Question about organization of gfortran source code

[Tobias Burnus]

Hi,

L R wrote:
> I am a high school student working on a project involving fortran.
> Since I am focusing more on the optimization side of a compiler, I do
> not want to spend ages writing a parser that will work for all of
> fortran's weird gotcha's.
It is a bit unclear what kind of optimizations you are thinking of.
Essentially all Fortran parts are done after the TREE is generated,
which is used by the middle end to do optimizations. (Though some
optimizations are done already before). To see how the information
generated for the middle-end looks like, use the option
-fdump-tree-original (and note that the dumped file does not contain all
information).

Without knowing more about your project, it is hard to guess what kind
of parser information you need. A simple parse information is given for
-fdump-parse-tree (which is useful for debugging but not necessarily for
post processing).

Tobias