my @history;
while (<>) {
    print @history if /pattern/;
    push @history, $_;
    shift @history if @history > 3;
}
[download]

No direction was given for handling overlapping matches.

Or if you don't want to test the length of your array forevermore:

my @history = ("\n") x 3;
while (<>) {
    print @history if /pattern/;
    push @history, $_;
    shift @history;
}
[download]

This works very well, thank all of you for your help.

Where can I place a print statement that seperates the output results for each match?

#!/Perl/bin/perl -w

open(INFILE, "< file.log")
         or die "Cant open file : $!";

open(OUT, "> results.txt")
         or die "Cant open new file : $!";

use strict;
my $pattern = "Error while detecting new item";
my @history = ("\n") x 3;
while (<INFILE>) {
print OUT @history if /$pattern/;
    push @history, $_;
    shift @history;
}

print $.;
close INFILE;
close OUT;
exit;
[download]

Example:

results.txt
matched line 1
matched line 2
matched line 3
results of first match
matched line 1
matched line 2
matched line 3
results of second match
and so on...
[download]

Thank you again so much, Neurotoxx

++, but that should be my @history = ('') x 3;

#!/usr/bin/perl

use strict;

my @file = <DATA>;

for my $i (0 .. $#file) {
    if ( $file[$i] =~ /pattern/ ) {
        print "$file[$_]" for ($i-3 .. $i-1);
    }
}

__DATA__
log info line yada yada 0
log info line yada yada 1
log info line yada yada 2
log info line yada yada 3
log info line (pattern) 4
log info line yada yada 5
[download]

That doesn't work if the pattern is found in the first three lines. (And what's with the quotes around a variable?) Change

print "$file[$_]" for ($i-3 .. $i-1);
[download]

to

print $file[$_] for grep $_ >= 0, $i-3 .. $i-1;
[download]

Thanks ikegami. Those are good points.

Not sure I understand this:

for my $i (0 .. $#file)

Is the #file a reference to my data file?

So then, for my $i (0 .. $data.log)

Thank you.

Sorry for my ignorance.

Is the #file a reference to my data file?

No, not directly. Instead, the idea presented was to capture the whole file in @file:

my @file = <DATA>;
[download]

In your case, it would be <INFILE> instead of <DATA>. (<DATA> is a special file handle, starting after the special tag __DATA__ in the script, but otherwise pretty much a normal-behaving file handle.)

The $# in front of an array identifier means "the last index of" that array.

Perhaps better is to keep the last three lines in a fifo:

open(INFILE, "< file.log")
         or die "Cant open file : $!";

open(OUT, "> results.txt")
         or die "Cant open new file : $!";

$pattern = "Error code 2";

my @file;
while (<INFILE>) {
    print OUT @file if (/$pattern/ and @file);
    push @file, $_;           # push on the current line
    shift @file while (@file > 3); # shorten to 3 elements
}

close INFILE;
close OUT;
[download]

Don't forget warnings and strict.

-QM
--
Quantum Mechanics: The dreams stuff is made of

$#file is the index of the last element in the array @file.

Otherwise grep will do it:

grep -B3 pattern logfile

Unfortunately, I need to have it run as a Windows task. I also need it to traverse the log directory and process each file with the pattern match.

Thanks,

Neurotoxx