ITNW 2310 - PERL
Allan Kochis,Adjunct Professor - CIT

Regular Expressions

  1. Matching
    We have been saying / /which is just shorthand for m/ / the match operator. When we omit the variable Perl uses $_ but we can check for a match(=~) or not a match(!~) in any variable.
    PerlOutput 
    #!/usr/bin/perl -w

$string="one two";

if($string =~ /one/) {
     print "Matched\n";
}

if($string !~ /three/) {
  print "True I did not find three\n";
}


    Matched
True I did not find three

  2. Memory and backreference variables.
    Whenever a regex sees () it saves the contents of the match in a memory location, First $1, second match $2 and so on.
    PerlOutput 
    #!/usr/bin/perl -w

$string="one two";

$_=$string;

m/(\w+)\W+(\w+)/;

print $1."\n";
print $2."\n";



    one
two

  3. Multiple matches
    In order to get more than the first match use the g flag. This will return an array reference.
    PerlOutput 
    #!/usr/bin/perl -w

$string="one two three four";

@a=($string =~ m/(\w+)/g);

for $item (@a) {
   print $item."\n";
}



    one
two
three
four

  4. The s and m flags.
    s treat multiple lines as a single line . will match a \n.
    m the pattern is over several lines. But . does not match \n
    PerlOutput 
    #!/usr/bin/perl -w

@a=(<<"FINIS" =~ m/(.+)/g);
Hello
my
name
is
mud
FINIS

print join(" ",@a)."\n";

@a=(<<"EOD"=~m/([\w'?]+)/gs);
What's
up
Doc?
EOD


print join(" ",@a)."\n";


    Hello my name is mud
What's up Doc?

  5. i and x modifiers
    i for case insensitive.
    x for multiple line and comments see page 171-172

  6. Substitution.

    s/find/replace/;
    PerlOutput 
    #!/usr/bin/perl -w


$a="The rain in Spain, stays mainly on the plain.";

$a=~s/ain/out/;
print $a."\n";


$a="The rain in Spain, stays mainly on the plain.";
$a=~s/ain/out/g;
print $a."\n";


    The rout in Spain, stays mainly on the plain.
The rout in Spout, stays moutly on the plout.

  7. Substitute with memory
    PerlOutput 
    #!/usr/bin/perl -w
#
#
print "----------------------------------------------\n";
print `cat ip.data`;
print "----------------------------------------------\n";
open(IN,"<ip.data") || die "could not open data file\n";

while(<IN>) {
   chomp;
   next if !(/ A /);
   @fields=split;
   $_=$fields[2];
   s/(\d{1,3})\.(\d{1,3})\.(\d{1,3})\.(\d{1,3})/$4.$3.$2.$1/;
   print $_."\t\tIN\tPTR\t".$fields[0].".mydomain.com.\n";
}

close(IN);


    ----------------------------------------------
@       IN      SOA     ns.mydomain.com. hostmaster.mydomain.com. (
                        1999021004      ; serial, todays date + todays serial #
                        8H              ; refresh, seconds
                        2H              ; retry, seconds
                        1W              ; expire, seconds
                        1D )            ; minimum, seconds

                NS      ns              ; Inet Address of name server
                MX      10 mail         ; Primary Mail Exchanger
localhost       A       127.0.0.1
ns              A       192.168.102.1
linuxhost       A       192.168.102.2
mail            A       192.168.102.3
www             A       192.168.102.4 
----------------------------------------------
1.0.0.127		IN	PTR	localhost.mydomain.com.
1.102.168.192		IN	PTR	ns.mydomain.com.
2.102.168.192		IN	PTR	linuxhost.mydomain.com.
3.102.168.192		IN	PTR	mail.mydomain.com.
4.102.168.192		IN	PTR	www.mydomain.com.

  8. Delimiters
    like q and qq can use any delimiter so can m and s.

  9. Lazy vs Greedy evaluation.
    LazyGreedy
    *?*
    +?+

    PerlOutput 
    #!/usr/bin/perl -w

$a=q/Test text "one" and "two" and "three"/;
print "Text   : ".$a."\n";

print "Greedy : ";

$a=~m/(".*")/;

print $1."\n";

print "Lazy   : ";

$a=~m/(".*?")/;

print $1."\n";




    Text   : Test text "one" and "two" and "three"
Greedy : "one" and "two" and "three"
Lazy   : "one"
© Allan Kochis Last revision 10/10/2005