From: <no...@so...> - 2002-02-08 21:19:32
|
Bugs item #514181, was opened at 2002-02-07 01:57 You can respond by visiting: http://sourceforge.net/tracker/?func=detail&atid=426093&aid=514181&group_id=39838 Category: Import/Install Difficulty Group: v1.0 (example) Status: Closed Resolution: Fixed Priority: 5 Submitted By: Jay Bonci (jaybonci) Assigned to: chromatic (chromatic) Summary: A Nodetype can extend itself Initial Comment: When editing a nodetype, if you happen to have a typo, or for whatever reason you are very silly, you can set a nodetype to "extend" itself. This send the request into what seems like an infinite loop. I waited a bit before killing apache to reload it. A simple check should be enough to fix it, but still, I thought I'd mention it here, as I don't exactly know the fix / checking procedure for ecore. Thanks. ---------------------------------------------------------------------- >Comment By: Nate Oostendorp (n_oostendorp) Date: 2002-02-08 13:19 Message: Logged In: YES user_id=297846 heh, re: tests -- yeah, i need to get off my ass and actually install everything on a new machine and try 'em out. ---------------------------------------------------------------------- Comment By: chromatic (chromatic) Date: 2002-02-08 13:01 Message: Logged In: YES user_id=2274 Patch really uploaded this time, but the fix has already been checked into CVS. You'll just have to trust me on the test case, until you rule one way or another on including them. ---------------------------------------------------------------------- Comment By: Nate Oostendorp (n_oostendorp) Date: 2002-02-08 12:53 Message: Logged In: YES user_id=297846 Hey guys. chromatic: yeah, fixing the top level insert function should do the trick. I don't see your patch anywhere though... You can email it to me, but you should have cvs access -- if that's still a problem let me know. ---------------------------------------------------------------------- Comment By: chromatic (chromatic) Date: 2002-02-08 11:58 Message: Logged In: YES user_id=2274 Easily enough fixed in insert(). I added a check to see if the type_nodetype and extends_nodetype fields match. If so, the node defaults to extending 'node'. This may not be the best approach, but I don't think this should be a failure, especially as we don't yet have a good way to handle logical constraints. I also added a test in the appropriate suite. Patch attached with my suggested fix. nate, any comments? ---------------------------------------------------------------------- You can respond by visiting: http://sourceforge.net/tracker/?func=detail&atid=426093&aid=514181&group_id=39838 |