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 testThatTheInPageModelIsStoredCorrectlyInContext()
{
$this->agent->storeModel('inpage');
$store = MwRdf::StoredModel();
$context = $this->agent->getContextNode('inpage');
$model = $this->agent->buildModel('inpage');
$this->assertModelContainsModelInContext($store, $model, $context, "The inpage RDF model has not been successfully stored and retrieved.");
}
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;
}
/**
* As with QueryPage, this is the actual workhorse. It does everything
* needed to make a real, honest-to-gosh query page.
*
* @param $offset database query offset
* @param $limit database query limit
* @param $shownavigation show navigation like "next 200"?
*/
function doQuery($offset, $limit, $shownavigation = true)
{
global $wgUser, $wgOut, $wgContLang;
$this->offset = $offset;
$this->limit = $limit;
$sname = $this->getName();
$fname = get_class($this) . '::doQuery';
$sk = $wgUser->getSkin();
$wgOut->setSyndicated($this->isSyndicated());
$query = MwRdf::Query($this->getQuery(), $this->getBaseUrl(), $this->getQueryLanguage());
$librdf_res = $query->execute(MwRdf::StoredModel());
# let's just dump the tuples into a normal php array shall
# we? This will avoid memory management hassels.
$res = array();
foreach ($librdf_res as $tuple) {
$res[] = $tuple;
}
$num = count($res);
$res = $this->preprocessResults($res);
if ($shownavigation) {
$wgOut->addHTML($this->getPageHeader());
$top = wfShowingResults($offset, $num);
$wgOut->addHTML("<p>{$top}\n");
# often disable 'next' link when we reach the end
$atend = $num < $limit;
$sl = wfViewPrevNext($offset, $limit, $wgContLang->specialPage($sname), wfArrayToCGI($this->linkParameters()), $atend);
$wgOut->addHTML("<br />{$sl}</p>\n");
}
if ($num > 0) {
$s = array();
if (!$this->listoutput) {
$s[] = "<ol start='" . ($offset + 1) . "' class='special'>";
}
# here's where we do the offset and limit
for ($i = $offset; $i < $num && $i < $offset + $limit; $i++) {
$format = $this->formatResult($sk, $res[$i]);
if ($format) {
$s[] = $this->listoutput ? $format : "<li>{$format}</li>\n";
}
}
if (!$this->listoutput) {
$s[] = '</ol>';
}
$str = $this->listoutput ? $wgContLang->listToText($s) : implode('', $s);
$wgOut->addHTML($str);
}
if ($shownavigation) {
$wgOut->addHTML("<p>{$sl}</p>\n");
}
return $num;
}
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 testLoad()
{
$agent = MwRdf::ModelingAgent($this->title);
for ($i = 0; $i < 10000; $i++) {
$model = $agent->retrieveModel('inpage');
$this->assertType('object', $model);
$this->assertTrue($model instanceof LibRDF_Model);
$statement = MwRdf::Statement(MwRdf::UriNode('http://www.w3.org/TR/rdf-syntax-grammar'), MwRdf::UriNode('http://purl.org/dc/elements/1.1/title'), MwRdf::LiteralNode('RDF/XML Syntax Specification (Revised)'));
$this->assertTrue($model->hasStatement($statement));
$statements = $model->findStatements(MwRdf::UriNode('http://www.w3.org/TR/rdf-syntax-grammar'), MwRdf::UriNode('http://example.org/stuff/1.0/editor'), null);
$this->assertType('object', $statements);
$this->assertTrue($statements instanceof LibRDF_StreamIterator);
$statement = $statements->current();
$this->assertTrue($statement instanceof LibRdf_Statement);
$editor = $statement->getObject();
