Hmm. Could this use base_get_info() instead?
On Fri, Jan 15, 2010 at 12:10, <hwri...@apache.org> wrote: > Author: hwright > Date: Fri Jan 15 17:10:10 2010 > New Revision: 899717 > > URL: http://svn.apache.org/viewvc?rev=899717&view=rev > Log: > * subversion/libsvn_wc/workqueue.c > (run_deletion_postcommit): Replace a read-only entry use with > svn_wc__db_read_info(). > > Modified: > subversion/trunk/subversion/libsvn_wc/workqueue.c > > Modified: subversion/trunk/subversion/libsvn_wc/workqueue.c > URL: > http://svn.apache.org/viewvc/subversion/trunk/subversion/libsvn_wc/workqueue.c?rev=899717&r1=899716&r2=899717&view=diff > ============================================================================== > --- subversion/trunk/subversion/libsvn_wc/workqueue.c (original) > +++ subversion/trunk/subversion/libsvn_wc/workqueue.c Fri Jan 15 17:10:10 2010 > @@ -895,7 +895,7 @@ > const char *repos_relpath; > const char *repos_root_url; > const char *repos_uuid; > - const svn_wc_entry_t *parent_entry; > + svn_revnum_t parent_revision; > > /* If we are suppose to delete "this dir", drop a 'killme' file > into my own administrative dir as a signal for svn_wc__run_log() > @@ -942,12 +942,14 @@ > FALSE, FALSE, cancel_func, cancel_baton, scratch_pool)); > > /* If the parent entry's working rev 'lags' behind new_rev... */ > - SVN_ERR(svn_wc__get_entry(&parent_entry, db, > - svn_dirent_dirname(local_abspath, > - scratch_pool), > - FALSE, svn_node_dir, FALSE, > - scratch_pool, scratch_pool)); > - if (new_revision > parent_entry->revision) > + SVN_ERR(svn_wc__db_read_info(NULL, NULL, &parent_revision, NULL, NULL, > + NULL, NULL, NULL, NULL, NULL, NULL, NULL, > + NULL, NULL, NULL, NULL, NULL, NULL, NULL, > + NULL, NULL, NULL, NULL, NULL, > + db, svn_dirent_dirname(local_abspath, > + scratch_pool), > + scratch_pool, scratch_pool)); > + if (new_revision > parent_revision) > { > /* ...then the parent's revision is now officially a > lie; therefore, it must remember the file as being > > >
