| 1 | package HTML::Filter; |
| 2 | |
| 3 | require HTML::Parser; |
| 4 | @ISA=qw(HTML::Parser); |
| 5 | |
| 6 | $VERSION = sprintf("%d.%02d", q$Revision: 2.9 $ =~ /(\d+)\.(\d+)/); |
| 7 | |
| 8 | sub declaration { $_[0]->output("<!$_[1]>") } |
| 9 | sub process { $_[0]->output($_[2]) } |
| 10 | sub comment { $_[0]->output("<!--$_[1]-->") } |
| 11 | sub start { $_[0]->output($_[4]) } |
| 12 | sub end { $_[0]->output($_[2]) } |
| 13 | sub text { $_[0]->output($_[1]) } |
| 14 | |
| 15 | sub output { print $_[1] } |
| 16 | |
| 17 | 1; |
| 18 | |
| 19 | __END__ |
| 20 | |
| 21 | =head1 NAME |
| 22 | |
| 23 | HTML::Filter - Filter HTML text through the parser |
| 24 | |
| 25 | =head1 NOTE |
| 26 | |
| 27 | This module is deprecated. C<HTML::Parser> now provides the |
| 28 | functionally of C<HTML::Filter> much more efficiently with the the |
| 29 | C<default> handler. |
| 30 | |
| 31 | =head1 SYNOPSIS |
| 32 | |
| 33 | require HTML::Filter; |
| 34 | $p = HTML::Filter->new->parse_file("index.html"); |
| 35 | |
| 36 | =head1 DESCRIPTION |
| 37 | |
| 38 | C<HTML::Filter> is an HTML parser that by default prints the |
| 39 | original text of each HTML element (a slow version of cat(1) basically). |
| 40 | The callback methods may be overridden to modify the filtering for some |
| 41 | HTML elements and you can override output() method which is called to |
| 42 | print the HTML text. |
| 43 | |
| 44 | C<HTML::Filter> is a subclass of C<HTML::Parser>. This means that |
| 45 | the document should be given to the parser by calling the $p->parse() |
| 46 | or $p->parse_file() methods. |
| 47 | |
| 48 | =head1 EXAMPLES |
| 49 | |
| 50 | The first example is a filter that will remove all comments from an |
| 51 | HTML file. This is achieved by simply overriding the comment method |
| 52 | to do nothing. |
| 53 | |
| 54 | package CommentStripper; |
| 55 | require HTML::Filter; |
| 56 | @ISA=qw(HTML::Filter); |
| 57 | sub comment { } # ignore comments |
| 58 | |
| 59 | The second example shows a filter that will remove any E<lt>TABLE>s |
| 60 | found in the HTML file. We specialize the start() and end() methods |
| 61 | to count table tags and then make output not happen when inside a |
| 62 | table. |
| 63 | |
| 64 | package TableStripper; |
| 65 | require HTML::Filter; |
| 66 | @ISA=qw(HTML::Filter); |
| 67 | sub start |
| 68 | { |
| 69 | my $self = shift; |
| 70 | $self->{table_seen}++ if $_[0] eq "table"; |
| 71 | $self->SUPER::start(@_); |
| 72 | } |
| 73 | |
| 74 | sub end |
| 75 | { |
| 76 | my $self = shift; |
| 77 | $self->SUPER::end(@_); |
| 78 | $self->{table_seen}-- if $_[0] eq "table"; |
| 79 | } |
| 80 | |
| 81 | sub output |
| 82 | { |
| 83 | my $self = shift; |
| 84 | unless ($self->{table_seen}) { |
| 85 | $self->SUPER::output(@_); |
| 86 | } |
| 87 | } |
| 88 | |
| 89 | If you want to collect the parsed text internally you might want to do |
| 90 | something like this: |
| 91 | |
| 92 | package FilterIntoString; |
| 93 | require HTML::Filter; |
| 94 | @ISA=qw(HTML::Filter); |
| 95 | sub output { push(@{$_[0]->{fhtml}}, $_[1]) } |
| 96 | sub filtered_html { join("", @{$_[0]->{fhtml}}) } |
| 97 | |
| 98 | =head1 SEE ALSO |
| 99 | |
| 100 | L<HTML::Parser> |
| 101 | |
| 102 | =head1 COPYRIGHT |
| 103 | |
| 104 | Copyright 1997-1999 Gisle Aas. |
| 105 | |
| 106 | This library is free software; you can redistribute it and/or |
| 107 | modify it under the same terms as Perl itself. |
| 108 | |
| 109 | =cut |