Linux Certif

Toute la documentation sur la certification Linux LPI

XML::SAX::Machines.3pm

NAME

     XML::SAX::Machines - manage collections of SAX processors
 
 

SYNOPSIS

     use XML::SAX::Machines qw( :all );
 
     my $m = Pipeline(
         "My::Filter1",   ## My::Filter1 autoloaded in Pipeline()
         "My::Filter2",   ## My::Filter2     "       "      "
         \*STDOUT,        ## XML::SAX::Writer also loaded
     );
 
     $m->parse_uri( $uri ); ## A parser is autoloaded via
                            ## XML::SAX::ParserFactory if
                            ## My::Filter1 isn't a parser.
 
     ## To import only individual machines:
     use XML::SAX::Machines qw( Manifold );
 
     ## Here's a multi-pass machine that reads one document, runs
     ## it through 5 filtering channels (one channel at a time) and
     ## reassembles it in to a single document.
     my $m = Manifold(
         "My::TableOfContentsExtractor",
         "My::AbstractExtractor",
         "My::BodyFitler",
         "My::EndNotesFilter",
         "My::IndexFilter",
     );
 
     $m->parse_string( $doc );
 
 

DESCRIPTION

SAX machines are a way to gather and manage SAX processors without going nuts. Or at least without going completely nuts. Individual machines can also be like SAX processors; they don't need to parse or write anything:

    my $w = XML::SAX::Writer->new( Output => \*STDOUT );
    my $m = Pipeline( "My::Filter1", "My::Filter2", { Handler => $w } );
    my $p = XML::SAX::ParserFactory->new( handler => $p );
 
 

More documentation to come; see XML::SAX::Pipeline, XML::SAX::Manifold, and XML::SAX::Machine for now.

Here are the machines this module knows about:

     ByRecord  Record oriented processing of documents.
               L<XML::SAX::ByRecord>
 
     Machine   Generic "directed graph of SAX processors" machines.
               L<XML::SAX::Machine>
 
     Manifold  Multipass document processing
               L<XML::SAX::Manifold>
 
     Pipeline  A linear sequence of SAX processors
               L<XML::SAX::Pipeline>
 
     Tap       An insertable pass through that examines the
               events without altering them using SAX processors.
               L<XML::SAX::Tap>
 
 

Config file

As mentioned in ``LIMITATIONS'', you might occasionally need to edit the config file to tell XML::SAX::Machine how to handle a particular SAX processor (SAX processors use a wide variety of API conventions).

The config file is a the Perl module XML::SAX::Machines::SiteConfig, which contains a Perl data structure like:

     package XML::SAX::Machines::SiteConfig;
 
     $ProcessorClassOptions = {
         "XML::Filter::Tee" => {
             ConstructWithHashedOptions => 1,
         },
     };
 
 

So far $Processors is the only available configuration structure. It contains a list of SAX processors with known special needs.

Also, so far the only special need is the ConstructWithHashes option which tells XML::SAX::Machine to construct such classes like:

     XML::Filter::Tee->new(
         { Handler => $h }
     );
 
 

instead of

     XML::Filter::Tee->new( Handler => $h );
 
 

WARNING If you modify anything, apply your changes in a new file created from XML::SAX::Machines::SiteConfig.pm. On Debian systems, this should be placed in /etc/perl so that it is not overwritten during upgrade. Do not alter XML::SAX::Machines::ConfigDefaults.pm or you will lose your changes when you upgrade.

TODO: Allow per-app and per-machine overrides of options. When needed.

AUTHORS

Barrie Slaymaker

LICENCE

Copyright 2002-2009 by Barrie Slaymaker.

This software is free. It is licensed under the same terms as Perl itself.