I made a Python Detokenizer for Axe. However, it only works on some files. The source can be found here.

However, it only seems to work for some files, here's the output for some programs:

http://paste.bwns.be/p/f7a31ac22 Doesn't work
By the way, what it prints first is the file in hexadecimal.

http://paste.bwns.be/p/f4125158 Works

http://paste.bwns.be/p/f59f46d85 Doesn't work

http://paste.bwns.be/p/f68869d0f Works

I really don't know what makes it work for some programs, but for others. I think it's the file structure, but all the programs are unlocked and unarchived.

Any idea of what I'm doing wrong here? Thanks

Do all of the programs have a mix of one- and two-byte tokens? Can you identify any patterns in terms of program size?

Kerm, thank you a lot for the answer. I did figure out the error though and fixed it. I'll make a topic on this later.

Very good, I look forward to hearing what the answer was. I know the program is not completely relevant to the members of Cemetech, but I hope that you will indeed upload it to the Cemetech archives! I should also mention that I didn't get a chance to look at your code to help debug; I was planning on doing that if you were unable to find the problem.