Lambda the Ultimate

inactiveTopic Funk
started 10/25/2000; 2:10:07 AM - last post 10/25/2000; 1:38:50 PM
andrew cooke - Funk  blueArrow
10/25/2000; 2:10:07 AM (reads: 498, responses: 3)
The summary here might make more sense than the main site.

Seems to involve rewriting strings with backtracking? Confused me. And what's a URM?
Posted to "" by andrew cooke on 10/25/00; 2:23:28 AM

Chris Rathman - Re: Funk  blueArrow
10/25/2000; 8:10:30 AM (reads: 525, responses: 0)
And what's a URM?
URM = Universal Register Machine

A URM is a Turing Machine which is a more friendly to progamming - i.e. a bit more advanced than the programming language used in a pure Turing Machine.

Seems to involve rewriting strings with backtracking?
As useful as the Regular Expression syntax is, I find it to be rather suboptimal in terms of readibility and maintainibility. May not be the author's intention, but I think the main advantage to use using a backtracking approach on strings is as an alternative to the unix RE engine.

Then again, I'm a little biased in that I think Icon is the best string engine ever invented for solving the problem of searching (and backtracking) through patterns within a string.

Ehud Lamm - Re: Funk  blueArrow
10/25/2000; 1:15:34 PM (reads: 505, responses: 1)
Icon better in this respect than SNOBOL?


Chris Rathman - Re: Funk  blueArrow
10/25/2000; 1:38:50 PM (reads: 553, responses: 0)
Icon better in this respect than SNOBOL? Why?
Considering that Icon and SNOBOL were developed by many of the same people, the languages probably have much in common with respect to text processing. Unfortunately, I've never worked with SNOBOL and don't know much about the language, other than the references I've seen scattered in the Icon literature. I have heard that SNOBOL is really showing it's age these days (line numbers, etc...).

As for Icon, text processing is much easier to deal with than regular expressions for the simple reason that the language natively supports generators, backtracking, and a success-failure mechanism. It's the only approach to string processing that I've seen that comes close to the power afforded by RE's.