Python for Reverse Engineering

[Brad Tilley]

A friend of mine wrote an algorithm that generates strings. He says that 
it's impossible to figure out exactly how the algorithm works. That may 
be true, but I think if I had enough sample strings that I could write a 
program to identify patterns in the strings and then produce strings 
with similar patterns. He disagrees with me. He gave me 100 strings to 
analyze.

I wrote a script to read each string and build a list of characters 
present. I've discovered that he only uses 20 alpha/numeric chars to 
generate the strings and that each string sums up to a value between 
1000 and 1200 when I assign each char in the string its ASCII value.

What else might I do to analyze these things? As it stands now, I can 
generate an acceptable string on ever 100th attempt or so, but I'd like 
to do better than this.

-b