I have been working on a small virtual machine called Carp. Right now I am working on writing my own tokenizer (as opposed to relying upon strtok()) and lexer, and then also a bytecode reader/writer.

http://github.com/tekknolagi/carp