Path: utzoo!utgpu!news-server.csri.toronto.edu!rutgers!sun-barr!cs.utexas.edu!uunet!rbj
From: rbj@uunet.UU.NET (Root Boy Jim)
Newsgroups: comp.lang.perl
Subject: Re: Dcmp - recursive directory comparison
Message-ID: <125391@uunet.UU.NET>
Date: 12 Mar 91 23:11:16 GMT
References: <lgy.668812268@newton>
Organization: UUNET Communications Services, Falls Church, VA
Lines: 30

<lgy.668812268@newton> lgy@phys.washington.edu (Laurence G. Yaffe) writes:
>
>    I frequently use a recursive directory comparison program to compare
>the contents of entire directory trees.  I recently decided to rewrite
>my previous shell based 'dcmp' program in Perl (and of course, add various
>bells and whistles).  Appended below is the result (my first serious Perl
>programming).  If others find it useful, great.  Comments on Perl programming
>style welcome.

Funny how things tend to get invented at about the same time.
I am planning to post my version soon. It does have one
other feature, in that it removes identical versions
from one half of the tree.

>    By default, this program reports on files present in one directory tree,
>but absent in the other, and files present in both trees but with differing
>contents, ownerships or permissions.  'Diff' may run on differing text files
>to display differences.  Its output is designed to be terse but informative.
>(I find the sysV "dircmp" unpleasantly verbose and insufficiently flexible.)
>It's wrapped with its own man page (but not shar'ed).

So get GNU diff instead. It does recursive comparison nicer.

I will look at your version more thoroughly soon. However, one
thing SCREAMS for attention. DON'T FOLLOW SYMBOLIC LINKS BY DEFAULT!
This is a dangerous thing to do in any program that deals
with the filesystem as a recursive structure.
-- 
		[rbj@uunet 1] stty sane
		unknown mode: sane