blib/lib/HTML/LinkExtor.pm

filter: Statementstmt 93.0% 40 / 43 Branchbran 92.8% 13 / 14 Conditioncond n/a MC/DCmcdc n/a Subroutinesub 85.7% 6 / 7 Podpod 100% 3 / 3 totaltotal 92.5% 62 / 67 CC 9 SCAR 12.4
CC
9
Cov
93%
_start_tag
6.0
new
2.0
_found_link
2.0
CRAP 3.5
« blib/lib/HTML/Entities.pm ↑ Summary hparser.c »
1 package HTML::LinkExtor;
2  
3 2 require HTML::Parser;
4 2 our @ISA = qw(HTML::Parser);
5 2 our $VERSION = '3.87';
6  
7 =head1 NAME
8  
9 HTML::LinkExtor - Extract links from an HTML document
10  
11 =head1 SYNOPSIS
12  
13 require HTML::LinkExtor;
14 $p = HTML::LinkExtor->new(\&cb, "http://www.perl.org/");
15 sub cb {
16 my($tag, %links) = @_;
17 print "$tag @{[%links]}\n";
18 }
19 $p->parse_file("index.html");
20  
21 =head1 DESCRIPTION
22  
23 I<HTML::LinkExtor> is an HTML parser that extracts links from an
24 HTML document. The I<HTML::LinkExtor> is a subclass of
25 I<HTML::Parser>. This means that the document should be given to the
26 parser by calling the $p->parse() or $p->parse_file() methods.
27  
28 =cut
29  
30 2 ▶use strict;
Subroutine BEGIN: called CC=1 SCAR=0.0
31 2 ▶use HTML::Tagset ();
Subroutine BEGIN: called CC=1 SCAR=0.0
32  
33 # legacy (some applications grabs this hash directly)
34 2 our %LINK_ELEMENT;
35 2 *LINK_ELEMENT = \%HTML::Tagset::linkElements;
36  
37 =over 4
38  
39 =item $p = HTML::LinkExtor->new
40  
41 =item $p = HTML::LinkExtor->new( $callback )
42  
43 =item $p = HTML::LinkExtor->new( $callback, $base )
44  
45 The constructor takes two optional arguments. The first is a reference
46 to a callback routine. It will be called as links are found. If a
47 callback is not provided, then links are just accumulated internally
48 and can be retrieved by calling the $p->links() method.
49  
50 The $base argument is an optional base URL used to absolutize all URLs found.
51 You need to have the I<URI> module installed if you provide $base.
52  
53 The callback is called with the lowercase tag name as first argument,
54 and then all link attributes as separate key/value pairs. All
55 non-link attributes are removed.
56  
57 =cut
58  
59 sub new
60 3 ▶{
Subroutine new: called CC=2 SCAR=6.9
Pod documented
61 3 my($class, $cb, $base) = @_;
62 3 my $self = $class->SUPER::new(
63 start_h => ["_start_tag", "self,tagname,attr"],
64 report_tags => [keys %HTML::Tagset::linkElements],
65 );
66 3 $self->{extractlink_cb} = $cb;
67 3 ▶ if ($base) {
Branch: 1 branch
BranchTrueFalseTotal
if ($base) 1 2 3
68 1 require URI;
69 1 $self->{extractlink_base} = URI->new($base);
70 }
71 3 $self;
72 }
73  
74 sub _start_tag
75 11 ▶{
Subroutine _start_tag: called CC=6 SCAR=17.9
76 11 my($self, $tag, $attr) = @_;
77  
78 11 my $base = $self->{extractlink_base};
79 11 my $links = $HTML::Tagset::linkElements{$tag};
80 11 ▶ $links = [$links] unless ref $links;
Branch: 1 branch
BranchTrueFalseTotal
unless ref $links 0 11 11
81  
82 11 my @links;
83 my $a;
84 11 for $a (@$links) {
85 17 ▶ next unless exists $attr->{$a};
Branch: 1 branch
BranchTrueFalseTotal
unless exists $attr->{$a} 10 7 17
86 7 (my $link = $attr->{$a}) =~ s/^\s+//; $link =~ s/\s+$//; # HTML5
