[plt-scheme] Good compiler texts?

From: Todd O'Bryan (toddobryan at gmail.com)
Date: Thu Dec 27 17:17:02 EST 2007

On Dec 27, 2007 11:45 AM, Matthias Felleisen <matthias at ccs.neu.edu> wrote:
> One person's front-end is another person's middle end.
> Todd, if Shriram's conjecture is correct and you are interested
> in lexing and parsing (and as you say the 'theory'), I recommend
> working through a text on 'automata' (as in finite-state, push-down,
> Turing and friends). You get the theory and you get in a manner
> that is related to the mathematics you could do in 12th grade.
> If I am correct, and your phrase really meant the ideas of, say,
> context-sensitive checking (e.g., variables, types) and the
> assignment of meaning, then PLAI is the right text for you.
Shriram had the advantage of discussions at my TeachScheme! workshop
to remind him that I'm a linguist by training. There's also  that
sixth-sense about knowing what people want even when they don't that
he mentioned. :-) So, points to Shriram--I am interested in the lexing
and parsing part.

I've actually had a Theory of Computation class, so I'm pretty
comfortable with those ideas and I've actually written top-down and
bottom-up parsers, in Lisp, believe it or not. What I'm really
interested in is, for example, how to deal with ambiguous grammars,
either by converting them to unambiguous grammars or by using a parser
that can handle ambiguous grammars. I'd also like some more
information on the data structures that get used to store parse states

My guess is that my weird combination of knowledge puts me in the
unenviable position of having to go to the primary literature to get
what I want. On the one hand that's exciting. On the other hand, I
wanted someone else to have distilled the big ideas.


Posted on the users mailing list.