RDFStore

Changes for version 0.5 - 2004-08-19

A brand new indexing model for RDF data has been developed in this release - such a indexing model allows to store quite efficienty "pedantic" RDF descriptions by leveraging on compression of the index of triples using a custom RLE+VLE compression algorithm written in C and XS. It is under investigation the real need to have a custom compression algorithm for RDFStore instead of using one in the public domain; some tests showed that general purpose algorithms like LZO or LZF perfom worse in the specific case of RDFStore where the sparse matrix contains well-known patterns.
This version start to support the RDF Data Query language (RDQL/Squish) via a proper DBD::RDFStore DBI driver (see http://www.hpl.hp.co.uk/people/afs/RDQL/ and http://swordfish.rdfweb.org/rdfquery/) The software support tha basic query language with some extensions to run free-text queries over literals (property values). See also the RDF Query and Rules page at http://www.w3.org/2003/03/rdfqr-tests/
RDFStore::Model now support basic single statement-grouping (i.e. "context") per statement that allow to run storage operations giving a kind of context to them e.g. a spatial or temporal context or just the source URI; all basic model method have been updated to support that. This stuff is really much close to the on going work of RDF Core WG and other proposals about adding context information to individual triples. This software part is just meant to be a useful and practical artifact and does not want or pretend any of the complex issues being discussed in the RDF community.
The RDFStore::Parser::OpenHealth parser has been removed from the distribution
Being that the new storage model does not require Data::MagicTie anymore, the module is being distributed separatly (not on CPAN even ;)
Some useless modules has been removed from the distribution and some has been renamed; in general the basic storage functionalities are provided through the RDFStore module, while the Stanford API specific code is available in the RDFStore::* modules. The SiRPAC RDF parser is still being distributed with the main storage sources but they will be also distributed separately in the future (XML::SAX::RDF::Filter or something)
UTF-8 (Unicode) support has been natively introduced into this release - custom C/XS code has been written to allow convertion from/to utf8 to other char encodings - the perl use can also use the RDFStore::Util::UTF8(3) to get general string utility functions which work on UTF8/Unicode.
Things in CVS now also into the sourceoforge RDFStore CVS repository. In addition the system should compile fine on most of the UNIX systems out there (including the dbms/ tcp/ip deamon)
Backend now fully general - software ships with BerkeleyDB, DBMS and caching (in-memory) backends
Basic caching layer added which should speed up all the DB and I/O operations
t/
- updated all test cases
- added 04_rdql.t, 05_serializier.t and 06_objects tests
RDFStore
- updated to be the corner stone of the RDF storage actually implemented in C and XS (lots of C and XS code really hoping to gain some speed and credibility here :) Here is the place where all perl world magics happen....almost
- now brand new C low-level API interfaced to perl via XS code
RDQL::Parser
- new module
DBD::RDFStore
- new DBI driver to run RDQL/Squish SQL like queries :-) allows to dump query results in several different formats
RDFStore::Iterator
- new module
RDFStore::Object
- new module
RDFStore::Serializer
- new module
RDFStore::Serializer::RDFXML
- new module
RDFStore::Serializer::NTriples
- new module
RDFStore::Serializer::RSS
- new module
RDFStore::Util::UTF8
- new module
RDFStore::Model
- brand new design now using the faster C/XS RDFStore(3) module....finally :)
- updated methods to avoid a full copy of statements across when the model is shared if possible
- added basic support for statements grouping - see setContext(), getContext() and resetContext()
- zapped toStrawmanRDF() method
- added serialize() method to generally dump a model/graph to a string or filehanlde
- added isConnected() and isRemote() methods
- added unite(), subtract(), intersect(), complement() and exor() methods
