File Coverage

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

Criterion	Covered	Total	%
statement	75	80	93.7
branch	16	36	44.4
condition	8	18	44.4
subroutine	7	7	100.0
pod	1	1	100.0
total	107	142	75.3

package WWW::Scraper::ISBN::Booktopia_Driver;

use strict;
use warnings;

use vars qw($VERSION @ISA);
$VERSION = '0.21';

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

=head1 NAME

WWW::Scraper::ISBN::Booktopia_Driver - Search driver for Booktopia online book catalog.

=head1 SYNOPSIS

See parent class documentation (L)

=head1 DESCRIPTION

Searches for book information from Booktopia online book catalog

=cut

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

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

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

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

use WWW::Mechanize;

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

use constant	SEARCH	=> 'http://www.booktopia.com.au/search.ep?cID=&submit.x=44&submit.y=7&submit=search&keywords=';
my ($BAU_URL1,$BAU_URL2,$BAU_URL3) = ('http://www.booktopia.com.au','/[^/]+/prod','.html');

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

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

=head1 METHODS

=over 4

=item C

Creates a query string, then passes the appropriate form fields to the
Booktopia 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)
  depth         (if known) (in millimetres)

The book_link and image_link refer back to the Booktopia 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)));

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

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

#print STDERR "\n# url=[".(SEARCH . $isbn)."]\n";

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

    my $pattern = $isbn;
    if(length $isbn == 10) {
        $pattern = '978' . $isbn;
        $pattern =~ s/.$/./;
    }

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

	return $self->handler("Failed to find that book on Booktopia website. [$isbn]")
		if($html =~ m!Sorry, we couldn't find any matches for!si);

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

    my $data;
    ($data->{publisher})                = $html =~ m!\s*Publisher:\s*\s*([^<]+)!si;
    ($data->{pubdate})                  = $html =~ m!\s*Published:\s*\s*([^<]+)!si;

    $data->{publisher} =~ s!<[^>]+>!!g  if($data->{publisher});
    $data->{pubdate} =~ s!\s+! !g       if($data->{pubdate});

    ($data->{image})                    = $html =~ m!(http://covers.booktopia.com.au/big/\d+/[-\w]+\.jpg)!si;
    ($data->{thumb})                    = $html =~ m!(http://covers.booktopia.com.au/\d+/\d+/[-\w]+\.jpg)!si;
    ($data->{isbn13})                   = $html =~ m!\s*ISBN:\s*\s*(\d+)!si;
    ($data->{isbn10})                   = $html =~ m!\s*ISBN-10:\s*\s*(\d+)!si;
    ($data->{author})                   = $html =~ m!\s*(?:By|Author):\s*(.*?)!si;
    ($data->{title})                    = $html =~ m!
    ($data->{title})                    = $html =~ m!([^<]+)!si  unless($data->{title});
    ($data->{description})              = $html =~ m!(.*?)!si;
    ($data->{binding})                  = $html =~ m!\s*Format:\s*\s*([^<]+)!si;
    ($data->{pages})                    = $html =~ m!\s*Number Of Pages:\s*\s*([\d.]+)!si;
    ($data->{weight})                   = $html =~ m!\s*Weight \(kg\):\s*\s*([\d.]+)!si;
    ($data->{height},$data->{width},$data->{depth})
                                        = $html =~ m!\s*Dimensions \(cm\):\s*([\d.]+)\s* x \s*([\d.]+)\s* x \s*([\d.]+)!si;

    # despite it saying Kg (kilogrammes) the weight seems to vary between widely!
    if($data->{weight}) {
        if(   $data->{weight} < 1)      {$data->{weight} = int($data->{weight} * 1000)}
        elsif($data->{weight} < 100)    {$data->{weight} = int($data->{weight} * 10)}
        elsif($data->{weight} < 1000)   {$data->{weight} = int($data->{weight})}
        else                            {$data->{weight} = int($data->{weight})}
    }

    $data->{height} = int($data->{height} * 10)     if($data->{height});
    $data->{width}  = int($data->{width}  * 10)     if($data->{width});
    $data->{depth}  = int($data->{depth}  * 10)     if($data->{depth});

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

    if($data->{description}) {
        $data->{description} =~ s!Click on the Google Preview[^<]+!!s;
        $data->{description} =~ s!!\n!gi;
        $data->{description} =~ s!<[^>]+>!!g;
    }

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

	return $self->handler("Could not extract data from Booktopia 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'		=> $mech->uri(),
		'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},
		'depth'		    => $data->{depth},
        'html'          => $html
	};

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

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

1;

__END__