&
 	 test_commit B &&
-	 git checkout -b c master &&
+	 git checkout -b xc master &&
 	 test_commit C &&
-	 git checkout -b bc b -- &&
-	 git merge c &&
-	 git checkout -b cb c -- &&
-	 git merge b &&
+	 git checkout -b xbc xb -- &&
+	 git merge xc &&
+	 git checkout -b xcb xc -- &&
+	 git merge xb &&
 	 git checkout master)
 '
 
 # no commits in bc descend from cb
 test_expect_success 'criss-cross: rev-list --ancestry-path cb..bc' '
 	(cd criss-cross &&
-	 git rev-list --ancestry-path cb..bc > actual &&
+	 git rev-list --ancestry-path xcb..xbc > actual &&
 	 test -z "$(cat actual)")
 '
 
 # no commits in repository descend from cb
 test_expect_success 'criss-cross: rev-list --ancestry-path --all ^cb' '
 	(cd criss-cross &&
-	 git rev-list --ancestry-path --all ^cb > actual &&
+	 git rev-list --ancestry-path --all ^xcb > actual &&
 	 test -z "$(cat actual)")
 '
 
-- 
1.7.7.rc1.366.ge210a6
Brad King· Sep 15, 2011, 12:45 UTC · re: Thomas Rast · lore

Re: [PATCH] t6019: avoid refname collision on case-insensitive systems

On 9/15/2011 4:34 AM, Thomas Rast wrote:
> The criss-cross tests kept failing for me because of collisions of 'a'
> with 'A' etc.  Prefix the lowercase refnames with an extra letter to
> disambiguate.
Thanks.
-Brad
Junio C Hamano· Sep 15, 2011, 15:52 UTC · re: Thomas Rast · lore

Re: [PATCH] t6019: avoid refname collision on case-insensitive systems

Thomas Rast <trast@student.ethz.ch> writes:
Show 9 quoted lines
> The criss-cross tests kept failing for me because of collisions of 'a'
> with 'A' etc.  Prefix the lowercase refnames with an extra letter to
> disambiguate.
>
> Signed-off-by: Thomas Rast <trast@student.ethz.ch>
> ---
>
> This fixes the tests on OS X.  Together with Peff's fix to the poll
> issue, it now tests clean again.

Hmm, I must have misplaced Peff's poll fix. Care to point me in the right direction?

Jeff King· Sep 15, 2011, 15:53 UTC · re: Junio C Hamano · lore

Re: [PATCH] t6019: avoid refname collision on case-insensitive systems

On Thu, Sep 15, 2011 at 08:52:01AM -0700, Junio C Hamano wrote:
Show 5 quoted lines
> > This fixes the tests on OS X.  Together with Peff's fix to the poll
> > issue, it now tests clean again.
> 
> Hmm, I must have misplaced Peff's poll fix. Care to point me in the right
> direction?

You have it already; it's part of the new http-auth commits I sent yesterday (patch 2/5).

-Peff
Junio C Hamano· Sep 15, 2011, 16:45 UTC · re: Jeff King · lore

Re: [PATCH] t6019: avoid refname collision on case-insensitive systems

Jeff King <peff@peff.net> writes:
Show 10 quoted lines
> On Thu, Sep 15, 2011 at 08:52:01AM -0700, Junio C Hamano wrote:
>
>> > This fixes the tests on OS X.  Together with Peff's fix to the poll
>> > issue, it now tests clean again.
>> 
>> Hmm, I must have misplaced Peff's poll fix. Care to point me in the right
>> direction?
>
> You have it already; it's part of the new http-auth commits I sent
> yesterday (patch 2/5).
Ah, that one. Thanks, yes I do have it.

I somehow thought Thomas was talking about a test fix for 5800 "intermittent hang".

← back to recent threads