public function build()
{
$article = $this->Agent->getArticle();
$text = $article->getContent(true);
# Strip comments and <nowiki>
$text = preg_replace("/<!--.*?-->/s", "", $text);
$text = preg_replace("@<nowiki>.*?</nowiki>@s", "", $text);
# change template usage to substitution; note that this is WRONG
#$tchars = Title::legalChars();
#$text = preg_replace("/(?<!{){{([$tchars]+)(\|.*?)?}}(?!})/", "{{subst:$1$2}}", $text);
$parser = new Parser();
# so the magic variables work out right
$parser->mOptions = new ParserOptions();
$parser->mTitle = $this->Agent->getTitle();
$parser->mOutputType = OT_WIKI;
$parser->initialiseVariables();
$parser->clearState();
$text = $parser->replaceVariables($text);
preg_match_all("@<rdf>(.*?)</rdf>@s", $text, $matches, PREG_PATTERN_ORDER);
$content = $matches[1];
$rdf = implode(' ', array_values($content));
$model = MwRdf::Model();
if (strlen($rdf) > 0) {
$parser->mOutputType = OT_HTML;
$rdf = $parser->replaceVariables($rdf);
$turtle_parser = MwRdf::Parser('turtle');
$base_uri = $this->Agent->getTitle()->getFullUrl();
$prelude = MwRdf::getNamespacePrelude();
$model->loadStatementsFromString($turtle_parser, $prelude . $rdf);
}
return $model;
}
public function retrieve()
{
$sm = MwRdf::StoredModel();
$model = MwRdf::Model();
$context = $this->Agent->getContextNode($this->getName());
$stream = librdf_model_context_as_stream($sm->getModel(), $context->getNode());
librdf_model_add_statements($model->getModel(), $stream);
librdf_free_stream($stream);
$sm->rewind();
return $model;
}
public function retrieve()
{
$base_uri = $this->Agent->getTitle()->getFullUrl();
$query = MwRdf::Query($this->getQueryString(), $base_uri, 'sparql');
$store = MwRdf::StoredModel();
$res = $query->execute($store);
$model = MwRdf::Model();
foreach ($res as $statement) {
$model->addStatement($statement);
}
return $model;
}
public function testInPageModelTurnsUpInStoredModel()
{
$this->agent->storeModel('inpage');
$store = MwRdf::StoredModel();
# strip contexts (expensive for larger stores, fine here)
$smodel = MwRdf::Model();
foreach ($store as $s) {
$smodel->addStatement($s);
}
$store->rewind();
$model = $this->agent->buildModel('inpage');
$this->assertModelContainsModel($smodel, $model, "The inpage RDF model has not been successfully stored and retrieved.");
}
public function build()
{
$dc = MwRdf::Vocabulary('dc');
$model = MwRdf::Model();
$ar = $this->Agent->titleResource();
$categories = $this->Agent->getTitle()->getParentCategories();
if (is_array($categories)) {
foreach (array_keys($categories) as $category) {
$catmf = MwRdf::ModelingAgent(Title::newFromText($category));
$model->addStatement(MwRdf::Statement($ar, $dc->subject, $catmf->titleResource()));
}
}
return $model;
}
public function build()
{
$dcterms = MwRdf::Vocabulary('dcterms');
$model = MwRdf::Model();
$tr = $this->Agent->titleResource();
$dbr = wfGetDB(DB_SLAVE);
$res = $dbr->select('pagelinks', array('pl_namespace', 'pl_title'), array('pl_from = ' . $this->Agent->getTitle()->getArticleID()), 'MwRdfLinksFrom', array('ORDER BY' => 'pl_namespace, pl_title'));
while ($res && ($row = $dbr->fetchObject($res))) {
$lt = Title::makeTitle($row->pl_namespace, $row->pl_title);
$ltmf = MwRdf::ModelingAgent($lt);
$model->addStatement(MwRdf::Statement($tr, $dcterms->references, $ltmf->titleResource()));
}
$dbr->freeResult($res);
return $model;
}
public function build()
{
global $wgRightsUrl;
$rdf = MwRdf::Vocabulary('rdf');
$cc = MwRdf::Vocabulary('cc');
$model = MwRdf::Model();
$tr = $this->Agent->titleResource();
$model->addStatement(MwRdf::Statement($tr, $rdf->a, $cc->Work));
if (!isset($wgRightsUrl)) {
return $model;
}
$lr = MwRdf::UriNode($wgRightsUrl);
$model->addStatement(MwRdf::Statement($tr, $cc->license, $lr));
$model->addStatement(MwRdf::Statement($lr, $rdf->a, $cc->License));
$terms = MwRdf::GetCcTerms($wgRightsUrl);
if (!isset($terms)) {
return $model;
}
foreach ($terms as $term) {
switch ($term) {
case 're':
$model->addStatement(MwRdf::Statement($lr, $cc->permits, $cc->Reproduction));
break;
case 'di':
$model->addStatement(MwRdf::Statement($lr, $cc->permits, $cc->Distribution));
