HELP with ideas for analyzing 250KSLOC of Fortran (IFTRAN)

Greetings Experts.


We are about to take on a large Fortran software maintenance effort. The
software is written in "IFTRAN" which allows for the use of DO WHILE, CASE,
CASEOF, etc. The entire system is embedded and the Fortran is called by
an executive hosted by a Rolm MSE-14 microprocessor.


I was hoping someone has had some experiences similar to mine and could
make some recommendations regarding analysis of the code. I am looking for
any public-domain scanners, parsers or code analyzers that could be
modified to understand IFTRAN/Fortran-77. There exists no graphical
description of the architetcure but it is very well commented. My
feeling is that I at least need to produce structure charts, DAGs, or
graphs that will help our programmers understand the relationships.


I have little experience with scanners/parsers but have a keen need to
come up to speed as required. The software development environment is
a Data General MV-10K with AOS/VS. I would like to analyze the code on
a Sun or PC.


Could f2c be easily modified to understand the additional IFTRAN
constructs - or would it be easier to start from scratch? I know of Tcl/Tk
and have thought that might be an easy to set up a GUI environment.


I have even thought that it could be beneficial to setup some kind of
hyper-information server (WWW/?) that would allow users to search the code
by keyword.




I greatly appreciate any suggestions you may have. I am looking at CASE
tools but would first like to explore what we can do ourselves.




Buried in code,
Thanks much,
Ronan McAllister
--

Return to the comp.compilers page.
Search the comp.compilers archives again.