HTML::Truncate(3) (beta software) truncate HTML by percentage or character count while preserving well-formedness.

VERSION

0.20

ABSTRACT

When working with text it is common to want to truncate strings to make them fit a desired context. E.g., you might have a menu that is only 100px wide and prefer text doesn't wrap so you'd truncate it around 15-30 characters, depending on preference and typeface size. This is trivial with plain text using substr but with HTML it is somewhat difficult because whitespace has fluid significance and open tags that are not properly closed destroy well-formedness and can wreck an entire layout.

HTML::Truncate attempts to account for those two problems by padding truncation for spacing and entities and closing any tags that remain open at the point of truncation.

SYNOPSIS


use strict;
use HTML::Truncate;
my $html = '<p><i>We</i> have to test <b>something</b>.</p>';
my $readmore = '... <a href="/full-article">[readmore]</a>';
my $html_truncate = HTML::Truncate->new();
$html_truncate->chars(20);
$html_truncate->ellipsis($readmore);
print $html_truncate->truncate($html);
# or
use Encode;
my $ht = HTML::Truncate->new( utf8_mode => 1,
chars => 1_000,
);
print Encode::encode_utf8( $ht->truncate($html) );

XHTML

This module is designed to work with XHTML-style nested tags. More below.

WHITESPACE AND ENTITIES

Repeated natural whitespace (i.e., ``\s+'' and not `` &nbsp; '') in HTML --- with rare exception (pre tags or user defined styles) --- is not meaningful. Therefore it is normalized when truncating. Entities are also normalized. The following is only counted 14 chars long.

  \n<p>\nthis     is   &#8216;text&#8217;\n\n</p>
  ^^^^^^^12345----678--9------01234------^^^^^^^^

METHODS

new
Can take all the methods as hash style args. ``percent'' and ``chars'' are incompatible so don't use them both. Whichever is set most recently will erase the other.

 my $ht = HTML::Truncate->new(utf8_mode => 1,
                              chars => 500, # default is 100
                              );
utf8_mode
Set/get, true/false. If "utf8_mode" is set, utf8_mode(1) is also set in the underlying HTML::Parser, entities will be transformed with decode and the default ellipsis will be a literal ellipsis and not the default of "&#8230;".
chars
Set/get. The number of characters remaining after truncation, excluding the ``ellipsis''.

Entities are counted as single characters. E.g., "&copy;" is one character for truncation counts.

Default is ``100.'' Side-effect: clears any ``percent'' that has been set.

percent
Set/get. A percentage to keep while truncating the rest. For a document of 1,000 chars, percent('15%') and chars(150) would be equivalent. The actual amount of character that the percent represents cannot be known until the given HTML is parsed.

Side-effect: clears any ``chars'' that has been set.

ellipsis
Set/get. Ellipsis in this case means --

    The omission of a word or phrase necessary for a complete
    syntactical construction but not necessary for understanding.
                         http://www.answers.com/topic/ellipsis

What it will probably mean in most real applications is ``read more.'' The default is "&#8230;" which if the utf8 flag is true will render as a literal ellipsis, "chr(8230)".

The reason the default is "&#8230;" and not ``...'' is this is meant for use in HTML environments, not plain text, and ``...'' (dot-dot-dot) is not typographically correct or equivalent to a real horizontal ellipsis character.

truncate
It returns the truncated XHTML if asked for a return value.

 my $truncated = $ht->truncate($html);

It will truncate the string in place if no return value is expected (wantarray is not defined).

   $ht->truncate($html);
   print $html;

Also can be called with inline arguments-

   print $ht->truncate( $html,
                        $chars_or_percent,
                        $ellipsis );

No arguments are strictly required. Without HTML to operate upon it returns undef. The two optional arguments may be preset with the methods ``chars'' (or ``percent'') and ``ellipsis''.

Valid nesting of tags is required (alla XHTML). Therefore some old HTML habits like <p> without a </p> are not supported and may cause a fatal error. See ``repair'' for help with badly formed HTML.

Certain tags are omitted by default from the truncated output.

  • Skipped tags

    These will not be included in truncated output by default.

