Community
Participate
Working Groups
Xtend needs to have a serializer running in negative time and error-free in all cases. I'm looking at you Mr. Eysholdt :-)
I'm on it!
Sounds very promising!
This scares me. Will it run as perfect as my Zip algo which crunches all files down to max 100 bytes (OK, Unzip algo still pending)?
For sure you can safely delete your zip algo, since Moritz' serializer will also take care of your archives. @Moritz: Shouldn't be a problem, is it? I mean the new architecture is quite flexible.
as long as you can describe your archive's contents with an Xtext grammar, everything is possible...
I've enabled the new serializer for xbase and xtend2: SerializerFragment: is used in the workflows for Xbase, Xtype and Xtend2 Interesting files are: org.eclipse.xtext.xbase.serializer.XbaseSemanticSequencer org.eclipse.xtext.xtend2.serializer.Xtend2SemanticSequencer org.eclipse.xtext.xbase.tests.serializer.XbaseSerializerTest org.eclipse.xtext.xtend2.tests.serializer.SerializerTest the generic sequencer can handle Xbase, except for: - determine what kind of operation a BinaryOperation is. - determine if an XClosure is a XShortClosure. to handle these cases I've adapted the XbaseSemanticSequencer. Thanks to AbstractXbaseEvaluationTest, the test for Xbase is pretty extensive. For Xtend2, however, I've got only one test case so far.
Woohoo! That sounds pretty good. So we need tests for the different Xtend fetures, such as RichStrings, dispatch and create extensions. We could use the AbstractXbaseEvaluationTest for Xtend too. It might not be that interesting but you ever know and it is relatively easy to do. A first shot on the formatter would also be great :-)
Preliminary scheduled for 2.0RC1
please open new bugzillas if concrete problems occur.
Closing all bugs that were set to RESOLVED before Neon.0