Community
Participate
Working Groups
The current implementation of the org.eclipse.xtext.mwe.Reader is based on the index. The SlotEntry traverses all exported objects of a resource and filters it by name and type. Elements which don't have a name are not exported by default. In some situations, the root element does not have a name, but should be the input to the code generator. We should provide a index independent version of the SlotEntry that is able to deliver unnamed/unexported objects.
Created attachment 170527 [details] An alternative implementation of SlotEnty + tests The attached patch introduces a new class ResourceSetBasedSlotEntry, that uses the resource set's contents instead of its indexing information.
The patch looks good to me.
Patch applied with two slight changes: - removed TODO comment - made logger static final
Closing bug which were set to RESOLVED before Eclipse Neon.0.