87 7 ▶ push(@links, $a, $base ? URI->new($link, $base)->abs($base) : $link);
Branch: 1 branch
BranchTrueFalseTotal
$base ? : 4 3 7
88 }
89 11 ▶ return unless @links;
Branch: 1 branch
BranchTrueFalseTotal
unless @links 4 7 11
90 7 $self->_found_link($tag, @links);
91 }
92  
93 sub _found_link
94 7 ▶{
Subroutine _found_link: called CC=2 SCAR=6.9
95 7 my $self = shift;
96 7 my $cb = $self->{extractlink_cb};
97 7 ▶ if ($cb) {
Branch: 1 branch
BranchTrueFalseTotal
if ($cb) { } 3 4 7
98 3 &$cb(@_);
99 } else {
100 4 push(@{$self->{'links'}}, [@_]);
101 }
102 }
103  
104 =item $p->links
105  
106 Returns a list of all links found in the document. The returned
107 values will be anonymous arrays with the following elements:
108  
109 [$tag, $attr => $url1, $attr2 => $url2,...]
110  
111 The $p->links method will also truncate the internal link list. This
112 means that if the method is called twice without any parsing
113 between them the second call will return an empty list.
114  
115 Also note that $p->links will always be empty if a callback routine
116 was provided when the I<HTML::LinkExtor> was created.
117  
118 =cut
119  
120 sub links
121 2 ▶{
Subroutine links: called CC=2 SCAR=6.9
Pod documented
122 2 my $self = shift;
123 2 ▶ exists($self->{'links'}) ? @{delete $self->{'links'}} : ();
Branch: 1 branch
BranchTrueFalseTotal
exists $self->{'links'} ? : 1 1 2
124 }
125  
126 # We override the parse_file() method so that we can clear the links
127 # before we start a new file.
128 sub parse_file
129 0 ▶{
Subroutine parse_file: not called CC=1 SCAR=6.9
Pod documented
130 0 my $self = shift;
131 0 delete $self->{'links'};
132 0 $self->SUPER::parse_file(@_);
133 }
134  
135 =back
136  
137 =head1 EXAMPLE
138  
139 This is an example showing how you can extract links from a document
140 received using LWP:
141  
142 use LWP::UserAgent;
143 use HTML::LinkExtor;
144 use URI::URL;
145  
146 $url = "http://www.perl.org/"; # for instance
147 $ua = LWP::UserAgent->new;
148  
149 # Set up a callback that collect image links
150 my @imgs = ();
151 sub callback {
152 my($tag, %attr) = @_;
153 return if $tag ne 'img'; # we only look closer at <img ...>
154 push(@imgs, values %attr);
155 }
156  
157 # Make the parser. Unfortunately, we don't know the base yet
158 # (it might be different from $url)
159 $p = HTML::LinkExtor->new(\&callback);
160  
161 # Request document and parse it as it arrives
162 $res = $ua->request(HTTP::Request->new(GET => $url),
163 sub {$p->parse($_[0])});
164  
165 # Expand all image URLs to absolute ones
166 my $base = $res->base;
167 @imgs = map { $_ = url($_, $base)->abs; } @imgs;
168  
169 # Print them out
170 print join("\n", @imgs), "\n";
171  
172 =head1 SEE ALSO
173  
174 L<HTML::Parser>, L<HTML::Tagset>, L<LWP>, L<URI::URL>
175  
176 =head1 COPYRIGHT
177  
178 Copyright 1996-2001 Gisle Aas.
179  
180 This library is free software; you can redistribute it and/or
181 modify it under the same terms as Perl itself.
182  
183 =cut
184  
185 2 1;
« blib/lib/HTML/Entities.pm ↑ Summary hparser.c »