[Perl] DH Spider 0.5

#!usr/bin/perl
# DH Spider 0.5
#(C) Doddy Hackman 2016
#Credits : Regex based on
#http://stackoverflow.com/questions/15710275/print-email-addresses-to-a-file-in-perl
#Thanks to motherconfessor & amon
#http://doddyhackman.webcindario.com/?tipo=perl&id=137

use DH_Searcher;
use LWP::UserAgent;
use IO::Socket::SSL;
use Getopt::Long;
use Color::Output;
Color::Output::Init;
use Try::Tiny;

my @agents = (
'Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:25.0) Gecko/20100101 Firefox/25.0',
    'Opera/9.80 (Windows NT 6.0) Presto/2.12.388 Version/12.14',
'Mozilla/5.0 (Windows NT 6.2; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/32.0.1667.0 Safari/537.36',
'Mozilla/5.0 (compatible; MSIE 10.6; Windows NT 6.1; Trident/5.0; InfoPath.2; SLCC1; .NET CLR 3.0.4506.2152; .NET CLR 3.5.30729; .NET CLR 2.0.50727) 3gpp-gba UNTRUSTED/1.0',
'Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.8.1.8pre) Gecko/20070928 Firefox/2.0.0.7 Navigator/9.0RC1',
    'Mozilla/5.0 (Windows; U; MSIE 9.0; WIndows NT 9.0; en-US))',
'Mozilla/5.0 (Windows NT 6.0; rv:2.0) Gecko/20100101 Firefox/4.0 Opera 12.14',
'Mozilla/5.0 (Windows; U; Windows NT 6.1; tr-TR) AppleWebKit/533.20.25 (KHTML, like Gecko) Version/5.0.4 Safari/533.20.27'
);

my $buscador = qr/[A-Z0-9._%+-]+\@[A-Z0-9.-]+\.[A-Z]{2,4}/i; # Thanks to motherconfessor & amon

GetOptions(
    "google"   => \$google,
    "bing"  => \$bing,
    "dork=s"  => \$dork,
    "count=i"  => \$count,
    "filename=s"   => \$filename,
    "page=s"   => \$page,
    "savefile=s"   => \$savefile
);

head();

if ($google) {
    if($dork ne "" && $count ne "") {
        dh_spider($dork,$count,"","google");
    } else {
        printear("[-] Use dork and count to scan in google\n");
    }
}
elsif ($bing) {
    if($dork ne "" && $count ne "") {
        dh_spider($dork,$count,"","bing");
    } else {
        printear("[-] Use dork and count to scan in google\n");
    }
}
elsif ($filename) {
    if(-f $filename) {
        dh_spider("","",$filename,"filename");
    } else {
        printear_rojo("[-] File $filename not found\n");
    }
} elsif($page) {
    if($page ne "") {
        dh_spider_single($page);
    }
} else {
    sintax();
}

copyright();

# Functions

sub dh_spider_single {
    my $page = shift;
    my $contador = "0";
    my $algo = 0;
    my @emails_found;
    printear_logo("[+] Searching emails in page ...\n\n");
    my $code = toma($page);
    while ( $code =~ /($buscador)/g ) {
        my $email = $1;
        push( @emails, $email );
        printear_titulo("[+] $page : ");
        printear_azul("$email\n");
        $algo = 1;
        $contador++;
        push(@emails_found,$email);
    }
    if($algo ne "1") {
        printear_titulo("[-] $page : ");
        printear_rojo("Emails not found\n");
    }
    $algo = 0;
    printear("\n[+] Emails Found : ");
    print "$contador\n";
    if($savefile && $contador > 0) {
        my @emails = repes(@emails_found);
        for my $email(@emails) {
            chomp $mail;
            savefile($savefile,$email);
        }
        printear("\n[+] Logs $savefile saved\n");
    }
    printear_logo("\n[+] Finished\n");
}

sub dh_spider {
    my($dork,$count,$filename,$option) = @_;

    my $contador = "0";
    my $algo = 0;

    my @links;
    my @emails_found;

    my $searcher = DH_Searcher->new();

