[PATCH 2 of 2] tests: fixup tests after 551aa6e27929

[Adrian Buehlmann]

On 31.10.2010 00:46, Adrian Buehlmann wrote:
> (adding Martin to cc)
> 
> On 31.10.2010 00:05, steve at borho.org wrote:
>> # HG changeset patch
>> # User Steve Borho <steve at borho.org>
>> # Date 1288475969 18000
>> # Branch stable
>> # Node ID 2637232ad33d829b475d9d530e055c7af97f698a
>> # Parent  1ef0e920cf260f62517730fec156f71a1a9ee1db
>> tests: fixup tests after 551aa6e27929
>>
>> diff -r 1ef0e920cf26 -r 2637232ad33d tests/test-mq-qnew.t
>> --- a/tests/test-mq-qnew.t	Sat Oct 30 16:58:15 2010 -0500
>> +++ b/tests/test-mq-qnew.t	Sat Oct 30 16:59:29 2010 -0500
>> @@ -107,7 +107,7 @@
>>    abort: "foo#bar" cannot be used as the name of a patch
>>    abort: "foo:bar" cannot be used as the name of a patch
>>    % qnew with name containing slash
>> -  abort: cannot write patch "foo/": Is a directory
>> +  abort: cannot write patch "foo/": No such file or directory
>>    abort: "foo" already exists as a directory
>>    foo/bar.patch
>>    popping foo/bar.patch
>> @@ -172,7 +172,7 @@
>>    abort: "foo#bar" cannot be used as the name of a patch
>>    abort: "foo:bar" cannot be used as the name of a patch
>>    % qnew with name containing slash
>> -  abort: cannot write patch "foo/": Is a directory
>> +  abort: cannot write patch "foo/": No such file or directory
>>    abort: "foo" already exists as a directory
>>    foo/bar.patch
>>    popping foo/bar.patch
> 
> Perhaps there is a platform dependent variation in that error message...
> 
> On my Ubuntu with Python 2.6.4, the test passes as is.
> 
> Judging from the mailheader, you ran the test on Mac OS?

Another question: if you apply my recent patch

  "opener: raise IOError on paths ending in '/'"

does the test still fail on your machine?