use strict;

my $counter=1;
$/="</Rec $counter>";
my $first_file;
my $second_file;

while (my $record=<DATA>) {
    $first_file .= $record unless $record=~m/var2 = 3/;
    $second_file .= $record if $record=~m/var2 = 3/;
    $counter++;
    $/="</Rec $counter>";
}
print "FIRST FILE: $first_file\nSECOND FILE: $second_file\n";
[download]

use strict;

open OF1, ">var2_equal_to_3.txt";
open OF2, ">var2_not_equal_to_3.txt";
my ($flipflop, $flag, $prev);
while (<DATA>) {
    chomp;
    $flag = 1 if m/var2\s+=\s+3/i;
    $flipflop = /<rec\s+[0-9]+>/i .. /<\/rec\s+[0-9]+>/i;
    $prev = $_ if $flipflop == 1;
    if ($flag) {
        print OF1 "$prev\n"  if $flipflop == 2;
        print OF1 "$_\n";
    }else{
        print OF2 "$_\n";
    }
    $flag  = $prev = undef if $flipflop =~ /E0$/;
}

close OF1;
close OF2;
[download]

use strict;

my $match = 'var2 = 3';
my $in = 'in.txt';
my $tmp = 'temp.txt';
my $ext = 'extract.txt';

my ($inh, $tmph, $exth, $last);

open($inh, $in);
open($tmph, ">$tmp");
open($exth, ">$ext");

while (<$inh>) {
    if (index($_, $match) == -1) {
        print $tmph $last; $last = $_; next;
    }
    print $exth $last, $_; $last = '';
    print $exth (scalar<$inh>) for (1..9);
}

close($inh);
close($tmph);
close($exth);

unlink($in);
rename($tmp, $in);
[download]

I tried this, and it worked perfect
Thanks a bunch :)

$record_length = 11;
$string_sought = 'whatever';

open (SOURCEFILE, "file_and_path");
@SOURCEFILE_LINES = <SOURCEFILE>;
close(SOURCEFILE);

$m = 0;
$n = 0;
for $i (0 .. $#SOURCEFILE_LINES-$record_length) {
     if ($SOURCEFILE_LINES[$i+1] =~ m/$string_sought/) {
          for $j (0 .. $record_length-1) {
               $TYPE_1_LINES[$m] = $SOURCEFILE_LINES[$i+$j];
               $m++;
          }
          $i = $i+$record_length;
     }
     else {
          $OTHER_LINES[$n] = $SOURCEFILE_LINES[$i];
          $n++;
     }
}
finish off last few lines - if sought record not found by now, there c
+an't be another one
for $i ($#SOURCEFILE_LINES-$record_length+1 .. $#SOURCEFILE_LINES) {
     $OTHER_LINES[$n] = $SOURCEFILE_LINES[$i];
     $n++;
}

#now write your two arrays to separate files, overwriting SOURCEFILE i
+f you like with @OTHER_LINES.
[download]

If the tagging for the data structure went like this, it would qualify as valid XML:

<rec id=4>
varFoo = bar
some text
some text
...
</rec>
[download]

# command line perl script to put all "var2 = 3" chunks into a separat
+e file:

perl -ne 'BEGIN{ $/="</rec>\n" } print if /var2 = 3/' input > var2_3.o
+utput

# just do the opposite (change "if" to "unless") to save the other chu
+nks elsewhere
[download]