=head1 NAME

HTML::Microformats::Documentation::Notes - misc usage and design notes

=head1 NOTES

=head2 Byzantine Internals

The internals of HTML::Microformats are pretty complicated - best to steer clear
of them. Here are three usage patterns that mostly avoid dealing with the internals:

=over

=item * B<Parse a page and use it as a single RDF graph.>

A page can be parsed into an L<RDF::Trine::Model> and queried
using SPARQL.

	use HTML::Microformats;
	use LWP::Simple qw[get];
	use RDF::Query;
	
	my $page  = 'http://example.net/';
	my $graph = HTML::Microformats
	               ->new_document(get($page), $page)
	               ->assume_all_profiles
	               ->parse_microformats
	               ->model;
	
	my $query = RDF::Query->new(<<SPARQL);
	PREFIX foaf: <http://xmlns.com/foaf/0.1/>
	SELECT DISTINCT ?friendname ?friendpage
	WHERE {
		<$page> ?p ?friendpage .
		?person foaf:name ?friendname ;
			foaf:page ?friendpage .
		FILTER (
			isURI(?friendpage)
			&& isLiteral(?friendname) 
			&& regex(str(?p), "^http://vocab.sindice.com/xfn#(.+)-hyperlink")
		)
	}
	SPARQL
	
	my $results = $query->execute($graph);
	while (my $result = $results->next)
	{
		printf("%s <%s>\n",
			$result->{friendname}->literal_value,
			$result->{friendpage}->uri,
			);
	}

=item * B<Use the data method on each object.>

The C<data> method on microformat objects returns a hashref of useful data.

	use HTML::Microformats;
	use LWP::Simple qw[get];
	
	my $page     = 'http://example.net/';
	my @xfn_objs = HTML::Microformats
	               ->new_document(get($page), $page)
	               ->assume_all_profiles
	               ->parse_microformats
	               ->objects('XFN');
	
	while (my $xfn = shift @xfn_objs)
	{
		printf("%s <%s>\n",
			$xfn->data->{title},
			$xfn->data->{href},
			);
	}

(If you're wondering why the second example's simpler it's because it returns
somewhat dumber data.)

=item * B<Convert to other formats.>

Various microformat objects have C<to_foo> methods allowing the data to be
exported in various formats..

	use HTML::Microformats;
	use LWP::Simple qw[get];
	
	my $page     = 'http://example.net/';
	my @hcards   = HTML::Microformats
	               ->new_document(get($page), $page)
	               ->assume_all_profiles
	               ->parse_microformats
	               ->objects('hCard');
	
	print $_->to_vcard foreach @hcards;

Methods available are:

=over

=item * C<to_vcard> (hCard objects)

Exports as vCard 3.0.

=item * C<to_vcard4> (hCard objects)

Exports as vCard 4.0.

=item * C<to_vcard4_xml> (hCard objects)

Exports as vCard XML.

=item * C<to_icalendar> (hCalendar, hEvent, hTodo, hFreebusy, hAlarm and hEntry objects)

Exports as iCalendar.

=item * C<to_atom> (hAtom and hEntry objects)

Exports as Atom 1.0.

=item * C<to_kml> (geo objects)

Exports as KML 2.0.

=item * C<< serialialise_model(as => $format) >> (all microformat objects)

Exports as RDF, serialised as C<$format>. (Format can be 'RDFXML', 'Turtle',
'NTriples', 'RDFJSON'.)

=back

=back

=head2 Stuff that's b0rked

The C<get_foo>, C<set_foo>, C<add_foo>, C<clear_foo> methods defined in
L<HTML::Microformats::Format> work unreliably and are poorly documented.
You're better off using the C<data> method and inspecting the returned structure
for the data you need. This will be fixed in the future.

=head2 Here be monsters

There are several parts of the code which are incredibly complicated and desperately
need refactoring. This will be done at some point, so don't rely too much on their current
behaviour.

C<stringify> and C<_stringify_helper> in L<HTML::Microformats::Utilities>.
The whole of L<HTML::Microformats::Mixin::Parser>.

=head1 SEE ALSO

L<HTML::Microformats>.

=head1 AUTHOR

Toby Inkster E<lt>tobyink@cpan.orgE<gt>.

=head1 COPYRIGHT AND LICENCE

Copyright 2008-2012 Toby Inkster

This library is free software; you can redistribute it and/or modify it
under the same terms as Perl itself.

=cut