    if($option eq "filename") {
        printear("[+] Opening file : ");
        print "$filename\n\n";
        @links = open_file($filename);
    } elsif($option eq "bing") {
        printear("[+] Service : ");
        print "Bing\n";
        printear("[+] Dork : ");
        print "$dork\n";
        printear("[+] Count : ");
        print "$count\n\n";
        printear_logo("[+] Searching pages ...\n\n");
        @links = $searcher->search_bing($dork,$count*10);
    } elsif($option eq "google") {
        printear("[+] Service : ");
        print "Google\n";
        printear("[+] Dork : ");
        print "$dork\n";
        printear("[+] Count : ");
        print "$count\n\n";
        printear_logo("[+] Searching pages ...\n\n");
        @links = $searcher->search_google($dork,$count*10);
    }

    printear("[+] Links found : ");
    print int(@links)."\n\n";
    printear_logo("[+] Searching emails ...\n\n");
    for my $link(@links) {
        chomp $link;
        my $code = toma($link);
        while ( $code =~ /($buscador)/g ) {
            my $email = $1;
            push( @emails, $email );
            printear_titulo("[+] $link : ");
            printear_azul("$email\n");
            $algo = 1;
            $contador++;
            push(@emails_found,$email);
        }
        if($algo ne "1") {
            printear_titulo("[-] $link : ");
            printear_rojo("Emails not found\n");
        }
        $algo = 0;
    }
    printear("\n[+] Emails Found : ");
    print "$contador\n";
    if($savefile && $contador > 0) {
        my @emails = repes(@emails_found);
        for my $email(@emails) {
            chomp $mail;
            savefile($savefile,$email);
        }
        printear("\n[+] Logs $savefile saved\n");
    }
    printear_logo("\n[+] Finished\n");
    $searcher->destroy();
}

# More Functions

sub toma {

    my $page = shift;

    my $contenido = "";

    if($page ne "") {
        try {
            my $nave = LWP::UserAgent->new(ssl_opts => {verify_hostname => 0,SSL_verify_mode => IO::Socket::SSL::SSL_VERIFY_NONE});
            $nave->agent( $agents[ rand @agents ] );
            $nave->timeout($timeout);
            $contenido = $nave->get($page)->content;
        } catch {
            $contenido = "Error";
        };

    } else {
        $contenido = "Error";
    }

    return $contenido;
}

sub open_file {
    my $file = shift;
    my @words;
    open( FILE,$file);
    @words = <FILE>;
    close FILE;
    return @words;
}

sub repes {
    my @limpio;
    foreach $test (@_) {
        push @limpio, $test unless $repe{$test}++;
    }
    return @limpio;
}

sub savefile {
    my ($filename,$text) = @_;
    open( SAVE, ">>" . $filename );
    print SAVE $text . "\n";
    close SAVE;
}

sub printear {
    cprint( "\x036" . $_[0] . "\x030" );
}

sub printear_logo {
    cprint( "\x037" . $_[0] . "\x030" );
}

sub printear_titulo {
    cprint( "\x0310" . $_[0] . "\x030" );
}

sub printear_rojo {
    cprint( "\x035" . $_[0] . "\x030" );
}

sub printear_azul {
    cprint( "\x033" . $_[0] . "\x030" );
}

sub sintax {

    printear("[+] Sintax : ");
    print "perl $0 <option> <value>\n";
    printear("\n[+] Options : \n\n");
    print "-google -dork <dork> -count <count> : Find emails in google\n";
    print "-bing -dork <dork> -count <count> : Find bing in google\n";
    print "-filename <filename> : Find emails in filename with URLS\n";
    print "-page <page> : Find emails in a single page\n";
    print "-savefile <savefile> : Save emails in logs\n";
    printear("\n[+] Example : ");
    print "perl dh_spider.pl -bing -dork contacto.php -count 5 -savefile emails_found.txt\n";
    copyright();
}

sub head {
    printear_logo("\n-- == DH Spider 0.5 == --\n\n\n");
}

sub copyright {
    printear_logo("\n\n-- == (C) Doddy Hackman 2016 == --\n\n");
    exit(1);
}

#The End ?