oracular (3) RDF::Generator::HTTP.3pm.gz
NAME
RDF::Generator::HTTP - Generate RDF from a HTTP message
SYNOPSIS
use LWP::UserAgent;
my $ua = LWP::UserAgent->new;
my $response = $ua->get('http://search.cpan.org/');
use RDF::Generator::HTTP;
use RDF::Trine qw(iri);
my $g = RDF::Generator::HTTP->new(message => $response,
graph => iri('http://example.org/graphname'),
blacklist => ['Last-Modified', 'Accept']);
my $model = $g->generate;
print $model->size;
my $s = RDF::Trine::Serializer->new('turtle', namespaces =>
{ httph => 'http://www.w3.org/2007/ont/httph#',
http => 'http://www.w3.org/2007/ont/http#' } );
$s->serialize_model_to_file(\*STDOUT, $model);
DESCRIPTION
This module simply takes a HTTP::Message object, and based on its content, especially the content the
HTTP::Header object(s) it contains, creates a simple RDF representation of the contents. It is useful
chiefly for recording data when crawling resources on the Web, but it may also have other uses.
Constructor
"new(%attributes)"
Moose-style constructor function.
Attributes
These attributes may be passed to the constructor to set them, or called like methods to get them.
"message"
A HTTP::Message (or subclass thereof) object to generate RDF for. Required.
"blacklist"
An "ArrayRef" of header field names that you do not want to see in the output.
"whitelist"
An "ArrayRef" of the only header field names that you want to see in the output. The whitelist will
be ignored if the blacklist is set.
"graph"
You may pass an optional graph name to be used for all triples in the output. This must be an object
of RDF::Trine::Node::Resource.
"ns"
An URI::NamespaceMap object containing namespace prefixes used in the module. You should probably not
override this even though you can.
"request_subject"
An RDF::Trine::Node object containing the subject of any statements describing requests. If unset, it
will default to a blank node.
"response_subject"
An RDF::Trine::Node object containing the subject of any statements describing responses. If unset,
it will default to a blank node.
Methods
The above attributes all have read-accessors by the same name. "blacklist", "whitelist" and "graph" also
has writers and predicates, which is used to test if the attribute has been set, by prefixing "has_" to
the attribute name.
This class has two methods:
"generate ( [ $model ] )"
This method will generate the RDF. It may optionally take an RDF::Trine::Model as parameter. If it
exists, the RDF will be added to this model, if not, a new Memory model will be created and returned.
"ok_to_add ( $field )"
This method will look up in the blacklists and whitelists and return true if the given field and
value may be added to the model.
EXAMPLES
For an example of what the module can be used to create, consider the example in the "SYNOPSIS", which at
the time of this writing outputs the following Turtle:
@prefix http: <http://www.w3.org/2007/ont/http#> .
@prefix httph: <http://www.w3.org/2007/ont/httph#> .
[] a http:RequestMessage ;
http:hasResponse [
a http:ResponseMessage ;
http:status "200" ;
httph:client_date "Sun, 14 Dec 2014 21:28:21 GMT" ;
httph:client_peer "207.171.7.59:80" ;
httph:client_response_num "1" ;
httph:connection "close" ;
httph:content_length "3643" ;
httph:content_type "text/html" ;
httph:date "Sun, 14 Dec 2014 21:28:21 GMT" ;
httph:link "<http://search.cpan.org/uploads.rdf>; rel=\"alternate\"; title=\"RSS 1.0\"; type=\"application/rss+xml\"", "<http://st.pimg.net/tucs/opensearch.xml>; rel=\"search\"; title=\"SearchCPAN\"; type=\"application/opensearchdescription+xml\"", "<http://st.pimg.net/tucs/print.css>; media=\"print\"; rel=\"stylesheet\"; type=\"text/css\"", "<http://st.pimg.net/tucs/style.css?3>; rel=\"stylesheet\"; type=\"text/css\"" ;
httph:server "Plack/Starman (Perl)" ;
httph:title "The CPAN Search Site - search.cpan.org" ;
httph:x_proxy "proxy2"
] ;
http:method "GET" ;
http:requestURI <http://search.cpan.org/> ;
httph:user_agent "libwww-perl/6.05" .
NOTES
HTTP Vocabularies
There have been many efforts to create HTTP vocabularies (or ontologies), where the most elaborate and
complete is the HTTP Vocabulary in RDF 1.0 <http://www.w3.org/TR/HTTP-in-RDF/>. Nevertheless, I decided
not to support this, but rather support an older and much less complete vocabulary that has been in the
Tabulator <https://github.com/linkeddata/tabulator-firefox> project, with the namespace prefixes
<http://www.w3.org/2007/ont/http#> and <http://www.w3.org/2007/ont/httph#>. The problem of modelling HTTP
is that headers modify each other, so if you want to record the HTTP headers so that they can be used in
an actual HTTP dialogue afterwards, they have to be in a container so that the order can be
reconstructed. Moreover, there is a lot of microstructure in the values, and that also adds complexity
if you want to translate all that to RDF. That's what the former vocabulary does. However, for now, all
the author wants to do is to record them, and then neither of these concerns are important. Therefore, I
opted to go for a much simpler vocabulary, where each field is a simple predicate. That is not to say
that the former approach isn't valid, it is just not something I need now.
BUGS
This is a very early release, but it works for the author.
Please report any bugs to <https://github.com/kjetilk/p5-rdf-generator-http/issues>.
AUTHOR
Kjetil Kjernsmo <kjetilk@cpan.org>.
COPYRIGHT AND LICENCE
This software is copyright (c) 2014 by Kjetil Kjernsmo.
This is free software; you can redistribute it and/or modify it under the same terms as the Perl 5
programming language system itself.
DISCLAIMER OF WARRANTIES
THIS PACKAGE IS PROVIDED "AS IS" AND WITHOUT ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, WITHOUT
LIMITATION, THE IMPLIED WARRANTIES OF MERCHANTIBILITY AND FITNESS FOR A PARTICULAR PURPOSE.