svn commit: r1028123 - /subversion/trunk/subversion/libsvn_client/patch.c

[st...@apache.org]

Author: stsp
Date: Wed Oct 27 22:05:26 2010
New Revision: 1028123

URL: http://svn.apache.org/viewvc?rev=1028123&view=rev
Log:
* subversion/libsvn_client/patch.c
  (resolve_target_path): Remove question in comment, the answer to which is
   that the path "." (current directory) is canonicalised to an empty path.
   So it's OK for the parser to return an empty path if the patch changes
   properties on the current directory.

Modified:
    subversion/trunk/subversion/libsvn_client/patch.c

Modified: subversion/trunk/subversion/libsvn_client/patch.c
URL: http://svn.apache.org/viewvc/subversion/trunk/subversion/libsvn_client/patch.c?rev=1028123&r1=1028122&r2=1028123&view=diff
==============================================================================
--- subversion/trunk/subversion/libsvn_client/patch.c (original)
+++ subversion/trunk/subversion/libsvn_client/patch.c Wed Oct 27 22:05:26 2010
@@ -346,9 +346,7 @@ resolve_target_path(patch_target_t *targ
   target->canon_path_from_patchfile = svn_dirent_internal_style(
                                         path_from_patchfile, result_pool);
 
-  /* We allow properties to be set on the wc root dir.
-   * ### Do we need to check for empty paths here, shouldn't the parser
-   * ### guarentee that the paths returned are non-empty? */
+  /* We allow properties to be set on the wc root dir. */
   if (! prop_changes_only && target->canon_path_from_patchfile[0] == '\0')
     {
       /* An empty patch target path? What gives? Skip this. */