File Coverage

blib/lib/WWW/Scraper/ISBN/WHSmith_Driver.pm

Criterion	Covered	Total	%
statement	84	84	100.0
branch	13	24	54.1
condition	8	18	44.4
subroutine	10	10	100.0
pod	1	1	100.0
total	116	137	84.6

package WWW::Scraper::ISBN::WHSmith_Driver;

use strict;
use warnings;

use vars qw($VERSION @ISA);
$VERSION = '0.07';

#--------------------------------------------------------------------------

=head1 NAME

WWW::Scraper::ISBN::WHSmith_Driver - Search driver for the WHSmith online book catalog.

=head1 SYNOPSIS

See parent class documentation (L)

=head1 DESCRIPTION

Searches for book information from the WHSmith online book catalog

=cut

#--------------------------------------------------------------------------

###########################################################################
# Inheritence

use base qw(WWW::Scraper::ISBN::Driver);

###########################################################################
# Modules

use WWW::Mechanize;

###########################################################################
# Constants

use constant	REFERER	=> 'http://www.whsmith.co.uk';
use constant	SEARCH	=> 'http://www.whsmith.co.uk/pws/ProductDetails.ice?ProductID=%s&keywords=%s&redirect=true';
use constant    PRODUCT => '/products/[^/]+/product/';

#--------------------------------------------------------------------------

###########################################################################
# Public Interface

=head1 METHODS

=over 4

=item C

Creates a query string, then passes the appropriate form fields to the
WHSmith server.

The returned page should be the correct catalog page for that ISBN. If not the
function returns zero and allows the next driver in the chain to have a go. If
a valid page is returned, the following fields are returned via the book hash:

  isbn          (now returns isbn13)
  isbn10
  isbn13
  ean13         (industry name)
  author
  title
  book_link
  image_link
  description
  pubdate
  publisher
  binding       (if known)
  pages         (if known)
  weight        (if known) (in grammes)
  width         (if known) (in millimetres)
  height        (if known) (in millimetres)

The book_link and image_link refer back to the WHSmith website.

=back

=cut

sub search {
	my $self = shift;
	my $isbn = shift;
	$self->found(0);
	$self->book(undef);

    # validate and convert into EAN13 format
    my $ean = $self->convert_to_ean13($isbn);
    return $self->handler("Invalid ISBN specified [$isbn]")
        if(!$ean || (length $isbn == 13 && $isbn ne $ean)
                 || (length $isbn == 10 && $isbn ne $self->convert_to_isbn10($ean)));

    $isbn = $ean;
#print STDERR "\n# isbn=[\n$isbn\n]\n";

	my $mech = WWW::Mechanize->new();
    $mech->agent_alias( 'Linux Mozilla' );
    $mech->add_header( 'Accept-Encoding' => undef );
    $mech->add_header( 'Referer' => REFERER );

    my $search = sprintf SEARCH, $isbn, $isbn;
#print STDERR "\n# search=[$search]\n";
    eval { $mech->get( $search ) };
    return $self->handler("the WHSmith website appears to be unavailable.")
	    if($@ || !$mech->success() || !$mech->content());

  	# The Book page
    my $html = $mech->content();
	return $self->handler("Failed to find that book on the WHSmith website. [$isbn]")
		if($html =~ m!Sorry, no products were found!si);

    my $url = $mech->uri();
	return $self->handler("Failed to find that book on the WHSmith website. [$isbn]")
		if($url =~ m!Error.aspx!si);

    $html =~ s/&/&/g;
    $html =~ s/�?39;/'/g;
    $html =~ s/ / /g;

#print STDERR "\n# html=[\n$html\n]\n";

    my $data;
    ($data->{isbn13})           = $html =~ m!\s*ISBN13:\s*(.*?)!si;
    ($data->{isbn10})           = $html =~ m!\s*ISBN10:\s*(.*?)!si;
    ($data->{publisher})        = $html =~ m!Publisher:\s*([^<]+)!si;
    ($data->{pubdate})          = $html =~ m!\s*publication date:\s*(.*?)!si;
    ($data->{title})            = $html =~ m!([^<]*)!si;
    ($data->{binding})          = $html =~ m!.*?([^<]+)!si;
    ($data->{binding})          = $html =~ m!\s*Format:\s*(.*?)!si  unless($data->{binding});
    ($data->{pages})            = $html =~ m!\s*Number Of Pages:\s*(.*?)!si;
    ($data->{author})           = $html =~ m!By:(.*?)!si;
    ($data->{image})            = $html =~ m!!si;
    ($data->{description})      = $html =~ m!!si;

    if($data->{image}) {
        $data->{thumb} = $data->{image};
        $data->{thumb} =~ s!/x?large/!/small/!;
    }

    # currently not provided
    ($data->{width})            = $html =~ m!Width:\s*([^<]+)!si;
    ($data->{height})           = $html =~ m!Height:\s*([^<]+)!si;
    ($data->{weight})           = $html =~ m!\s*weight:\s*(.*?)!s;

    $data->{width}  = int($data->{width})   if($data->{width});
    $data->{height} = int($data->{height})  if($data->{height});
    $data->{weight} = int($data->{weight})  if($data->{weight});

    if($data->{author}) {
        $data->{author} =~ s/<[^>]*>//g;
        $data->{author} =~ s/\(author\)/ /g;
        $data->{author} =~ s/\s+/ /g;
        $data->{author} =~ s/\s+,\s+/, /g;
    }

#use Data::Dumper;
#print STDERR "\n# data=" . Dumper($data);

	return $self->handler("Could not extract data from The WHSmith result page.")
		unless(defined $data);

	# trim top and tail
	foreach (keys %$data) {
        next unless(defined $data->{$_});
        $data->{$_} =~ s/^\s+//;
        $data->{$_} =~ s/\s+$//;
    }

	my $bk = {
		'ean13'		    => $data->{isbn13},
		'isbn13'		=> $data->{isbn13},
		'isbn10'		=> $data->{isbn10},
		'isbn'			=> $data->{isbn13},
		'author'		=> $data->{author},
		'title'			=> $data->{title},
		'book_link'		=> $url,
		'image_link'	=> $data->{image},
		'thumb_link'	=> $data->{thumb},
		'description'	=> $data->{description},
		'pubdate'		=> $data->{pubdate},
		'publisher'		=> $data->{publisher},
		'binding'	    => $data->{binding},
		'pages'		    => $data->{pages},
		'weight'		=> $data->{weight},
		'width'		    => $data->{width},
		'height'		=> $data->{height},
        'html'          => $html
	};

#use Data::Dumper;
#print STDERR "\n# book=".Dumper($bk);

    $self->book($bk);
	$self->found(1);
	return $self->book;
}

1;

__END__