- re-added RDFStore::Resource inheritance
- added getParser(), getReader(), getSerializer() and getWriter() methods
RDFStore::Parser
- new module
- includes wget() now using IO::Socket and other general purpose parsing methods
RDFStore::Parser::SiRPAC
- fixed bug in processDescription()
- fixed bug in processTypedNode() when removeAttribute
- fixed bug in makeAbsolute() when LocalName contains '#'
- removed xml:space handling in RDFXML_CharacterDataHandler()
- fixed bug in processPredicate() - does not generate triples when PCDATA is pure whitespaces stuff and there are XML subelements
- fixed bug in processListItem() when generate rdf:li elements
- added rdfcroak() instead of using $expat->xpcroak()
- updated newReificationID()
- added RDFMS_nodeID and RDFMS_datatype
- updated reify()
- added rdf:nodeID support
- added rdf:parseType="Collection" support to processPredicate()
- fixed bug in processPredicate() to force a resource node object for rdf:type on predicate with rdf:resource
- removed parse_start() method and added parsestream() to do expat no-blocking parseing of large XML streams
- removed processListItem() - new specs require to process containers as nomral predicates and simply enumerating elements
- fixed bug in processContainer() to treat RDF containers just like any other typed node but with rdf:li or rdf:_n nodes
- added xml:base support
- added xml:lang support
- added manage_bNodes callback/hanlder
- updated bNode identifier generation algorithm - now parser run wide unique - see newReificationID()
- various fixes when using getAttributeValue()
- updated rdfcroak() to return source name too when failing
- updated processXML() - removed the fetchSchema part
- force source to STDIN: if not defined
- removed RDFCore_Issues option - now default
- added rdfwarn()
- added warnings()
- moved common code to RDFStore::Parser
- added rdf:datatype support
- added rdfstore:context support
RDFStore::Parser::NTriples
- new module
RDFStore::Vocabulary::DC
- updated DCMI namespace to http://purl.org/dc/elements/1.1/
RDFStore::Parser::Styles::RDFStore::Model
- renamed from RDFStore::Parser::Styles::MagicTie
- added Context option to the storage
- fixed typing error when apssing the nodeFactory to the model
- allows to specify an existing RDFStore::Model as input
- added owl:imports support
utils/
- updated rdfquery.pl to use RDQL/Squish syntax and read some default storage parameters
- added context/provenance flags to rdf.pl and rdfingest.pl
- added pod2html.pl utility to HTMl format native perl POD documentation
RDFStore::Resource
- added anonymous resources support - see also RDFStore::NodeFactory(3) and RDFStore::Model(3)
- added isAnonymous() and isbNode()
- updated accordingly to new RDFStore API
- removed BLOB support
RDFStore::RDFNode
- updated accordingly to new RDFStore API
- removed BLOB support
RDFStore::Literal
- updated accordingly to new RDFStore API
- removed BLOB support
RDFStore::Statement
- updated accordingly to new RDFStore API
- added statements reification and context support
RDFStore::Vocabulary::Generator
- updated to use RDFStore::Model new API
RDFStore::Vocabulary::DCQ
- updated
RDFStore::Vocabulary::DCT
- new vocabulary for DCMI Type Vocabulary
RDFStore::Vocabulary::RSS
- new vocabulary for RSS1.0 see http://purl.org/rss/1.0/
RDFStore::Vocabulary::RDFStoreContext
- new vocabulary for RDFStore contexts
RDFStore::Vocabulary::OWL
- new vocabulary for OWL see http://www.w3.org/2002/07/owl#
RDFStore::NodeFactory
- added anonymous resource support via createAnonymousResource() and createbNode() - see also RDFStore::Resource(3)
- added statements reification support via createReifiedStatement() - see also RDFStore::Statement(3)
- updated accordingly to new RDFStore API
- added createNTriple() method
RDFStore::SetModel
- removed now included into main RDFStore::Model(3)
RDFStore::Util::Digest
- added as new module to manage SHA-1 cryptographic digests natively (using apache SHA-1 code) most of functionality like old module RDFStore::Stanford::Digest::Util(2)
RDFStore::Model::Iterator
- re-implementation of old RDFStore::Model::Statements using RDFStore::Iterator