NAME

XML::RSS::Parser - A liberal object-oriented parser for RSS feeds.

SYNOPSIS

 #!/usr/bin/perl -w
 use strict;
 
 use XML::RSS::Parser;
 use FileHandle;
 
 my $p = XML::RSS::Parser->new;
 my $fh = FileHandle->new('/path/to/some/rss/file');
 my $feed = $p->parse_file($fh);
 
 # output some values 
 my $feed_title = $feed->query('/channel/title');
 print $feed_title->text_content;
 my $count = $feed->item_count;
 print " ($count)\n";
 foreach my $i ( $feed->query('//item') ) { 
     my $node = $i->query('title');
     print '  '.$node->text_content;
     print "\n"; 
 }

DESCRIPTION

XML::RSS::Parser is a lightweight liberal parser of RSS feeds. This parser is "liberal" in that it does not demand compliance of a specific RSS version and will attempt to gracefully handle tags it does not expect or understand. The parser's only requirements is that the file is well-formed XML and remotely resembles RSS. Roughly speaking, well formed XML with a "channel" element as a direct sibling or the root tag and "item" elements etc.

There are a number of advantages to using this module then just using a standard parser-tree combination. There are a number of different RSS formats in use today. In very subtle ways these formats are not entirely compatible from one to another. XML::RSS::Parser makes a couple assumptions to "normalize" the parse tree into a more consistent form. For instance, it forces "channel" and "item" into a parent-child relationship. For more detail see "SPECIAL PROCESSING NOTES".

This module is leaner then XML::RSS -- the majority of code was for generating RSS files. It also provides a XPath-esque interface to the feed's tree.

While XML::RSS::Parser creates a normalized parse tree, it still leaves the mapping of overlapping and alternate tags common in the RSS format space to the developer. For this look at the XML::RAI (RSS Abstraction Interface) package which provides an object-oriented layer to XML::RSS::Parser trees that transparently maps these various tags to one common interface.

XML::RSS::Parser is based on XML::Elemental, a a SAX-based package for easily parsing XML documents into a more native and mostly object-oriented perl form.

SPECIAL PROCESSING NOTES

There are a number of different RSS formats in use today. In very subtle ways these formats are not entirely compatible from one to another. What's worse is that there are unlabeled versions within the standard in addition to tags with overlapping purposes and vague definitions. (See Mark Pilgrim's "The myth of RSS compatibility" "/diveintomark.org/archives/2004/02/04/incompatible- rss" in http: for just a sampling of what I mean.) To ease working with RSS data in different formats, the parser does not create the feed's parse tree verbatim. Instead it makes a few assumptions to "normalize" the parse tree into a more consistent form.

With the refactoring of this module and the switch to a true tree structure, the normalization process has been simplified. Some of the version 2x proved to be problematic with more advanced and complex feeds.

The RSS namespace (if any) is extracted from the first sibling of the root tag. We don't use the root tag because in RSS 1.0 the root tag is in the RDF namespace and not RSS. That namespace is treated as the '#default' (no prefix) namespace for the parse tree.
The parser will not include the root tags of "rss" or "RDF" in the tree. Namespace declaration information is still extracted.
The parser forces "channel" and "item" into a parent-child relationship. In versions 0.9 and 1.0, "channel" and "item" tags are siblings.

Two significant changes were made with the release of version 4.0.

XML::RSS::Parser is not a subclass of XML::Elemental.: This change should be transparent in most cases, but deemed necessary for the error handling and special handling of RSS data.
XML::RSS::Parser uses Clarkian Notation for element and attribute names.: This change is inherited from recent changes in XML::Elemental. The previous system was flawed and not widely adopted. Clarkian notation is the form used by XML::SAX and XML::Simple to name a few. Use the "process_name" in XML::Elemental::Util to parse element and attribute names intoo their namespace URI and local name parts.

