Some Eclipse Foundation services are deprecated, or will be soon. Please ensure you've read this important communication.

Bug 345587

Summary:	Acceleo is not working with some deployed EPackages
Product:	[Modeling] Acceleo	Reporter:	Cedric Brun <cedric.brun>
Component:	Core	Assignee:	Project Inbox <acceleo-inbox>
Status:	CLOSED FIXED	QA Contact:
Severity:	blocker
Priority:	P3	CC:	stephane.begaudeau
Version:	unspecified
Target Milestone:	---
Hardware:	PC
OS:	Linux
Whiteboard:

Description Cedric Brun

2011-05-12 07:55:56 EDT

I'm using an Acceleo generator and launching it as a plugin.

When I reference a given Ecore model which is deployed and installed in my IDE, I'm using the 'http://' URI but Acceleo compiles it in ../../../ in the .emtl file.

This very same metamodel is not appearing in the "Ecore Editor load resource" action if you check "runtime". I can't say why.

I'm blocked as I can't use the standalone launch as I have platform:/ URIs in my model and I can't use the plugin launch neither.

Any hint of what's wrong with my Ecore model ?

Comment 1 Stephane Begaudeau

2011-05-16 11:50:48 EDT

I couldn't reproduce the problem since any ecore files with its code generated thanks to EMF is normally register as a runtime metamodel when deployed as a plugin. And with runtime metamodel Acceleo save the uri of the metamodel with the uri written in the metamodel (http://wwww....) not a relative path to the ecore file.

I could reproduce your "corruption" problem that you've told me about. The problem was fixed previously by Laurent but since we changed the way we resolve the uri of a metamodel in the workspace we were looking for a physical uri while we are now using the logical uri.

The fix has been contributed on HEAD, it will be available in Acceleo 3.1.0 RC1 tomorrow.

Comment 2 Stephane Begaudeau

2011-05-16 11:54:38 EDT

Now that I'm thinking about it, the corrution problem was the caused of the disappearance of your metamodel from the runtime view :D
I've removed my fixed and tried, and it was indeed the problem.
Well it should be fixed now anyway.

Comment 3 Laurent Goubet

2011-08-18 04:22:03 EDT

Closing our "resolved" and "verified" bugs.