13 May
2001
13 May
'01
7:40 a.m.
Thomas Johnsson <johnsson@crt.se> wrote,
Happy and others like it generate an LR parser, which is a well-established technology since the late 60's (Knuth): efficient, deterministic, and checks the grammar for you. Parser combinators are usually nondeterministic ie backtracking (pre-Knuth!:-) though Cleverly Disguised in Haskell Higher Order clothes.... [..] PS would be cool to try to marry the two approaches....
Doaitse Swierstra et al have done this to a certain extent: http://www.cs.uu.nl/groups/ST/Software/UU_Parsing/ Parser combinators that are efficient, deterministic, and handle errors much better than the classic form of parser combinators. Manuel