File Coverage

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

Criterion	Covered	Total	%
statement	21	70	30.0
branch	0	18	0.0
condition	0	18	0.0
subroutine	7	8	87.5
pod	1	1	100.0
total	29	115	25.2

package WWW::Scraper::ISBN::Blackwell_Driver;

use strict;
use warnings;

use vars qw($VERSION @ISA);
$VERSION = '1.00';

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

=head1 NAME

WWW::Scraper::ISBN::Blackwell_Driver - Search driver for Blackwell's online book catalog.

=head1 SYNOPSIS

See parent class documentation (L)

=head1 DESCRIPTION

Searches for book information from Blackwell's online book catalog.

=cut

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

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

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

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

use WWW::Mechanize;

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

use constant REFERER => 'http://bookshop.blackwell.co.uk';
use constant SEARCH  => 'http://bookshop.blackwell.co.uk/jsp/search_results.jsp?wcp=1&quicksearch=1&cntType=&searchType=keywords&searchData=%s&x=10&y=10';

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

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

=head1 METHODS

=over 4

=item C

Creates a query string, then passes the appropriate form fields to the
Blackwell 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
  thumb_link
  description
  pubdate
  publisher
  binding       (if known)
  weight        (if known) (in grammes)
  width         (if known) (in millimetres)
  height        (if known) (in millimetres)

The book_link, image_link and thumb_link all refer back to the Blackwell 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 $self->handler("Invalid ISBN specified")
        if(!$ean || (length $isbn == 13 && $isbn ne $ean)
                 || (length $isbn == 10 && $isbn ne $self->convert_to_isbn10($ean)));
    my $isbn10 = $self->convert_to_isbn10($ean);

    my $mech = WWW::Mechanize->new();
    $mech->agent_alias( 'Windows IE 6' );
    $mech->add_header( 'Accept-Encoding' => undef );

    my $search = sprintf SEARCH , $ean;
#print STDERR "\n# search=[$search]\n";

    eval { $mech->get( $search ) };
    return $self->handler("Blackwell's website appears to be unavailable.")
        if($@ || !$mech->success() || !$mech->content());

    # The Book page
    my $html = $mech->content();

    return $self->handler("Blackwell's website appears to be unavailable.")
        if($html =~ m!Blocked IP Address due to Suspicious Activity!si);

    return $self->handler("Failed to find that book on the Blackwell website. [$isbn]")
        if($html =~ m!Sorry, there are no results for!si);

    $html =~ s/&/&/g;
    $html =~ s/ / /g;
    $html =~ s/–/-/g;
#print STDERR "\n# html=[\n$html\n]\n";

    my $data;
    ($data->{isbn13})      = $html =~ m!!si;
    ($data->{title})       = $html =~ m!!si;
    ($data->{author})      = $html =~ m!]*>\s*]+>([^<]+)!si;
    ($data->{binding})     = $html =~ m!\s*([^<\(]+)!si;
    ($data->{publisher})   = $html =~ m!
    ($data->{pubdate})     = $html =~ m!
    ($data->{description}) = $html =~ m!!si;
    ($data->{pages})       = $html =~ m!
    ($data->{weight})      = $html =~ m!
    ($data->{height})      = $html =~ m!
    ($data->{width})       = $html =~ m!
    ($data->{image})       = $html =~ m!!si;

    $data->{image}     =~ s!https:///!https://bookshop.blackwell.co.uk/! if($data->{image});
    $data->{thumb}     = $data->{image};
    $data->{isbn10}    = $isbn10;
    $data->{author}    =~ s/\s*\(author\)//si if($data->{author});
    $data->{publisher} =~ s/�?39;/'/g if($data->{publisher});

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

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

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

    my $url = $mech->uri();

    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},
        'height'      => $data->{height},
        'width'       => $data->{width},
        'html'        => $html
    };

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

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

1;

__END__