torsten.anders at beds.ac.uk
Tue Oct 30 23:27:54 CET 2012
> the new parser is based on a PEG/Packrat parser written purely in Oz. The grammar and its associated semantic actions are written in Oz. The Packrat engine will be available at the developer level, so that you can use it for your DSL's.
Sounds like this is all I need :) Rewriting existing grammars should be fine.
Unfortunately, ozh (the javadoc complement) is then also gone, because it depends on Gump. But once we have a full grammar of the Oz language itself then it would in principle be possible to port ozh.
BTW: as the Oz parser itself is defined in Oz, does this mean we already have a full grammar of the Oz language itself? Would it also be possible to define extensions of the Oz syntax by extending this grammar?
Dr Torsten Anders
Course Leader, Music Technology
University of Bedfordshire
Park Square, Room A315
On 30 Oct 2012, at 21:09, Sébastien Doeraene wrote:
> On Tue, Oct 30, 2012 at 3:01 PM, Torsten Anders <torsten.anders at beds.ac.uk> wrote:
> On 30 Oct 2012, at 11:26, alexgian wrote:
> >> The next step is the distribution subsystem. Mozart 2.0 has a reflective language interface, where you can write Oz code to change the semantics of language operations.
> > Brilliant. I got nothing against Gump, but that whole C/yacc route was a bit convoluted for my tastes.
> Actually, Gump is more intended for defining new languages from scratch, not for changing the semantics of Oz itself. (I hope either Gump or something similar will still be available in the medium/long term for Mozart 2 :)
> We plan to drop Gump altogether. It was too much of a hacky thing. At most we'll have a side-project for legacy code. But the new parser is based on a PEG/Packrat parser written purely in Oz. The grammar and its associated semantic actions are written in Oz. The Packrat engine will be available at the developer level, so that you can use it for your DSL's.
> mozart-users mailing list mozart-users at mozart-oz.org
More information about the mozart-users