[SOLVED] Parsing S-Expressions in Python

MTK358 · 02-01-2011, 08:13 PM

I've been loosely following this:

http://norvig.com/lispy.html

And I have a problem: the parsing function throws an array out of bounds exception. I thought that maybe I'm doing it wrong, so I copy and pasted the code from the page, and still the same error!

I'm using Python 3 (not 2), if it matters.

bgeddy · 02-01-2011, 10:12 PM

I've looked at the page you link to - it's quite interesting. I had a look at the lis.py module and noticed it's definitely python2 and not python3 compliant having print statements and calling the now removed operator.div method of the operator module. I fixed the print statements running the 2to3 utility on the code and replaced the operator.div with operator.truediv calls manually. This meant I could now import it and run repl() to get the prompt. I also tested some examples of program(), tokenize() and parse() which worked too. This example, "(count (quote the) (quote (the more the merrier the bigger the better)))", of using the prompt to evaluate statements failed though so I left it at that. The example works however in python2.

This is undoubtedly a python3 incompatibility and I suspect your problems are the same. Possibly try your tests in python2 ?

MTK358 · 02-02-2011, 08:00 AM

Turns out it was a dumb mistake. I wrote this:

Code:

return s.replace('(',' ( ').replace('(',' ( ').split()

instead of this:

Code:

return s.replace('(',' ( ').replace(')',' ) ').split()

in the tokenizing function.