$this->assertTrue($editor instanceof LibRdf_BlankNode);
$statement = MwRdf::Statement($editor, MwRdf::UriNode('http://example.org/stuff/1.0/fullname'), MwRdf::LiteralNode("Dave Beckett"));
$this->assertTrue($model->hasStatement($statement));
$statement = MwRdf::Statement($editor, MwRdf::UriNode('http://example.org/stuff/1.0/homePage'), MwRdf::UriNode('http://purl.org/net/dajobe/'));
$this->assertTrue($model->hasStatement($statement));
$model->rewind();
}
}
public function getQuery()
{
$ns = MwRdf::Vocabulary('dc')->getNS();
return "\n\t\t\tPREFIX dc: <{$ns}>\n\t\t\tSELECT ?url ?title\n\t\t\tWHERE { ?url dc:title ?title } ";
}
function wfRdfSpecialPage($par)
{
global $wgRequest, $wgOut, $_SERVER, $_REQUEST;
MwRdf::setup();
$target = $wgRequest->getVal('target');
if (!isset($target) || $target == null) {
# no target parameter
MwRdf::ShowForm();
return;
}
if (strlen($target) == 0) {
# no target contents
MwRdf::ShowForm(wfMsg('badtitle'));
return;
}
$nt = Title::newFromText($target);
if ($nt->getArticleID() == 0) {
# not an article
MwRdf::ShowForm(wfMsg('badtitle'));
return;
}
$format = $wgRequest->getVal('format', 'rdfxml');
$accept = isset($_SERVER['HTTP_ACCEPT']) ? $_SERVER['HTTP_ACCEPT'] : null;
$rdftype = wfNegotiateType(wfAcceptToPrefs($accept), wfAcceptToPrefs(MwRdf::getTypePrefs($format)));
if (!$rdftype) {
wfHttpError(406, "Not Acceptable", wfMsg("notacceptable"));
return false;
}
$wgOut->disable();
if (!headers_sent()) {
header("Content-type: {$rdftype}");
}
$wgOut->sendCacheControl();
$mf = MwRdf::ModelingAgent($nt);
# Note: WebRequest chokes on arrays here
$modelnames = null;
if (isset($_REQUEST['modelnames'])) {
$modelnames = $_REQUEST['modelnames'];
}
if (is_null($modelnames)) {
$modelnames = $mf->listDefaultModels();
}
if (is_string($modelnames)) {
$modelnames = explode(',', $modelnames);
}
if (!$modelnames) {
MwRdf::ShowForm(wfMsg('nomodelsselected'));
return;
}
$model = $mf->retrieveModel($modelnames);
if (!$model->current()) {
$mf->storeAllModels();
$model = $mf->retrieveModel($modelnames);
}
$ser = MwRdf::Serializer($format);
$text = $model->serializeStatements($ser);
# XXX: Test Hook: it would be better if we could capture the
# print statement below with an output buffer, but that is
# disabled for the CLI in PHP 5
if (isset($_SERVER['CONTEXT']) && $_SERVER['CONTEXT'] == 'phpunit test') {
return $text;
} else {
print $text;
return true;
}
}
public function testTimestamp()
{
$this->markTestIncomplete();
$pagename = self::$username . '/NonArticle';
$title = Title::newFromText($pagename, NS_USER);
if ($title->exists()) {
$article = new Article($title);
$article->doPurge();
}
$agent = MwRdf::ModelingAgent($title);
$this->assertType('boolean', $agent->getTimestampResource());
$this->assertFalse($agent->getTimestampResource());
}
public function getQueryString()
{
$url = $this->Agent->getTitle()->getFullUrl();
$ns = MwRdf::Vocabulary('dcterms')->getNS();
return "PREFIX dcterms: <{$ns}>\n\t\t\tCONSTRUCT { <{$url}> dcterms:isReferencedBy ?x }\n\t\t\tWHERE { ?x dcterms:references <{$url}> } ";
}
public function testGetCcTerms()
{
$license = 'http://creativecommons.org/licenses/by-sa/1.0/';
$terms = MwRdf::GetCcTerms($license);
$this->assertType('array', $terms, $license);
$this->assertContains('re', $terms);
