#!/usr/bin/env perl

# Parsing an XML feed which we get off a web server

$urlsource = "http://www.firstgreatwestern.info/coffeeshop/index.php?type=rss;action=.xml";

# Use the Library for Web Processes, "User Agent" (i.e. browser) module

use LWP::UserAgent;

# Load from web, or local copy if there's a recent one available

if (-e "latest.txt" and -M "latest.txt" > 0.01) {
        $agent = LWP::UserAgent->new; # Create me a browser
        $agent->agent("Well House Consultants"); # Set the browser name
        $req = HTTP::Request->new(GET => ($urlsource)); # Set up the request we'll run
        $res = $agent->request($req); # Run the request
        $page = $res -> content(); # Pull back the content!

        open (FHO,">latest.txt") ; # Save the raw XML (for caching purposes)
        print FHO $page;
        close FHO;

        print STDERR "Data read from server\n";

} else {

        open (FH,"latest.txt");
        read (FH,$page,-s "latest.txt");
        close FH;

        print STDERR "Cached recent copy of data used\n";
}

use XML::Parser; # XML Parser

        # $struct is a pointer to a LIST of tags in the XML document
        # within a start / end pair, you'll have other lists of tags
        # attributes are hashes

if ($ARGV[0] eq "-tree") { # DOM like
        $parser = new XML::Parser(Style => "Tree");
        $struct = $parser -> parse($page);
        displaylist("",$struct);

} elsif ($ARGV[0] eq "-stream") { # SAX like
        $parser = new XML::Parser(Style => "Stream");
        $parser -> parse($page);

} elsif ($ARGV[0] eq "-handle") { # SAX / a bit more selective
        $parser = new XML::Parser(Handlers => {Start => \&entering,
                                Char => \&handle_char});
        $parser -> parse($page);

} else {
        die ("Usage: $0 -[tree|stream|handle]\n");
}

################################################################

sub displaylist {
        my ($inset,$current) = @_;
        $inset .= " ";
        my $k;
        for ($k=0;$k<@$current;$k++) {
                if ($type = ref $$current[$k]) {
                        print ("$inset$k: $type\n");
                        if ($type eq "ARRAY") {
                                displaylist($inset,$$current[$k]);
                        } elsif ($type eq "HASH") {
                                displayhash($inset,$$current[$k]);
                        }
                } else {
                        my $info = $$current[$k];
                        $info =~ tr/ -~/!/c;
                        print ("$inset$k: $info\n");
                }
        }
}

sub displayhash {
        my ($inset,$current) = @_;
        $inset .= " ";
        my $k;
        foreach $k (sort keys %$current) {
                if ($type = ref $$current{$k}) {
                        print ("$inset$k: $type\n");
                        if ($type eq "ARRAY") {
                                displaylist($inset,$$current{$k});
                        } elsif ($type eq "HASH") {
                                displayhash($inset,$$current{$k});
                        }
                } else {
                        my $info = $$current{$k};
                        $info =~ tr/ -z/!/c;
                        print ("$inset$k: $info\n");
                }
        }
}

##########################################################################################

sub StartTag {
        print "Start Tag -";
        ppars(@_);
        }
sub EndTag {
        print "End Tag -";
        ppars(@_);
        }
sub Text {
        print "Text -";
        ppars(@_);
        }
sub StartDocument {
        print "Start Document -";
        ppars(@_);
        }
sub EndDocument {
        print "End Document -";
        ppars(@_);
        }

sub ppars {
        my($obj,$el) = @_;
        print " Object: $obj ";
        print "Element: $el\n";
        }

###################################################################################################

sub entering {
        # say_pars("tag", @_); # for debug / demo
        $latest = $_[1];
        }
sub handle_char {
        # say_pars("text", @_); # for debug / demo
        if ($latest eq "guid") {
                say_pars("url", @_);
        }
        if ($latest eq "title") {
                say_pars("text", @_);
        }
}

sub say_pars {
        my($what,$obj,$el) = @_;
        return unless ($el =~ /\S+/);
        print "$what: $el\n";
        }

__END__

Sample output

munchkin:s12perl grahamellis$ perl xmx -handle
Cached recent copy of data used
text: First Great Western Coffee Shop
text: Re: A good problem to have?
url: http://www.firstgreatwestern.info/coffeeshop/index.php?topic=11326.msg118375#msg118375
text: Re: Unique opportunity to travel the Portbury Branch Line - Saturday 29 Sept 2012
url: http://www.firstgreatwestern.info/coffeeshop/index.php?topic=11167.msg118374#msg118374
text: Re: A good problem to have?
url: http://www.firstgreatwestern.info/coffeeshop/index.php?topic=11326.msg118373#msg118373
text: Re: FGW (bad) experience 27/9/2012 - Maidenhead to Paddington
url: http://www.firstgreatwestern.info/coffeeshop/index.php?topic=11318.msg118371#msg118371
text: Re: GWS Steam Railmotor on Looe Valley Line Suns 4 and 11 November
url: http://www.firstgreatwestern.info/coffeeshop/index.php?topic=11317.msg118370#msg118370
munchkin:s12perl grahamellis$