File::DirWalk makes the directory recursion easy. Perl's builtins trivialize the rest. The canonical way to use DirWalk is something like so:

#!/usr/bin/env perl

use 5.014;
use warnings;
use File::DirWalk;

# Process commandline arguments
die "usage: $0 pattern path ..." if (@ARGV < 2);
my $pattern = shift;
# @ARGV now contains one or more paths to recurse

# Set up DirWalk object with callback to grep each file
my $dw = new File::DirWalk;
$dw->onFile(sub {
    my $file = shift;
    
    if (-T $file) {
        open my $fh, '<', $file or die "Can't open $file: $!";
        print map { "$file: $_" } grep { /$pattern/ } <$fh>;
        close $fh;
    }

    return File::DirWalk::SUCCESS;
});
    
$dw->walk($_) for @ARGV;
[download]

The above gives you a very simple grep -R command. Probably more instructive than directly useful.

The -T operator is one of Perl's file test operators that evaluates to true if the file is an ASCII text file (based on some heuristic guesswork). There is an equivalent -B test for binary files.

#!/usr/bin/perl

use Modern::Perl;
use autodie;

my ($dir, $match) = @ARGV;   

die "directory does not exists $dir" unless -d $dir;
die "No match" unless $match;

open_readdir($dir, \&process);

sub process { 
        my ($dir, $file) = @_;
        if (-T "$dir/$file") {

                open my $fh, '<', "$dir/$file";
                say "$dir/$file: $_"
                        for grep /$match/, $fh->getlines;
                $fh->close;
        }
}

sub open_readdir {
        (my $dir = shift) =~ s/\/$//;
        my $code = shift;
        my $db;

        opendir $dh, $dir;
        my @files = grep !/^\./, readdir $dh;
        
        for my $file (@files) {
                if (-d "$dir/$file") {
                        open_readdir("$dir/$file", \&process);
                } else {
                        &$code($dir, $file);
                }
        }
        closedir($dh);
}
[download]

use File::Find::Rule;

my @specs = ( qr/pattern/i, [ qr/notpattern/i ], \&oneline );
find(
  file =>
  ascii =>
  exec => \&onename,
  grep => \@specs,
  in => \@dirs,
);

sub onename {
  my( $shortname, $path, $fullname )= @_;
  $_ = $shortname;
  1;
}
sub oneline {
  my( $shortname, $path, $fullname )= @_;
  my $line = $_;
  1;
}
[download]