Have a look at this http://www.expasy.ch/tools/peptidecutter/

There should be a plenty of such scripts. BTW, you can look at function "cleave" (line 764) in this source

http://code.google.com/p/nsilk/source/browse/trunk/xllib/sequence.py?r=143

Nsilk is a python tool to spot protein-protein interactions, but that function will predict peptides from a sequence. You only have to be inspired as it uses some classes and objects that you won't have.

d

This is pretty simple to do with python, no fancy scripts are required.

>>> protein = 'MLTFKFYLPKQATELKDLQCLEDELGPLRHVLDLTESKSFQLEDAENFIS'
>>> 
>>> print protein.replace('K','-').replace('R','-').split('-')
['MLTF', 'FYLP', 'QATEL', 'DLQCLEDELGPL', 'HVLDLTES', 'SFQLEDAENFIS']

You can plug this in anywhere in your program.

You could make the code a bit more dynamic I suppose by using a dictionary to hold the enzyme:recognition site combinations and building the command that way, rather than hardwiring it.

I'd start with a tool designed for this purpose. One option is digest, from the EMBOSS suite of programs.

It should not be too difficult to integrate this tool into a Python script, either by calling it and parsing the output, or running the EMBOSS tool independently then parsing. I believe that BioPython has some options for working with EMBOSS via the Bio.Emboss.Applications module.