mkent@dewey.soe.berkeley.edu (Marty Kent) (12/11/90)
I'm about to take on the task of making sense out of a large and entirely undocumented and uncommented body of common lisp code - yes thank you for your sympathies - and I'm looking for some tools and perhaps methodological suggestions. I've undertaken similar tasks in the distant past and remember feeling vastly aided by MASTERSCOPE, which as you may recall was a lexical analysis tool in Interlisp which examined source code and compiled a database about the call structure and variable usage of the routines. Being able to look at graphs of "who calls who" proved very helpful in making sense of the complex set of relationships among many unknown functions, and the natural-languagey interface was really a joy to use. The first time I sat down with the thing I wanted to change the name of some global, and without knowing any of the actual MASTERSCOPE commands and with no manual, just interacting with the program, I was able to come up with the satisfactory line (something like "show where any uses global-name", a very natural formulation) on the second try! Very effective interface, I thought... Well let me not wax overly nostalgic here. It strikes me as plausible there might be some kind of comparable tool available now, written in common lisp and prepared to analyse common lisp code - something in source code form, waiting to be ported and compiled on the local system - I seem to recall a long-ago sense of lots of lisp code available at various sites in the public domain, many peoples' labors of love or at least infatuation - lots of my own stuff out there as well at one time or another- and I wonder if that's still the case. I'd appreciate getting information on the availability of lex analysis tools, or common lisp software tools in general, so if you have any suggestions for me in this regard, please drop a line. And if you have any tips or anecdotes about eureka experiences in coming to understand (whatever that means) the structure of large lisp systems from undocumented source code, I'd really like to hear about that too. I also have some interest in commercial products, but I'm more interested in public domain software - for several reasons... Thanks for your time and input (and for the implicit corroboration of my social-image existence...)
mkent@dewey.soe.berkeley.edu (Marty Kent) (12/11/90)
Sorry but apparently my mailer neglected to append my signature to my
request for info on CommonLisp lexical analysis tools (subject heading
CommonLisp Software Tools Availability?) My address is as follows:
Marty Kent
Sixth Sense Research and Development
415/548 9129
MKent@dewey.soe.berkeley.edu
{uwvax, decvax, inhp4}!ucbvax!mkent%dewey.soe.berkeley.edu
Kent's Law of Human Organizations: Nothing Big Works