Re: LR(k)

[John Cowan <cowan@LOCKE.CCIL.ORG>]

la paulos. cusku di'e

> This is a grammar problem, not a language one. Knuth's theorem states
> that *all* deterministic context-free languages have an LR(1) grammar.

I'm not familiar with the theorem.  Surely not all languages are LALR(1),
though, which is what standard Yaccs implement.  I looked into using an
LR(2) parser at one time, but it only ran on OS/2, which I no longer run,
and is proprietary.  (It was downward compatible with Yacc, but called
a routine yylex2() when necessary.)

--
John Cowan                                      cowan@ccil.org
                e'osai ko sarji la lojban.