Path: utzoo!utgpu!news-server.csri.toronto.edu!rpi!zaphod.mps.ohio-state.edu!wuarchive!uunet!interet!iris
From: iris@interet.UUCP (User)
Newsgroups: comp.lang.fortran
Subject: Re: FORTRAN restructuring tools
Message-ID: <16@interet.UUCP>
Date: 7 Jun 91 14:13:37 GMT
References: <7341@s3.ireq.hydro.qc.ca>
Reply-To: iris@interet.UUCP (User)
Distribution: comp.lang.fortran
Organization: Interet, Maplewood, NJ
Lines: 30

In article <7341@s3.ireq.hydro.qc.ca> langlo@ireq.hydro.qc.ca (Luc Langlois 8159) writes:
>I'm looking for tools to analyse and restructure existing FORTRAN 77 code. 
>Has anyone had any experience with any of the above ? 

Of the list you mention, we have two: FORWARN from  Quibus  and  MAT  from
SAIC.   We also acquired Fortran-lint from IPT (415-494-7500) about a year
ago.

Of the three, I find that Fortran-lint works the best.   FORWARN  is  fast
but  often incomplete.  MAT is slow and cumbersome and often TOO complete:
it generates reams of messages.  Fortran-lint seems to be just right: it's
fast and very convenient to use (unlike the other two, I don't have to run
my source code through any filters or combine my 1000  source  files  into
one).   It does complete interface checking even with the routines not  in
calltree  order  and  is able to detect an uninitialized variable which is
referenced several subroutine calls deep.  The messages can be individual-
ly  suppressed  and  the  output  is  clean  and  understandable.   I find
Fortran-lint to be an indispensable tool for managing a large FORTRAN  ap-
plication.

============
Iris Engelson
Director of Software Development
Interet
111 Dunnell Road 
Maplewood, NJ 07040
Tel: 201-763-1200
Fax: 201-763-5120
iris@interet