File Coverage

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

Criterion	Covered	Total	%
statement	13	15	86.6
branch			n/a
condition			n/a
subroutine	5	5	100.0
pod			n/a
total	18	20	90.0

line	stmt	sub	time	code
1				# ex:ts=8
2
3				package WWW::Scraper::ISBN::TWCwbook_Driver;
4
5	1	1	2433	use strict;
	1		3
	1		55
6	1	1	8	use warnings;
	1		1
	1		37
7
8	1	1	6	use vars qw($VERSION @ISA);
	1		12
	1		109
9				$VERSION = '0.02';
10
11				#--------------------------------------------------------------------------
12
13				=head1 NAME
14
15				WWW::Scraper::ISBN::TWCwbook_Driver - Search driver for TWCwbook's online catalog.
16
17				=head1 SYNOPSIS
18
19				See parent class documentation (L)
20
21				=head1 DESCRIPTION
22
23				Searches for book information from the TWCwbook's online catalog.
24
25				=cut
26
27				#--------------------------------------------------------------------------
28
29				###########################################################################
30				#Library Modules #
31				###########################################################################
32
33	1	1	6	use WWW::Scraper::ISBN::Driver;
	1		1
	1		148
34	1	1	518	use WWW::Mechanize;
	0
	0
35				use Template::Extract;
36				use Text::Iconv;
37
38				###########################################################################
39				#Constants #
40				###########################################################################
41
42				use constant QUERY => "http://www.cwbook.com.tw/search/result1.jsp?field=2&keyWord=%s";
43
44				#--------------------------------------------------------------------------
45
46				###########################################################################
47				#Inheritence #
48				###########################################################################
49
50				@ISA = qw(WWW::Scraper::ISBN::Driver);
51
52				###########################################################################
53				#Interface Functions #
54				###########################################################################
55
56				=head1 METHODS
57
58				=over 4
59
60				=item C
61
62				Creates a query string, then passes the appropriate form fields to the Cwbook
63				server.
64
65				The returned page should be the correct catalog page for that ISBN. If not the
66				function returns zero and allows the next driver in the chain to have a go. If
67				a valid page is returned, the following fields are returned via the book hash:
68
69				isbn
70				title
71				author
72				book_link
73				image_link
74				pubdate
75				publisher
76				price_list
77
78				The book_link and image_link refer back to the Cwbook website.
79
80				=back
81
82				=cut
83
84				sub search {
85				my $self = shift;
86				my $isbn = shift;
87				$self->found(0);
88				$self->book(undef);
89
90				my $url = sprintf(QUERY, $isbn);
91				my $mechanize = WWW::Mechanize->new();
92				$mechanize->get($url);
93				return undef unless($mechanize->success());
94
95				my $template = <
96
97				END
98
99				my $extract = Template::Extract->new;
100				my $data = $extract->extract($template, $mechanize->content());
101
102				return $self->handler("Could not extract data from TWCwbook result page.")
103				unless(defined $data);
104
105				$url = "http://www.cwbook.com.tw/common/book.jsp?productID=$data->{bookid}";
106				$mechanize->get($url);
107				return undef unless($mechanize->success());
108
109				$template = <
110				[% ... %]
111				[% title %] [% ... %]
112				[% ... %]
113				[% author %] [% ... %]
114				[% publisher %] [% ... %]
115				[% pubdate %] [% ... %]
116				[% price_list %]
117				END
118
119				$data = $extract->extract($template, $mechanize->content());
120				return $self->handler("Could not extract data from TWCwbook result page.")
121				unless(defined $data);
122
123				my $conv = Text::Iconv->new("utf-8", "big5");
124				$data->{title} = $conv->convert($data->{title});
125				$data->{title} =~ s/[ \r\t]//g;
126				$data->{author} = $conv->convert($data->{author});
127				$data->{author} =~ s/�@�̡G(.*)/$1/;
128				$data->{publisher} = $conv->convert($data->{publisher});
129				$data->{publisher} =~ s/�X��G(.*)/$1/;
130				$data->{pubdate} =~ s/\D(\d+\/\d+\/\d+)./$1/;
131				$data->{price_list} =~ s/\D+(\d+).*/$1/;
132
133				my $bk = {
134				'isbn' => $isbn,
135				'title' => $data->{title},
136				'author' => $data->{author},
137				'book_link' => $url,
138				'image_link' => "http://www.cwbook.com.tw".$data->{image_link},
139				'pubdate' => $data->{pubdate},
140				'publisher' => $data->{publisher},
141				'price_list' => $data->{price_list},
142				};
143
144				$self->book($bk);
145				$self->found(1);
146				return $self->book;
147				}
148
149				1;
150				__END__

File Coverage

[% title %]