Hello, I've recently checked in modified template code for the GLR mode of Happy. It handles densely ambiguous grammars a lot better, although may be slower on more reasonable grammars, particularly if empty productions are heavily used (eg for gap rules in biological sequence grammars). For example, it can parse "E -> E + E | i" on sentences with 201 tokens in a small number of seconds; the previous version could not handle more than 41 in reasonable time. You can get the code from CVS, or download the files via the following: http://www.dur.ac.uk/p.c.callaghan/happy-glr/new-templates.tgz http://www.dur.ac.uk/p.c.callaghan/happy-glr/new-templates.zip There's still plenty of room for further speed improvements. Contact me if interested. Further information can be found at: http://www.dur.ac.uk/p.c.callaghan/happy-glr Paul -------------------------------------------------------------------------- Dr Paul Callaghan (Lecturer) Email: P.C.Callaghan@durham.ac.uk Dept of Computer Science, Durham http://www.dur.ac.uk/p.c.callaghan