On 11/04/2013 06:52 AM, Philippe Sigaud wrote:

On Mon, Nov 4, 2013 at 1:55 AM, Timothee Cour <thelastmammoth@gmail.com> wrote:

I guess I'll have to write a CT-compatible LALR(1) engine...
 
D does not fit into LALR(1), you need glr.

Well, too bad. GLR is also on my plate, but I fear its cubic behavior (IIRC, it degrades gracefully, though).
Thats one part, and even worse is that you need a decider function if more than one rule accepts.
Do you know what part of the D grammar makes it non-LALR(1)?
I had big trouble with the IdentifierList rules.