RSS Parser Email Script 2

#!/usr/bin/perl

##################################
# Andres                         #
# afelotreyu @ gmail . com       #
# Date:        June  20  2011    #
##################################
#
# This perl Script reads RSS URLS and emails its contents to a given address.
# The RSS urls must be in a file see below.
# ########### Log Format
# Please note the 'rss_log' file must have the following format
# URL|||title
# http://file.xml|||title
# http://file.xml

use vars qw/ $REVISION $VERSION /;
$REVISION='0.01';
$VERSION="1.0 (Build $REVISION)";

# Use CPAN modules
use strict;
use XML::Simple;
use Data::Dumper;
use CGI;
use LWP;
use URI;
use Net::SMTP;
my $cgi = new CGI;
my $browser = LWP::UserAgent->new;

########################
# Configuration
# please modify carefully as needed
my %configval = (
  'rss_log'       => 'somelog.txt',              # File where to find RSS list and log it
  'tmp_xml_file'    => 'sometimefile.xml',
  'subject_email'    => 'RSS Parse email',
  'email_method'    => 'sendmail',                # sendmail or smtp
  'email_to'      => 'some_user@somedomain.com',        # To - Email Address(es)
  'email_from'     => 'some_user@somedomain.com',        # From - email address
  'sendmail'      => '/usr/sbin/sendmail -t',         # Sendmail Location and params
  'smtp_server'     => 'somedomain.local',            # This is the SMTP
  'smtp_timeout'    => '30',
);
########################

# Important Variables
my ($data,$source,$content);
my $new = 0;      # New items = no!
my $send = 0;     # Send Email = no!
# Temporal variables that we use over the script
my ($tmp,$a,$e,$line,$link,$last);

########################
# Lest get started
# First we read the log file
$source = ReadLOG_file();
# $content builds the email message
$content = "RSS Parse Email Listing\n\n";
# Be nice, print some output for html
print $cgi->header('text/html');

# Loop over every RSS link
for ( $a=0;$a < = $#{$source}; $a++) {
        $last = 0;
        $content = $content."RSS->[$a]: \n";
        # Download the File of the RSS
        &DownloadXML_file($source->[$a]->[0]);
        # Read the file
        $data = &ReadXML_file($configval{'tmp_xml_file'});
        # Check if there are any items defined
        if ($data->{channel}->{item}) {
                # We have items, lets chech output
                if (ref($data->{channel}->{item}) eq 'ARRAY') {
                        foreach $e ( @{$data->{channel}->{item}} ) {
                                # OK we are in, now we store all posts since last one
                                if ("$e->{title}" eq "$source->[$a]->[1]" ) {
                                        $content = $content."\nLast Item: $source->[$a]->[1] \n";
                                        last;
                                } else {
                                        # New Item, we need to store it!
                                        if ( $new == 0 ) { $new = 1; }
                    if ( $last == 0 ) { $source->[$a]->[2] = $e->{title}; $last = 1; }
                                        $send = 1;    # Send Email
                                        $content = $content."\nTitle: $e->{title} \n";
                                        $content = $content."Link: $e->{link} \n";
                                        $content = $content."Description: $e->{description} \n";
                                        $content = $content."Date: $e->{pubDate} \n";
                                }
                        }
                } else {
                        # Only one item, is it the same as log?
                        if ("$data->{channel}->{item}->{title}" eq "$source->[$a]->[1]") {
                                # Same item as before
                                $content = $content."\nLast Item: $source->[$a]->[1] \n";
                        } else { # New Item
                                # New Item, we need to store it!
                                if ( $new == 0 ) { $new = 1; }
                if ( $last == 0 ) { $source->[$a]->[2] = $data->{channel}->{item}->{title}; $last = 1; }
                                $send = 1;    # Send Email
                                $content = $content."title: $data->{channel}->{item}->{title} \n";
                                $content = $content."Link: $data->{channel}->{item}->{link} \n";
                                $content = $content."Description: $data->{channel}->{item}->{description} \n";
                                $content = $content."Date: $data->{channel}->{item}->{pubDate} \n";
                        }
                }
        } else { # No items, maybe empty?
                $content = $content."RSS Feed is currently Empty - \n";
        }
        # Next RSS!
        $content = $content."\n_____________________________________\n\n";
};