break;
case 'de':
$model->addStatement(MwRdf::Statement($lr, $cc->permits, $cc->DerivativeWorks));
break;
case 'nc':
$model->addStatement(MwRdf::Statement($lr, $cc->prohibits, $cc->CommercialUse));
break;
case 'no':
$model->addStatement(MwRdf::Statement($lr, $cc->requires, $cc->Notice));
break;
case 'by':
$model->addStatement(MwRdf::Statement($lr, $cc->requires, $cc->Attribution));
break;
case 'sa':
$model->addStatement(MwRdf::Statement($lr, $cc->requires, $cc->ShareAlike));
break;
case 'sc':
$model->addStatement(MwRdf::Statement($lr, $cc->requires, $cc->SourceCode));
break;
}
}
return $model;
}
public function build()
{
global $wgContLang;
$dc = MwRdf::Vocabulary('dc');
$dcterms = MwRdf::Vocabulary('dcterms');
$rdfs = MwRdf::Vocabulary('rdfs');
$model = MwRdf::Model();
$tr = $this->Agent->titleResource();
$article = $this->Agent->getArticle();
$text = $article->getContent(true);
$parser = new Parser();
$parser->mOptions = new ParserOptions();
$parser->mTitle = $this;
$parser->initialiseVariables();
$parser->clearState();
$tags = array('nowiki');
$m = array();
$text = $parser->extractTagsAndParams($tags, $text, $m);
# XXX: maybe it would actually be better to do this at another
# stage after the parser has already identified interwiki and
# lang links
# Find prefixed links
preg_match_all("/\\[\\[([^|\\]]+:[^|\\]]+)(\\|.*)?\\]\\]/", $text, $m);
if (!isset($m[0])) {
return $model;
}
// nothing found so nevermind
foreach ($m[1] as $linktext) {
$iwlink = Title::newFromText($linktext);
if (isset($iwlink)) {
$pfx = $iwlink->getInterwiki();
if (strlen($pfx) > 0) {
$iwlinkmf = MwRdf::ModelingAgent($iwlink);
$iwr = $iwlinkmf->titleResource();
# XXX: Wikitravel uses some 4+ prefixes for sister site links
if ($wgContLang->getLanguageName($pfx) && strlen($pfx) < 4) {
$model->addStatement(MwRdf::Statement($tr, $dcterms->hasVersion, $iwr));
$model->addStatement(MwRdf::Statement($iwr, $dc->language, MwRdf::Language($pfx)));
} else {
# XXX: Express the "sister site" relationship better
$model->addStatement(MwRdf::Statement($tr, $rdfs->seeAlso, $iwr));
}
}
}
}
return $model;
}
public function build()
{
global $wgServer;
$dc = MwRdf::Vocabulary('dc');
$dcterms = MwRdf::Vocabulary('dcterms');
$dctype = MwRdf::Vocabulary('dctype');
$model = MwRdf::Model();
$tr = $this->Agent->titleResource();
$dbr = wfGetDB(DB_SLAVE);
$res = $dbr->select('imagelinks', array('il_to'), array('il_from = ' . $this->Agent->getTitle()->getArticleID()), 'MwRdfImage');
while ($res && ($row = $dbr->fetchObject($res))) {
$img = wfFindFile($row->il_to);
if (!$img->exists()) {
continue;
}
$iuri = $img->getURL();
if ($iuri[0] == '/') {
$iuri = $wgServer . $iuri;
}
$ir = MwRdf::UriNode($iuri);
$model->addStatement(MwRdf::Statement($tr, $dcterms->hasPart, $ir));
$model->addStatement(MwRdf::Statement($ir, $dc->type, $dctype->Image));
$mt = $img->getMimeType();
if (isset($mt)) {
$model->addStatement(MwRdf::Statement($ir, $dc->format, MwRdf::MediaType($mt)));
}
$hist_line = $img->nextHistoryLine();
if (!isset($hist_line)) {
continue;
}
$creator = MwRdf::PersonToResource($hist_line->img_user, $hist_line->img_user_text, User::whoIsReal($hist_line->img_user));
$model->addStatement(MwRdf::Statement($ir, $dc->creator, $creator));
$model->addStatement(MwRdf::Statement($ir, $dc->date, MwRdf::TimestampResource($hist_line->img_timestamp)));
$seen = array($hist_line->img_user => true);
while ($hist_line = $img->nextHistoryLine()) {
if (isset($seen[$hist_line->img_user])) {
continue;
}
$contributor = MwRdf::PersonToResource($hist_line->img_user, $hist_line->img_user_text, User::whoIsReal($hist_line->img_user));
$model->addStatement(MwRdf::Statement($ir, $dc->contributor, $contributor));
$seen[$hist_line->img_user] = true;
}
}
$dbr->freeResult($res);
return $model;
}
public function build()
{
global $wgSitename, $wgContLanguageCode;
$article = $this->Agent->getArticle();
$model = MwRdf::Model();
$rdf = MwRdf::Vocabulary('rdf');
$dc = MwRdf::Vocabulary('dc');
$dctype = MwRdf::Vocabulary('dctype');
$artres = $this->Agent->titleResource();
