[RFC/RFA] Introduce new struct parse_context

[Joel Brobecker]

> > If we were to put some documentation in gdbint for this patch,
> > what would you put?
> 
> Both.  This is a significant new feature, and at least its main
> principles should be described in gdbint, enough to get the reader off
> the ground with the knowledge she could use to find the parts of code
> which implement this feature (and read the comments there ;-).

I don't see what you would describe in the documentation that would
not be in the code as a comment. I think I don't see what to say because
I don't see this patch as a significant new feature, but as a modification
whose purpose is to fix a bug (the parser uses the global current_language)
to do the parsing, but this global gets overwritten as a side effect of
a function called between the moment we set the current_language and
the moment we actually do the parsing).

How about the following compromise: I describe again roughly what the
problem is and how I am going to address it. Then you can decide what
to add to gdbint, and where. I would appreciate your help in writing
that addition, because I just simply suck in writing documentations,
and as a result, I'm really slow.  Just the time it takes me to write
good comments makes me mad, sometimes.

-- 
Joel