Uniform the AST to find more duplicate mutations

[zimmski]

The AST should be cleaned from all comments, empty lines and odd formatting (i.e. usage of ";") before taking a checksum. This would lead a uniform coding style and could lead to finding more duplications. At least for the code in "example" it would find some.

This should be also done with the content of the original file, so that we can identify if a mutation produces no changes.