# A new item was set, update log file
if ( $new == 1 ) {
        open (NEWLOG,">",$configval{'rss_log'}) or die (print "$configval{'rss_log'} - File could not open to write $!");
        # Loop over every RSS link
        for ( $a=0;$a < = $#{$source}; $a++) {
                if ($source->[$a]->[2]) {
                        print NEWLOG $source->[$a]->[0].'|||'.$source->[$a]->[2]."\n";
                } else {
                        print NEWLOG $source->[$a]->[0].'|||'.$source->[$a]->[1]."\n";
                }
        }
    close NEWLOG;

}
# Do we need to email?
if ($send == 1) {
        &SendMail($configval{'subject_email'},$content);
}
# Print something for HTML access
print "<html><body><h1>Finished</h1><h1><body></body></h1></body></html>";

########################
# Subroutines go here
########################
# ReadXML_file
# Needs to get a file location to read its contents
# It returns a dereference hash ref
########################
sub ReadXML_file() {

        my $file = $_[0];
        # Read the XML file
        my $xml = new XML::Simple (
                KeyAttr       => [],
                SuppressEmpty => undef);

        # read XML file
        $data = $xml->XMLin($file);
        return $data

        ##### Sample code!
        # Use of the XML::Simple output!
        # dereference hash ref
        # access <employee> array
        #foreach $e (@{$data->{data}->{robots}->{robot}})
        #{
                #print $e->{name}, "\n";
                #print "Age/Sex: ", $e->{id}, "/",  $e->{icon}, "\n";
                #print "Department: ", $e->{url_company}, "\n";
                #print "\n";
        #}
}

########################
# ReadLOG_file
########################
sub ReadLOG_file() {
        open (READLOG,"< ",$configval{'rss_log'}) or die "File $configval{'rss_log'} could not open $! \n";
        flock (READLOG,2);
        seek(READLOG,0,0);
        $e = 0;
        foreach $line (<READLOG>) {
                chomp $line;
                if ($line =~ m/^http/ ) {
                        my @pair = split(/\|\|\|/,$line);
                        $source->[$e]->[0] = $pair[0];
                        $source->[$e]->[1] = $pair[1];
                        $e++;
                }
        }
        close READLOG;
        return $source;
}

########################
# DownloadXML_file
# needs a URL to connect to and download the XML file.
########################
sub DownloadXML_file() {
    # Download XML
    # We are going to connect to the page and download the XML file
    my $url = URI->new($_[0]);
    $url->query_form();
    my $mfilecont = $browser->get($url);
   
    $tmp = $configval{'tmp_xml_file'};
    open (TEMPORAL,">", $tmp) or die (print "$tmp - File could not open $!");
    print TEMPORAL $mfilecont->content;
    close TEMPORAL;

        # Return a success signal '
    return 1;
}

########################
# SendMail(subject,message);
# Needs Subject and Message content
# Uses SMTP configuration
########################
sub SendMail() {
        my $subject = $_[0];
        my $message = $_[1];

        #### Email
        # Here is where the email is created
        for ($configval{'email_method'}) {
                if (/^smtp$/) {
                        my $smtp = Net::SMTP->new(
                                        $configval{'smtp_server'},                              # MailHost
                                        Timeout => $configval{'smtp_timeout'},    # TimeOut
                        );
                        $smtp->mail("$configval{'email_from'}") || die (  print "From_email call bad $! \n" );     # Address Email is created from (Check if exist)
                        $smtp->to("$configval{'email_to'}") || die ( print "To email call bad $! \n" ); # Address Email is going to (Check if exist)
                        $smtp->data();
                        $smtp->datasend("To: $configval{'email_to'} \n"); # Address Email is going to
                        $smtp->datasend("Subject: $subject \n"); # Subject
                        $smtp->datasend("\n");
                        $smtp->datasend("$message");
                        $smtp->dataend();
                        $smtp->quit;
                } elsif (/^sendmail$/) {
                        open(SENDMAIL, "| $configval{'sendmail'} ") or die "Cannot open $configval{'sendmail'}: $!";
                        print SENDMAIL "Reply-to: ".$configval{'email_from'}."\n";
                        print SENDMAIL "Subject: ".$subject."\n";
                        print SENDMAIL "To: ".$configval{'email_to'}."\n";
                        print SENDMAIL "Content-type: text/plain\n\n";
                        print SENDMAIL $message."\n";
                        print SENDMAIL '.';
                        close(SENDMAIL);
                }
        }
        #### Done Email
}
########################

1;
__END__