You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Bert Huijben <be...@qqmail.nl> on 2010/04/29 10:55:49 UTC
RE: svn commit: r939150 - /subversion/trunk/subversion/libsvn_client/merge.c
> -----Original Message-----
> From: neels@apache.org [mailto:neels@apache.org]
> Sent: donderdag 29 april 2010 2:49
> To: commits@subversion.apache.org
> Subject: svn commit: r939150 -
> /subversion/trunk/subversion/libsvn_client/merge.c
>
> Author: neels
> Date: Thu Apr 29 00:49:16 2010
> New Revision: 939150
>
> URL: http://svn.apache.org/viewvc?rev=939150&view=rev
> Log:
> Another entry_t down.
>
> * subversion/libsvn_client/merge.c
> (obstructed_or_missing):
> Use wc-ng API instead of svn_wc__get_entry_versioned(), and much
> simplify
> this function in the process. Untangle, rinse and dry.
> (node_kind_working, node_kind_on_disk):
> Remove these ghastly two functions and incorporate into
> obstructed_or_missing(), the way it should always have been.
This patch breaks
FAIL: merge_authz_tests.py 1: skipped paths get overriding mergeinfo
(which is automatically skipped on ra_local)
Most likely your new check doesn't work correctly for 'absent' nodes.
I would recommend also checking the behavior for excluded node, as our test suite doesn't test those as often as it should.
(Nice effect is that you can check excluded on ra_local, but can't check absent)
Bert
Re: svn commit: r939150 - /subversion/trunk/subversion/libsvn_client/merge.c
Posted by Neels J Hofmeyr <ne...@elego.de>.
Philip Martin wrote:
> Neels J Hofmeyr <ne...@elego.de> writes:
>
>>> (Nice effect is that you can check excluded on ra_local, but can't
>>> check absent)
>> /me doesn't fully understand and wonders if he should.
>
> Absent can only occur with auth and auth does not apply to ra_local.
Heh, something I always knew but never realized. Thanks!
~Neels
Re: svn commit: r939150 - /subversion/trunk/subversion/libsvn_client/merge.c
Posted by Philip Martin <ph...@wandisco.com>.
Neels J Hofmeyr <ne...@elego.de> writes:
>> (Nice effect is that you can check excluded on ra_local, but can't
>> check absent)
> /me doesn't fully understand and wonders if he should.
Absent can only occur with auth and auth does not apply to ra_local.
--
Philip
Re: svn commit: r939150 - /subversion/trunk/subversion/libsvn_client/merge.c
Posted by Neels J Hofmeyr <ne...@elego.de>.
Bert Huijben wrote:
>
>> -----Original Message-----
>> From: neels@apache.org [mailto:neels@apache.org]
>> Sent: donderdag 29 april 2010 2:49
>> To: commits@subversion.apache.org
>> Subject: svn commit: r939150 -
>> /subversion/trunk/subversion/libsvn_client/merge.c
>>
>> Author: neels
>> Date: Thu Apr 29 00:49:16 2010
>> New Revision: 939150
>>
>> URL: http://svn.apache.org/viewvc?rev=939150&view=rev
>> Log:
>> Another entry_t down.
>>
>> * subversion/libsvn_client/merge.c
>> (obstructed_or_missing):
>> Use wc-ng API instead of svn_wc__get_entry_versioned(), and much
>> simplify
>> this function in the process. Untangle, rinse and dry.
>> (node_kind_working, node_kind_on_disk):
>> Remove these ghastly two functions and incorporate into
>> obstructed_or_missing(), the way it should always have been.
>
> This patch breaks
> FAIL: merge_authz_tests.py 1: skipped paths get overriding mergeinfo
weird...
> (which is automatically skipped on ra_local)
...ah! Darn it!
Thanks Bert, I'll try to resolve.
> Most likely your new check doesn't work correctly for 'absent' nodes.
thx
> I would recommend also checking the behavior for excluded node, as our test suite doesn't test those as often as it should.
and thx
> (Nice effect is that you can check excluded on ra_local, but can't check absent)
/me doesn't fully understand and wonders if he should.
~Neels