       <head>...</head> <script>...</script> <form>...</form>
       <iframe></iframe> <title>...</title> <style>...</style>
       <base/> <link/> <meta/>
    
  • Tags allowed to self-close

    See emptyElement in HTML::Tagset.

add_skip_tags( qw( tag list ) )
Put one or more new tags into the list of those to be omitted from truncated output. An example of when you might like to use this is if you're thumb-nailing articles and they start with "<h1>title</h1>" or such before the article body. The heading level would be absurd with a list of excerpts so you could drop it completely this way---

 $ht->add_skip_tags( 'h1' );
dont_skip_tags( qw( tag list ) )
Takes tags out of the current list to be omitted from truncated output.
repair
Set/get, true/false. If true, will attempt to repair unclosed HTML tags by adding close-tags as late as possible (eg. "<i><b>foobar</i>" becomes "<i><b>foobar</b></i>"). Unmatched close tags are dropped ("foobar</b>" becomes "foobar").
on_space
This will make the truncation back up to the first space it finds so it doesn't truncate in the the middle of a word. ``on_space'' runs before ``cleanly'' if both are set.
cleanly
Set/get --- a regular expression. This is on by default and the default cleaning regular expression is "cleanly(qr/[\s[:punct:]]+\z/)". It will make the truncation strip any trailing spacing and punctuation so you don't get things like ``The End....'' or ``What? ...'' You can cancel it with "$ht->cleanly(undef)" or provide your own regular expression.

COOKBOOK (well, a recipe)

Template Toolkit filter

For excerpting HTML in your Templates. Note the ``add_skip_tags'' which is set to drop any images from the truncated output.

 use Template;
 use HTML::Truncate;
 my %config =
    (
     FILTERS => {
         truncate_html => [ \&truncate_html_filter_factory, 1 ],
     },
     );
 my $tt = Template->new(\%config) or die $Template::ERROR;
 # ... etc ...
 sub truncate_html_filter_factory {
     my ( $context, $len, $ellipsis ) = @_;
     $len = 32 unless $len;
     $ellipsis = chr(8230) unless defined $ellipsis;
     my $ht = HTML::Truncate->new();
     $ht->add_skip_tags(qw( img ));
     return sub {
         my $html = shift || return '';
         return $ht->truncate( $html, $len, $ellipsis );
     }
 }

Then in your templates you can do things like this:

 [% FOR item IN search_results %]
   <div class="searchResult">
     <a href="[% item.uri %]">[% item.title %]</a><br />
     [% item.body | truncate_html(200) %]
   </div>
 [% END %]

See also Template::Filters.

AUTHOR

Ashley Pond V, "<[email protected]>".

LIMITATIONS

There may be places where this will break down right now. I'll pad out possible edge cases as I find them or they are sent to me via the CPAN bug ticket system.

This is not an HTML filter

Although this happens to do some crude HTML filtering to achieve its end, it is not a fully featured filter. If you are looking for one, check out HTML::Scrubber and HTML::Sanitizer.

BUGS, FEEDBACK, PATCHES

Please report any bugs or feature requests to "[email protected]", or through the web interface at <http://rt.cpan.org/NoAuth/ReportBug.html?Queue=HTML-Truncate>. I will get the ticket, and then you'll automatically be notified of progress as I make changes.

TO DO

Write a couple more tests (percent and skip stuff) then take out beta notice. Try to make the 5.6 stuff work without decode...? Try a "drop_tags" method?

Write an XML::LibXML based version to load when possible...? Or make that part of XHTML::Util?

THANKS TO

Kevin Riggle for the ``repair'' functionality; patch, Pod, and tests.

Lorenzo Iannuzzi for the ``on_space'' functionality.

COPYRIGHT & LICENSE

Copyright (X) 2005-2009 Ashley Pond V.

This program is free software; you can redistribute it or modify it or both under the same terms as Perl itself.

DISCLAIMER OF WARRANTY

Because this software is licensed free of charge, there is no warranty for the software, to the extent permitted by applicable law. Except when otherwise stated in writing the copyright holders or other parties provide the software ``as is'' without warranty of any kind, either expressed or implied, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. The entire risk as to the quality and performance of the software is with you. Should the software prove defective, you assume the cost of all necessary servicing, repair, or correction.

In no event unless required by applicable law or agreed to in writing will any copyright holder, or any other party who may modify and/or redistribute the software as permitted by the above licence, be liable to you for damages, including any general, special, incidental, or consequential damages arising out of the use or inability to use the software (including but not limited to loss of data or data being rendered inaccurate or losses sustained by you or third parties or a failure of the software to operate with any other software), even if such holder or other party has been advised of the possibility of such damages.