Ejemplos de Xml::getRSSTitle en C++ (Cpp)

Ejemplo n.º 1

Archivo: Matches.cpp Proyecto: privacore/open-source-search-engine

// . this was in Summary.cpp, but is more useful here
// . we can also use this to replace the proximity algo setup where it
//   fills in the matrix for title, link text, etc.
// . returns false and sets g_errno on error
bool Matches::set( Words *bodyWords, Phrases *bodyPhrases, Sections *bodySections, Bits *bodyBits,
				   Pos *bodyPos, Xml *bodyXml, Title *tt, Url *firstUrl, LinkInfo *linkInfo ) {
	// don't reset query info!
	reset2();

	// . first add all the matches in the body of the doc
	// . add it first since it will kick out early if too many matches
	//   and we get all the explicit bits matched
	if ( !addMatches( bodyWords, bodyPhrases, bodySections, bodyBits, bodyPos, MF_BODY ) ) {
		return false;
	}

	// add the title in
	if ( !addMatches( tt->getTitle(), tt->getTitleLen(), MF_TITLEGEN ) ) {
		return false;
	}

	// add in the url terms
	if ( !addMatches( firstUrl->getUrl(), firstUrl->getUrlLen(), MF_URL ) ) {
		return false;
	}

	// also use the title from the title tag, because sometimes it does not equal "tt->getTitle()"
	int32_t  a     = tt->getTitleTagStart();
	int32_t  b     = tt->getTitleTagEnd();

	char *start = NULL;
	char *end   = NULL;
	if ( a >= 0 && b >= 0 && b>a ) {
		start = bodyWords->getWord(a);
		end   = bodyWords->getWord(b-1) + bodyWords->getWordLen(b-1);
		if ( !addMatches( start, end - start, MF_TITLETAG ) ) {
			return false;
		}
	}

	// now add in the meta tags
	int32_t n = bodyXml->getNumNodes();
	XmlNode *nodes = bodyXml->getNodes();

	// find the first meta summary node
	for ( int32_t i = 0 ; i < n ; i++ ) {
		// continue if not a meta tag
		if ( nodes[i].m_nodeId != TAG_META ) continue;
		// only get content for <meta name=..> not <meta http-equiv=..>
		int32_t tagLen;
		char *tag = bodyXml->getString ( i , "name" , &tagLen );
		// is it an accepted meta tag?
		int32_t flag = 0;
		if (tagLen== 7&&strncasecmp(tag,"keyword"    , 7)== 0)
			flag = MF_METAKEYW;
		if (tagLen== 7&&strncasecmp(tag,"summary"    , 7)== 0)
			flag = MF_METASUMM;
		if (tagLen== 8&&strncasecmp(tag,"keywords"   , 8)== 0)
			flag = MF_METAKEYW;
		if (tagLen==11&&strncasecmp(tag,"description",11)== 0)
			flag = MF_METADESC;
		if ( ! flag ) continue;
		// get the content
		int32_t len;
		char *s = bodyXml->getString ( i , "content" , &len );
		if ( ! s || len <= 0 ) continue;
		// wordify
		if ( !addMatches( s, len, flag ) ) {
			return false;
		}
	}

	// . now the link text
	// . loop through each link text and it its matches

	// loop through the Inlinks
	Inlink *k = NULL;
	for ( ; (k = linkInfo->getNextInlink(k)) ; ) {
		// does it have link text? skip if not.
		if ( k->size_linkText <= 1 ) {
			continue;
		}

		// set the flag, the type of match
		mf_t flags = MF_LINK;

		// add it in
		if ( !addMatches( k->getLinkText(), k->size_linkText - 1, flags ) ) {
			return false;
		}

		// set flag for that
		flags = MF_HOOD;

		// add it in
		if ( !addMatches( k->getSurroundingText(), k->size_surroundingText - 1, flags ) ) {
			return false;
		}

		// parse the rss up into xml
		Xml rxml;
		if ( ! k->setXmlFromRSS ( &rxml ) ) {
			return false;
		}

		// add rss description
		bool isHtmlEncoded;
		int32_t rdlen;
		char *rd = rxml.getRSSDescription( &rdlen, &isHtmlEncoded );
		if ( !addMatches( rd, rdlen, MF_RSSDESC ) ) {
			return false;
		}

		// add rss title
		int32_t rtlen;
		char *rt = rxml.getRSSTitle( &rtlen, &isHtmlEncoded );
		if ( !addMatches( rt, rtlen, MF_RSSTITLE ) ) {
			return false;
		}
	}

	// that should be it
	return true;
}