Provided by: libhtml-truncate-perl_0.20-2_all bug


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




       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.


        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();
        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) );


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


       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>


       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

           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;".

           Set/get. The number of characters remaining after truncation, excluding the

           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.

           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.

           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.

           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,

           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.

           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

              print $html;

           Also can be called with inline arguments-
              print $ht->truncate( $html,
                                   $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

           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.

           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

           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/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) %]
        [% END %]

       See also Template::Filters.


       Ashley Pond V, "<>".


       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


       Please report any bugs or feature requests to "", or through
       the web interface at <>. I
       will get the ticket, and then you'll automatically be notified of progress as I make

   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


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

       Lorenzo Iannuzzi for the "on_space" functionality.


       HTML::Entities, HTML::TokeParser, the "truncate" filter in Template, and Text::Truncate.

       HTML::Scrubber and HTML::Sanitizer.


       Copyright (©) 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.


       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.