NAMESPACE PREFIXES

The following prefix and namespace combinations are recognized by default. Use "register_ns_prefix" to add more as needed.

    admin       http://webns.net/mvcb/
    ag          http://purl.org/rss/1.0/modules/aggregation/
    annotate    http://purl.org/rss/1.0/modules/annotate/
    atom        http://www.w3.org/2005/Atom
    audio       http://media.tangent.org/rss/1.0/
    cc          http://web.resource.org/cc/
    company     http://purl.org/rss/1.0/modules/company
    content     http://purl.org/rss/1.0/modules/content/
    cp          http://my.theinfo.org/changed/1.0/rss/
    dc          http://purl.org/dc/elements/1.1/
    dcterms     http://purl.org/dc/terms/
    email       http://purl.org/rss/1.0/modules/email/
    ev          http://purl.org/rss/1.0/modules/event/
    feedburner  http://rssnamespace.org/feedburner/ext/1.0
    foaf        http://xmlns.com/foaf/0.1/
    image       http://purl.org/rss/1.0/modules/image/
    itunes      http://www.itunes.com/DTDs/Podcast-1.0.dtd
    l           http://purl.org/rss/1.0/modules/link/
    openSearch  http://a9.com/-/spec/opensearchrss/1.0/
    rdf         http://www.w3.org/1999/02/22-rdf-syntax-ns#
    rdfs        http://www.w3.org/2000/01/rdf-schema#
    ref         http://purl.org/rss/1.0/modules/reference/
    reqv        http://purl.org/rss/1.0/modules/richequiv/
    rss091      http://purl.org/rss/1.0/modules/rss091#
    search      http://purl.org/rss/1.0/modules/search/
    slash       http://purl.org/rss/1.0/modules/slash/
    ss          http://purl.org/rss/1.0/modules/servicestatus/
    str         http://hacks.benhammersley.com/rss/streaming/
    sub         http://purl.org/rss/1.0/modules/subscription/
    sy          http://purl.org/rss/1.0/modules/syndication/
    tapi        http://api.technorati.com/dtd/tapi-001.xml#
    taxo        http://purl.org/rss/1.0/modules/taxonomy/
    thr         http://purl.org/rss/1.0/modules/threading/
    trackback   http://madskills.com/public/xml/rss/module/trackback/
    wiki        http://purl.org/rss/1.0/modules/wiki/
    xhtml       http://www.w3.org/1999/xhtml
    xml         http://www.w3.org/XML/1998/namespace/

    creativeCommons  http://backend.userland.com/creativeCommonsRssModule

METHODS

The following objects and methods are provided in this package.

XML::RSS::Parser->new

Constructor. Returns a reference to a new XML::RSS::Parser object.

$parser->parse =item $parser->parse_file =item $parser->parse_string =item $parser->parse_uri

These methods are mostly pass-thru to the underlying SAX parser provided by XML::Elemental. (See XML::SAX::Base for more.)

XML::RSS::Parser wraps these calls in eval statements and rather then dying returns undefined. Any parsing errors can be retreived by using the "errstr" method inherited from Class::ErrorHandler.

Once the markup has been parsed it is automatically passed through the "rss_normalize" method before the parse tree is returned to the caller.

XML::RSS::Parser->register_ns_prefix(prefix,curi)

Registers the given path with namespace URI for XPath lookups. Both parameters are required.

XML::RSS::Parser->ns_qualify(element, namespace_uri)

An simple utility implemented as an abstract method that will return a fully namespace qualified string for the supplied element. Return values are now in Clarkian notation.

XML::RSS::Parser->prefix(namespace_uri)

Returns the prefix to the given namespace URI. Returns "undef" if the prefix is not known.

XML::RSS::Parser->namespace(prefix)

Returns the namespace URI to the given prefix. Returns "undef" if the namespace is not registered.

error

Sets an error message for later retreival and returns "undef". Inherited from Class::ErrorHandler.

errstr

Returns the last error message set by "error". Inherited from Class:ErrorHandler.

DEPENDENCIES

XML::SAX, XML::Elemental, Class::ErrorHandler, Class::XPath 1.4*

Versions up to 1.4 have a design flaw that would cause it to choke on feeds with the / character in an attribute value. For example the Yahoo! feeds.

AUTHOR & COPYRIGHT

POD ERRORS

Hey! The above document had some coding errors, which are explained below:

Around line 127:: =begin without a target?
Around line 310:: '=item' outside of any '=over'
Around line 364:: You forgot a '=back' before '=head1'
Around line 390:: =back without =over
Around line 400:: '=end' without a target?