I started off on a project to write a simple little program (ha!) that would turn files written in a LaTeX-like (but much, much simpler) language (called Raskal) into a fully-formed web site. All of the "hard" stuff - i.e. the actual HTML, CSS, document structures, etc. will be defined in files elsewhere, and then glued together (and filled in with content) according to what's in a bunch of files written in Raskal. All I really have to do is read the Raskal files and then parse them and turn them into data trees and hand them off to another program (already written) which will take care of turning the data tree into actual HTML and putting it where it goes on the server.
Someone said that I'm writing a compiler and therefore need to learn all about context-free grammars, LALR(1) parsers, deterministic finite state automata, and all sorts of other sexy-sounding but ultimately painful stuff. Are they right or did they just want to give me headaches? I am aware that things like YACC exist but I'm wary of using such magic without understanding it first. Thanks, and (since no-one else is going to say it) welcome back to me. Dan /* PLUG: http://plug.org, #utah on irc.freenode.net Unsubscribe: http://plug.org/mailman/options/plug Don't fear the penguin. */
