Tokenize fails when the '.' character appears inside the code

[dpaz]

When the character '.' is used inside the code, for example, a string, the tokenize function, broke the code in wrong parts.

The standard erlang library tokenizer works with characters as separators only. So probably one possible solution can be split that works with regexp.

Another way to work with this can be using erlang preprocesor, with this way we probably solve at the same time #4