Formatter questions

I have a few questions about Mizar with regards to the formatter that I'm
working on:

1) Is there a comprehensive list of the meanings of all the intermediate
files that are generated for an article? I have some vague ideas like "parx"
seems to be related to tokenization, but really I'd like a list of what each
of these is for.

2) Related to 1), is there an intermediate file that is an abstract syntax
tree, or is it possible to have Mizar generate and abstract syntax tree for
an article? Ideally for formatting you'd just like an AST annotated with
identifiers.

3) Is the grammar posted at
http://mizar.uwb.edu.pl/version/current/doc/syntax.txt up-to-date?