You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by John Szakmeister <jo...@szakmeister.net> on 2004/12/19 01:07:25 UTC

Re: svn commit: r12386 - in branches/1.1.x: . subversion/libsvn_client subversion/libsvn_ra_dav subversion/mod_dav_svn

On Saturday 18 December 2004 18:05, jerenkrantz@tigris.org wrote:
[snip] 
=======================================================================
>======= --- branches/1.1.x/subversion/libsvn_client/ra.c	(original)
> +++ branches/1.1.x/subversion/libsvn_client/ra.c	Sat Dec 18 17:05:48
> 2004 @@ -170,11 +170,30 @@
>  {
>    svn_client__callback_baton_t *cb = baton;
>    svn_wc_adm_access_t *adm_access;
> +  const svn_wc_entry_t *entry;
>    const char *full_path = svn_path_join (cb->base_dir, path, pool);
>
> -  SVN_ERR (svn_wc_adm_probe_retrieve (&adm_access, cb->base_access,
> -                                      full_path, pool));
> -  return svn_wc_prop_set (name, value, full_path, adm_access, pool);
> +  SVN_ERR (svn_wc_entry (&entry, full_path, cb->base_access, FALSE,
> pool)); +  if (! entry)
> +    return svn_error_createf (SVN_ERR_UNVERSIONED_RESOURCE, NULL,
> +                              _("'%s' is not under version control"),
> +                              svn_path_local_style (full_path, pool));
> +
> +  SVN_ERR (svn_wc_adm_retrieve (&adm_access, cb->base_access,
> +                                (entry->kind == svn_node_dir
> +                                 ? full_path
> +                                 : svn_path_dirname (full_path,
> pool)), +                                pool));
> +
> +  /* We pass 1 for the 'force' parameter here.  Since the property is
> +     coming from the repository, we definitely want to accept it.
> +     Ideally, we'd raise a conflict if, say, the received property is
> +     svn:eol-style yet the file has a locally added svn:mime-type
> +     claiming that it's binary.  Probably the repository is still
> +     right, but the conflict would remind the user to make sure.
> +     Unfortunately, we don't have a clean mechanism for doing that
> +     here, so we just set the property and hope for the best. */
> +  return svn_wc_prop_set2 (name, value, full_path, adm_access, TRUE,
> pool); }

This commit broke the 1.1.x branch.  In particular, svn_wc_prop_set2() 
doesn't exist in 1.1.x, as indicated by the recent failures. :-)

-John

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: svn commit: r12386 - in branches/1.1.x: . subversion/libsvn_client subversion/libsvn_ra_dav subversion/mod_dav_svn

Posted by John Szakmeister <jo...@szakmeister.net>.
On Saturday 18 December 2004 21:02, Justin Erenkrantz wrote:
> --On Saturday, December 18, 2004 8:07 PM -0500 John Szakmeister
>
> <jo...@szakmeister.net> wrote:
> > This commit broke the 1.1.x branch.  In particular,
> > svn_wc_prop_set2() doesn't exist in 1.1.x, as indicated by the recent
> > failures. :-)
>
> That last line calling svn_wc_prop_set2() shouldn't have been there.
> libsvn_diff got overly greedy in the merge and I didn't catch it.
>
> Sorry.  I'll merge it back.  -- justin

No problem.  I just wanted to point it out the problem so it could be 
fixed. :-)

I am a bit concerned that merge inserted a line that wasn't present in the 
original diff though.  We should *never* do that.  I wonder where it's 
breaking down?

-John

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org

Re: svn commit: r12386 - in branches/1.1.x: . subversion/libsvn_client subversion/libsvn_ra_dav subversion/mod_dav_svn

Posted by Justin Erenkrantz <ju...@erenkrantz.com>.
--On Saturday, December 18, 2004 8:07 PM -0500 John Szakmeister 
<jo...@szakmeister.net> wrote:

> This commit broke the 1.1.x branch.  In particular, svn_wc_prop_set2()
> doesn't exist in 1.1.x, as indicated by the recent failures. :-)

That last line calling svn_wc_prop_set2() shouldn't have been there. 
libsvn_diff got overly greedy in the merge and I didn't catch it.

Sorry.  I'll merge it back.  -- justin

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org