File Coverage

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

Criterion	Covered	Total	%
statement	37	65	56.9
branch	1	16	6.2
condition	6	30	20.0
subroutine	9	9	100.0
pod	1	1	100.0
total	54	121	44.6

package WWW::Scraper::ISBN::Yahoo_Driver;

use strict;
use warnings;

use vars qw($VERSION @ISA);
$VERSION = '0.22';

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

=head1 NAME

WWW::Scraper::ISBN::Yahoo_Driver - Search driver for Yahoo Books online catalog.

=head1 SYNOPSIS

See parent class documentation (L)

=head1 DESCRIPTION

Searches for book information from the Yahoo Books online catalog.

=cut

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

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

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

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

use WWW::Scraper::ISBN::Driver;
use WWW::Mechanize;

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

use constant    YAHOO   => 'http://shopping.yahoo.com';
use constant    SEARCH  => 'https://shopping.yahoo.com/search?fr=yshoppingheader_test2&type=2button&did=0&p=';

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

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

=head1 METHODS

=over 4

=item C

Creates a query string, then passes the appropriate form fields to the
Yahoo Books 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        (no longer provided by Yahoo on page)
  isbn13
  ean13         (industry name)
  title
  author
  pubdate       (no longer provided by Yahoo on page)
  publisher
  book_link
  image_link
  thumb_link    (same as image_link)
  description
  binding       (if known)
  pages         (no longer provided by Yahoo on page)
  weight        (no longer provided by Yahoo on page)
  width         (no longer provided by Yahoo on page)
  height        (no longer provided by Yahoo on page)

The book_link and image_link refer back to the Yahoo Books website.

=back

=cut

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

    # 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;

    my $mech = WWW::Mechanize->new;
    $mech->agent_alias( 'Linux Mozilla' );

    eval { $mech->get( YAHOO . '/books' ) };
    return $self->handler("Yahoo! book website appears to be unavailable.")
	    if($@ || !$mech->success() || !$mech->content());

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


    # The Results page
    my $content = $mech->content();
    #print STDERR "\n# results=[\n$content\n]\n";

    my ($link,$thumb) = $content =~ m!]+>
    #print STDERR "\n# link=[$link]\n# thumb=[$thumb]\n";

    return $self->handler("Failed to find that book on Yahoo! book website.")
        unless(defined $link);

    $data->{book_link} = $link;

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


	# The Book page
    my $html = $mech->content();
    #print STDERR "\n# page=[\n$html\n]\n";

	return $self->handler("Could not extract data from Yahoo! result page.")
		unless($html =~ m!\s*!);

#    $data->{thumb_link}     = $thumb;
#    ($data->{image_link})   = $html =~ m!!is;
    ($data->{title})        = $html =~ m!]*>([^<]*)!is;
    ($data->{description})  = $html =~ m!(.*?)!is;
    ($data->{publisher})    = $html =~ m!
    ($data->{binding})      = $html =~ m!
    ($data->{author})       = $html =~ m!
    ($data->{isbn13})       = $ean;
    ($data->{isbn10})       = $self->convert_to_isbn10($ean);

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

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

    my $bk = {
        'isbn13'        => $data->{isbn13},
        'isbn10'        => $data->{isbn10},
        'isbn'          => $data->{isbn13},
        'ean13'         => $data->{isbn13},
        'author'        => $data->{author},
        'title'         => $data->{title},
		'book_link'		=> $data->{book_link},
        'image_link'    => $thumb,
        'thumb_link'    => $thumb,
		'description'	=> $data->{description},
        'publisher'     => $data->{publisher},
        'pubdate'       => $data->{pubdate},
		'binding'	    => $data->{binding},
		'pages'		    => $data->{pages},
		'weight'		=> $data->{weight},
		'width'		    => $data->{width},
		'height'		=> $data->{height},
		'depth'		    => $data->{depth},
        'html'          => $html
    };
    $self->book($bk);
    $self->found(1);
    return $self->book;
}

q{currently listening to: 'Drunk In Public' by The Levellers};

__END__