It’s been a good weekend!  I’ve made a ton of progress implementing the new description language.

FYI, the compiler is split into 4 layers: the Parser; the Model; the Plan; and the Code Generator.  At this point, the Parser and the vast majority of the Model have been rewritten to support the new description language.  Changes to the Plan will be mostly refactoring, to link up the old code to the new Model.  The Code Generator needs to be updated with the error recovery system and support for the new AST features.  That said, I think the Parser and Model rewrite comprise the majority of the work, which is now finished.

In particular, the “+” and “*” operators both work for terms and phrases in rules (a totally new feature), and slot naming works reliably in both singular and plural contexts.  Macro support is also working, as is the new unicode support in the lexer. 

I may still have a little bit of work to do in the interpreter — to ensure the output matches the input grammar — but the new model keeps enough information from the AST that, even without that work, the interpreter output should still make sense.

So, there you have it.  I might even have the interpreter back on its feet this week, time allowing.  Yay!