[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: svn commit: r922511 - /subversion/trunk/subversion/libsvn_client/commit.c

From: Philip Martin <philip.martin_at_wandisco.com>
Date: Thu, 18 Mar 2010 22:30:27 +0000

Greg Stein <gstein_at_gmail.com> writes:

>> @@ -1040,13 +1046,10 @@ check_nonrecursive_dir_delete(void *bato
>>   */
>>   if (btn->depth != svn_depth_infinity)
>>     {
>> -      svn_wc_status2_t *status;
>> -      svn_node_kind_t kind;
>> -
>> -      SVN_ERR(svn_io_check_path(target_abspath, &kind, pool));
>
> ... looks at the disk.
>
> Is this change proper and safe? What happens if a file/dir is
> obstructed by something of a different kind? Before, it would see what
> was *actually* there, not just what it thought *should* be there.

This code detects attempts to make a non-recursive commit of a deleted
directory with deleted children. In the past it would skip the check
if the directory was obstructed, now it makes the check. The new
behaviour seems more logical to me, but it makes little difference in
practice since the non-recursive commit of an obstructed directory
will fail.

-- 
Philip
Received on 2010-03-18 23:30:59 CET

This is an archived mail posted to the Subversion Dev mailing list.