$this->assertContains('di', $terms);
$this->assertContains('de', $terms);
$this->assertContains('no', $terms);
$this->assertContains('by', $terms);
$this->assertContains('sa', $terms);
$license = 'http://creativecommons.org/licenses/by-nc-sa/1.0/';
$terms = MwRdf::GetCcTerms($license);
$this->assertType('array', $terms, $license);
$this->assertContains('re', $terms);
$this->assertContains('di', $terms);
$this->assertContains('de', $terms);
$this->assertContains('no', $terms);
$this->assertContains('by', $terms);
$this->assertContains('sa', $terms);
$this->assertContains('nc', $terms);
$license = 'http://creativecommons.org/licenses/GPL/2.0/';
$terms = MwRdf::GetCcTerms($license);
$this->assertType('array', $terms, $license);
$this->assertContains('de', $terms);
$this->assertContains('re', $terms);
$this->assertContains('di', $terms);
$this->assertContains('no', $terms);
$this->assertContains('sa', $terms);
$this->assertContains('sc', $terms);
$license = 'http://creativecommons.org/licenses/LGPL/2.1/';
$terms = MwRdf::GetCcTerms($license);
$this->assertType('array', $terms, $license);
$this->assertContains('de', $terms);
$this->assertContains('re', $terms);
$this->assertContains('di', $terms);
$this->assertContains('no', $terms);
$this->assertContains('sa', $terms);
$this->assertContains('sc', $terms);
$license = 'http://www.gnu.org/copyleft/fdl.html';
$terms = MwRdf::GetCcTerms($license);
$this->assertType('array', $terms, $license);
$this->assertContains('de', $terms);
$this->assertContains('re', $terms);
$this->assertContains('di', $terms);
$this->assertContains('no', $terms);
$this->assertContains('sa', $terms);
$this->assertContains('sc', $terms);
}
public function __construct()
{
foreach ($this as $key => $value) {
$this->{$key} = MwRdf::UriNode($this->getNS() . $key);
}
}
public function clearStoredModel($name)
{
$context = $this->getContextNode($name);
$store = MwRdf::StoredModel();
librdf_model_context_remove_statements($store->getModel(), $context->getNode());
$store->rewind();
$store = null;
# __destruct() causes a segfault
}
public function cleanup()
{
if (self::$Store) {
self::$Store = null;
}
if (self::$StoredModel) {
self::$StoredModel = null;
}
}
public function testShowFormWithMessage()
{
$msg = 'this is a message';
MwRdf::ShowForm($msg);
$this->assertContains($msg, $this->html);
}
if ($page_count < $start) {
fwrite(STDOUT, "-1");
exit(1);
}
MwRdf::Setup();
$dbr = wfGetDB(DB_SLAVE);
$res = $dbr->query("\n\tSELECT page_id,\n\tpage_title,\n\tpage_namespace,\n\tpage_is_redirect\n\tFROM page\n\tORDER BY page_id\n\tLIMIT 200\n\tOFFSET {$start}");
while ($s = $dbr->fetchObject($res)) {
if ($s->page_is_redirect) {
continue;
}
$title = Title::makeTitle($s->page_namespace, $s->page_title);
if (!$title) {
continue;
}
$agent = MwRdf::ModelingAgent($title);
fwrite(STDERR, $s->page_id . ": " . $title->getPrefixedDbKey());
$agent->storeAllModels();
$agent = null;
$title = null;
fwrite(STDERR, "\n");
}
fwrite(STDOUT, $start + 200);
function get_total_number_of_articles()
{
$dbr = wfGetDB(DB_SLAVE);
# get the total number of articles
$page = $dbr->tableName('page');
$res = $dbr->query("select count( page_id ) as page_count from {$page};");
$s = $dbr->fetchObject($res);
return $s->page_count;
