NAME

DOMHandler - Implements a call-back interface to DOM.

SYNOPSIS

use DOMHandler;
use XML::LibXML;
$p = new XML::LibXML;
$doc = $p->parse_file( 'data.xml' );
$dh = new DOMHandler( handler_package => new testhandler );
$dh->traverse( $doc );

package testhandler;
sub new {
    return bless {};
}
sub A {
    my( $self, $agent, $node ) = @_;
    my $par = $node->parentNode->nodeName;
    print "I'm in an A element and my parent is $par.\n";
}
sub generic_element {
    my( $self, $agent, $node ) = @_;
    my $name = $node->nodeName;
    print "I'm in an element named '$name'.\n";
}
sub generic_text {
    print "Here's some text.\n";
}
sub generic_PI {
    print "Here's a processing instruction.\n";
}
sub generic_CDATA {
    print "Here's a CDATA Section.\n";
}

DESCRIPTION

This module creates a layer on top of DOM that allows you to program in a "push" style rather than "pull". Once the document has been parsed and you have a DOM object, you can call on the DOMHandler's traverse() method to apply a set of call-back routines to all the nodes in a tree. You supply the routines in a handler package when initializing the DOMHandler.

In your handler package, the names of routines determine which will be called for a given node. There are routines for node types, named "generic_" plus the node type. For elements, you can name routines after the element name and these will only be called for that type of element. A list of supported handlers follows:

else_generic_node()

Applied only to nodes that have not been handled by another routine.

generic_CDATA()

Applied to CDATA sections.

generic_comment()

Applied to XML comments.

generic_doctype()

Applied to DOCTYPE declarations.

generic_element()

Applied to all elements.

generic_node()

Applied to all nodes.

generic_PI()

Processing instruction

generic_text()

Applied to text nodes.

A handler routine takes three arguments: the $self reference, a reference to the DOMHandler object, and a reference to a node in the document being traversed. You can use DOM routines on that node to do any processing you want. At the moment, this module only supports XML::LibXML documents.

IMPORTANT NOTE: Some DOM operations may cause unwanted results. For example, if you delete the current node's parent, the program will likely crash.

METHODS

traverse( $doc )

Visits each node in a document, in order, applying the appropriate handler routines.

AUTHOR

Erik Ray (eray@oreilly.com), Production Tools Dept., O'Reilly and Associates Inc.

COPYRIGHT

Copyright (c) 2002 Erik Ray and O'Reilly & Associates.