The Common Lisp stuff now works passably well for prototyping, though it's nowhere *near* as mature as the Scheme implementation. Until someone creates a *real* Common Lisp implementation, we at least have a prototype that'll work for some purposes.
Below are some results after running some tests to see if the Common Lisp stuff *could* be used (it can). --- David A. Wheeler To test out the Common Lisp stuff, I snagged the Common Lisp source code from: http://norvig.com/paip/README.html and ran this: ./diff-s-sweet -C -w paip/*.lisp This program's "-w" option is primarily to help you find s-expressions that aren't well-formatted, but it also helps to test "unsweeten"'s Common Lisp implementation. Unsurprisingly, nearly all Common Lisp code (at least, in this sample) is also well-formed. In the guile test suite, *EVERY* file in the test suite was well-formed. In the Common Lisp set, I did find a few exceptions, though they were rare: * In a few places it uses "f(x)" to mean the two datums "f (x)". This is in a section where it tries to process lists that are math expressions. Of course, in sweet-expressions f(x) is (f x). This turns out to be rare, unsurprisingly, and easily fixed in the few places they occur by inserting a space between "f" and "(x)". The whole point of that code is to try to simulate a notation that in my opinion is better handled with sweet-expressions... so in fact that whole section of code could disappear. * In a few places there are multiple datums on one line, esp. in grammar.lisp where there are a number of declarations about words (multiple declarations are put on a line). Inserting 1+ spaces in front of the line fixes this completely, and it actually doesn't happen very often. I also found a few weaknesses in its handling of Common Lisp, which *could* be fixed: 1. The sequence ".," is currently interpreted as a symbol (as does guile!), which isn't correct for Common Lisp: (defun make-clause (clause) "Translate a message from define-class into a case clause." `(,(first clause) #'(lambda ,(second clause) .,(rest2 clause)))) That could be fixed in our reader, but for the moment, it doesn't seem worth the trouble. 2. The |...| is consumed by the reader, and not printed back out later, so |.| is mis-interpreted later as the set-cdr operator in "examples.lisp" section "(defexamples 7 ...". It's a rough count, but there are 11738 lines in this sample code set (per wc -l paip/*.lisp); note that is *not* pretty-printed. The # of lines of code where there were problems, was 197 (per ./diff-s-sweet -C -w paip/*.lisp | grep '^- ' | wc -l), which is 1.7%. Generally the problem was that one construct, say ".,", would cause an entire datum to fail, so it's not as bad as it looks. It's basically a very small set of lines that have to be changed before nearly all of them can be read as-is. ------------------------------------------------------------------------------ Minimize network downtime and maximize team effectiveness. Reduce network management and security costs.Learn how to hire the most talented Cisco Certified professionals. Visit the Employer Resources Portal http://www.cisco.com/web/learning/employer_resources/index.html _______________________________________________ Readable-discuss mailing list Readable-discuss@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/readable-discuss
