Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!zaphod.mps.ohio-state.edu!think.com!linus!linus!gong!lamour
From: lamour@gong.mitre.org (Michael Lamoureux)
Newsgroups: comp.lang.perl
Subject: Re: pattern matching question
Message-ID: <1991May24.133427.4721@linus.mitre.org>
Date: 24 May 91 13:34:27 GMT
References: <1991May22.193037.12166@cherokee.uswest.com> <1991May22.221507.12660@cherokee.uswest.com> <49420@ut-emx.uucp> <sherman.675057031@foster>
Sender: news@linus.mitre.org (News Service)
Reply-To: lamour@mitre.org
Organization: The MITRE Corporation, McLean, Va
Lines: 61
Nntp-Posting-Host: gong.mitre.org

In article <sherman.675057031@foster>, sherman@unx.sas.com (Chris Sherman) writes:
|> In <49420@ut-emx.uucp> dboles@ccwf.cc.utexas.edu (David Boles) writes:
|> 
|> >Warning: PERL NOVICE approaching !!!
|> 
|>    Warning:  PERL NOVICE answering!!!

	Ditto.  (But I am avidly reading the book...)

|> >I am using:
|> >while (<>) {
|> >    s/(p\d*) (p\d*) (\d*) (\d*)/$1 -$3 -$4\n$2 $3 $4\n/;
|> >    print;
|> >}
|> >Why aren't $3 and $4 "alive" in the first half of the replacement
|> >string?  What am I missing?
|> 
|> I think I got it.  Perl is taking your test string literally, space
|> for space.

	This is exactly it.

|> #!/usr/local/bin/perl
|> while (<>) {
|>   s/(p\d*) *(p\d*) *(\d*) *(\d*)/$1 -$3 -$4\n$2 $3 $4\n/;
|>   print;
|> }
|> 
|> Maybe perl pro's can tell me what the '*'s meant exactly, why
|> they are working, and if they would work in every case.
|> I was hoping to set up a one-or-more-number-of
|> spaces type thing, but I don't think I did that right).

	Well, your expression tests for 0 or more spaces.  An "*" tests
for 0 or more occurences, a "+" tests for 1 or more.  So using "+"
instead of "*" would fix that, but I think using a "\s" instead of a
" " would be more multi-purpose.  This matches any whitespace, not just
a space.  So I guess it should look like this:

while (<>) {
  s/(p\d+)\s+(p\d+)\s+(\d+)\s+(\d+)/$1 -$3 -$4\n$2 $3 $4\n/;
  print;
}

	This matches the p's only if they have numbers appended as well.
So even better...

while (<>) {
  if (/.*(p\d+)\s+(p\d+)\s+(\d+)\s+(\d+).*/) {
    print "$1 -$3 -$4\n$2 $3 $4\n";
  }
}

	This allows you to put comments or something else in the file
and only prints out the strings which match (and drops typos...you may
want to change this expression a bit and flag errors with an else).  Note
that "." matches any character.

Michael
lamour@mitre.org
Disclaimer:  Perl is addictive ;-)