$model->addStatement(MwRdf::Statement($artres, $dc->title, MwRdf::LiteralNode($this->Agent->getTitle()->getText())));
$model->addStatement(MwRdf::Statement($artres, $dc->publisher, MwRdf::PageOrString(wfMsg('aboutpage'), $wgSitename)));
$model->addStatement(MwRdf::Statement($artres, $dc->language, MwRdf::Language($wgContLanguageCode)));
$model->addStatement(MwRdf::Statement($artres, $dc->type, $dctype->Text));
$model->addStatement(MwRdf::Statement($artres, $dc->format, MwRdf::MediaType('text/html')));
if ($this->Agent->getTimestampResource()) {
$model->addStatement(MwRdf::Statement($artres, $dc->date, $this->Agent->getTimestampResource()));
}
if (MWNamespace::isTalk($this->Agent->getTitle()->getNamespace())) {
$model->addStatement(MwRdf::Statement($artres, $dc->subject, $this->Agent->subjectResource()));
} else {
$talk = MwRdf::ModelingAgent($this->Agent->getTitle()->getTalkPage());
$model->addStatement(MwRdf::Statement($talk->titleResource(), $dc->subject, $artres));
}
# 'Creator' is responsible for this version
$creator = MwRdf::PersonToResource($article->getUser());
$model->addStatement(MwRdf::Statement($artres, $dc->creator, $creator));
# 'Contributors' are all other version authors
$contributors = $article->getContributors();
foreach ($contributors as $user_parts) {
$contributor = MwRdf::PersonToResource($user_parts[0], $user_parts[1], $user_parts[2]);
$model->addStatement(MwRdf::Statement($artres, $dc->contributor, $contributor));
}
# Rights notification
global $wgRightsPage, $wgRightsUrl, $wgRightsText;
$rights = MwRdf::RightsResource();
if ($rights) {
$model->addStatement(MwRdf::Statement($artres, $dc->rights, $rights));
}
return $model;
}
public function build()
{
global $wgContLanguageCode;
$dc = MwRdf::Vocabulary('dc');
$dcterms = MwRdf::Vocabulary('dcterms');
$model = MwRdf::Model();
$tr = $this->Agent->titleResource();
$dbr = wfGetDB(DB_SLAVE);
$res = $dbr->select(array('page', 'revision'), array('rev_id', 'rev_timestamp', 'rev_user', 'rev_user_text'), array('page_namespace = ' . $this->Agent->getTitle()->getNamespace(), 'page_title = ' . $dbr->addQuotes($this->Agent->getTitle()->getDBkey()), 'rev_page = page_id', 'rev_id != page_latest'), 'MwRdfHistory', array('ORDER BY' => 'rev_timestamp DESC'));
while ($res && ($row = $dbr->fetchObject($res))) {
$url = $this->Agent->getTitle()->getFullURL('oldid=' . $row->rev_id);
$ur = MwRdf::UriNode($url);
$model->addStatement(MwRdf::Statement($tr, $dcterms->hasVersion, $ur));
$model->addStatement(MwRdf::Statement($ur, $dcterms->isVersionOf, $tr));
$model->addStatement(MwRdf::Statement($ur, $dc->language, MwRdf::Language($wgContLanguageCode)));
$realname = $row->rev_user == 0 ? null : User::whoIsReal($row->rev_user);
$pr = MwRdf::PersonToResource($row->rev_user, $row->rev_user_text, $realname);
$model->addStatement(MwRdf::Statement($ur, $dc->creator, $pr));
$model->addStatement(MwRdf::Statement($ur, $dc->date, MwRdf::TimestampResource($row->rev_timestamp)));
}
$dbr->freeResult($res);
return $model;
}
public function testSpecialPageProducesCorrectHistoryModel()
{
$this->format = 'turtle';
$parser = MwRdf::Parser('turtle');
$this->prepareRequest(null, array('history'));
$text = wfRdfSpecialPage(null);
$model = MwRdf::Model();
$model->loadStatementsFromString($parser, $text);
$expect = MwRdfTest::createExpectedHistoryModel();
$this->assertModelEqualsModel($expect, $model);
}
public function retrieveModel($par)
{
if (gettype($par) == 'string') {
$par = array($par);
}
if (!gettype($par) == 'array') {
throw new Exception('retrieveModel takes either a model ' . 'name string or an array of modelnames');
}
$model = MwRdf::Model();
foreach ($par as $name) {
$mm = $this->ModelMakers[$name];
### TODO Retrieve from Cache if it's there
$sub_model = $mm->retrieve($name);
### TODO Cache the sub_model
$stream = librdf_model_as_stream($sub_model->getModel());
librdf_model_add_statements($model->getModel(), $stream);
librdf_free_stream($stream);
$sub_model->rewind();
}
return $model;
}
public function testModel()
{
$m = MwRdf::Model();
$this->assertTrue($m instanceof LibRDF_Model);
}
