Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!uunet!munnari.oz.au!metro!usage.csd.unsw.oz.au!ccadfa!ghm
From: ghm@ccadfa.adfa.oz.au (Geoff Miller)
Newsgroups: comp.lang.fortran
Subject: Re: More Style Points -- One Entry, One Exit?
Message-ID: <1814@ccadfa.adfa.oz.au>
Date: 14 Aug 90 23:35:12 GMT
References: <1990Aug4.011300.17395@nmt.edu>
Organization: Computer Centre, University College, UNSW, ADFA, Canberra, Australia
Lines: 32

nraoaoc@nmt.edu (Daniel Briggs) writes:

>As long as we are discussing style points, here's one I have wondered
>about.  Why are people SO adamant about the ONE entry, ONE exit point
>of view?  

Maintainability.  When you have had to maintain large chunks of unstructured
FORTRAN in which each (long) subroutine has multiple entry and exit points,
you become slightly paranoid on the subject.  Also, as I pointed out in 
another connection, it is better to have one simple rule to follow than to 
have rules which depend on the surroundings.

>  ...examples and discussion deleted...

>....  So, style gurus, what are your opinions?  Do
>you stick religiously to one entry, one exit, or do you bend the rules
>a bit when it makes for clearer code?

Any Style Guide is an ideal.  How rigidly you apply it (or have your staff 
apply it) is up to you, and I would be the first to admit that in some cases 
the very nature of FORTRAN can make stylistically pure code appear ugly.
It comes back always to maintainability, and there will always be some room
for judgement.

Actually, the only point where I bend the one-entry-one-exit rule is in the
case of an emergency stop  -  for example, where a program is designed to 
maintain two consistent data files, and yet it detects an inconsistency.  In
this case I stop the program with an informative (to me :-) message, because
I have obviously got something drastically wrong somewhere.

Geoff Miller  (ghm@cc.adfa.oz.au)
Computer Centre, Australian Defence Force Academy