[antlr-interest] rules for re-factoring grammars?

[Benjamin Shropshire]

Does anyone known of a good wright up on the rules that can be used to 
re-factor productions? I'm thinking in particular of rules that can 
remove left recursion including with ambiguous grammars.

A :
    F |
    A B |
    D |
    A C |
    E;

D, E and/or F might map over an A.

I've Googled for it but haven't come up with much useful (I might be 
looking for the wrong terms or something) and I don't recall seeing 
anything near what I'm looking for in the antlr docs (but this is 
re-much more a language theory question than an Antlr question).