Merge branch 'master' into dev-siteinfo

BaseScoringParameters.cpp

@@ -70,3 +70,11 @@ void BaseScoringParameters::traceToLog(const char *prefix) {
 		log(LOG_TRACE,"%s:  m_flagRankAdjustment[%d]=%d",prefix,i,m_flagRankAdjustment[i]);
 	}
 }
+
+
+bool BaseScoringParameters::allLanguageWeightsAreTheSame() const {
+	for(unsigned i=0; i<64; i++)
+		if(m_languageWeights[i]!=m_languageWeights[0])
+			return false;
+	return true;
+}

BaseScoringParameters.h

@@ -40,6 +40,7 @@ struct BaseScoringParameters {
 	BaseScoringParameters() { clear(); }
 	void clear();
 	void traceToLog(const char *prefix);
+	bool allLanguageWeightsAreTheSame() const;
 };
 
 #endif

PageResults.cpp

@@ -345,12 +345,28 @@ bool sendPageResults ( TcpSocket *s , HttpRequest *hr ) {
 	                                                       tld_hint, strlen(tld_hint), true);
 
 
-	// Then try the external language detection server
-	if (g_queryLanguage.getLanguage(st, gotQueryLanguageWrapper, fx_qlang, fx_blang, fx_country, fx_fetld, query)) {
-		// blocked
-		return false;
+	//We can't use st->m_si.m_baseScoringParameters.allLanguageWeightsAreTheSame() because the cgi parameters have not set the members in SearchInput yet.
+	//So we have to check if any CGI parameter using the "lw_" prefix has been specified.
+	bool any_specific_language_weight_specified = false;
+	for(int32_t i = 0; i < hr->getNumFields(); i++) {
+		if(!hr->getValue(i))
+			continue;
+		const char *full_field_name = hr->getField(i);
+		if(strncmp(full_field_name,"lw_",3)==0) {
+			any_specific_language_weight_specified = true;
+			break;
+		}
+	}
+	
+	if(!any_specific_language_weight_specified) {
+		// Then try the external language detection server
+		if (g_queryLanguage.getLanguage(st, gotQueryLanguageWrapper, fx_qlang, fx_blang, fx_country, fx_fetld, query)) {
+			// blocked
+			return false;
+		}
 	} else
-		return gotQueryLanguage(st, {});
+		log(LOG_DEBUG,"Explicit language weight specified. Not contacting external language detection server");
+	return gotQueryLanguage(st, {});
 }
 
 static void gotQueryLanguageWrapper(void *state, const std::vector<std::pair<lang_t, double>> &language_weights) {

Parms.cpp

@@ -4016,7 +4016,7 @@ void Parms::init ( ) {
 		"set in the search controls and is usually something like "
 		"20.0. Which means that we multiply a result's score by 20 "
 		"if from the same language as the query or the language is "
-		"unknown.";
+		"unknown. Has no effect if any of the per-language weights have been specified.";
 	simple_m_set(SearchInput,m_sameLangWeight);
 	m->m_defOff= offsetof(CollectionRec,m_sameLangWeight);
 	m->m_cgi  = "langw";
@@ -4028,7 +4028,7 @@ void Parms::init ( ) {
 	m->m_desc  = "Use this to override the default uknown language weight "
 		"for this collection. We multiply a result's score by this value "
 		"if the user requested a specific language, but the language of the "
-		"indexed page could not be determined.";
+		"indexed page could not be determined. Has no effect if any of the per-language weights have been specified.";
 	simple_m_set(SearchInput,m_unknownLangWeight);
 	m->m_defOff= offsetof(CollectionRec,m_unknownLangWeight);
 	m->m_cgi  = "ulangw";
@@ -4036,6 +4036,23 @@ void Parms::init ( ) {
 	m->m_page  = PAGE_RESULTS;
 	m++;
 
+	for(int i=0; i<MAX_LANGUAGES; i++) {
+		static char title[MAX_LANGUAGES][64];
+		sprintf(title[i],"Language weight for %s", getLanguageString(i));
+		static char cgi[MAX_LANGUAGES][64];
+		sprintf(cgi[i],"lw_%s", getLanguageAbbr(i));  //note: sendPageResults() relies on this
+		m->m_title = title[i];
+		m->m_desc  = "";
+		m->m_obj = OBJ_SI;
+		m->m_off = offsetof(SearchInput,m_baseScoringParameters.m_languageWeights) + sizeof(float)*i;
+		m->m_type = TYPE_FLOAT;
+		m->m_def   = "1.0";
+		m->m_cgi   = cgi[i];
+		m->m_flags = PF_API;
+		m->m_page  = PAGE_RESULTS;
+		m++;
+	}
+	
 	m->m_title = "site-rank multiplier";
 	m->m_desc  = "formula: score = (siterank*multiplier)+1";
 	simple_m_set(SearchInput,m_baseScoringParameters.m_siteRankMultiplier);

SearchInput.cpp

@@ -262,16 +262,20 @@ bool SearchInput::set(TcpSocket *sock, HttpRequest *r, lang_t primaryQueryLangua
 	log(LOG_INFO, "query: using primary query lang of %s", getLanguageAbbr(m_queryLangId));
 
 	if(!language_weights.empty()) {
+		//external language detection server had results
 		for(int i=0; i<64; i++)
 			m_baseScoringParameters.m_languageWeights[i] = 0.01; //server returns weights in the range [0..1]
 		for(const auto &e : language_weights)
 			m_baseScoringParameters.m_languageWeights[e.first] = e.second;
 	} else {
-		//query language server is unavailable.
-		for(int i=0; i<64; i++)
-			m_baseScoringParameters.m_languageWeights[i] = 1.0;
-		m_baseScoringParameters.m_languageWeights[langUnknown] = m_unknownLangWeight; //backward compatible. questionable behaviour
-		m_baseScoringParameters.m_languageWeights[m_queryLangId] = m_sameLangWeight;
+		if(m_baseScoringParameters.allLanguageWeightsAreTheSame()) {
+			//query language server is unavailable.
+			for(int i=0; i<64; i++)
+				m_baseScoringParameters.m_languageWeights[i] = 1.0;
+			m_baseScoringParameters.m_languageWeights[langUnknown] = m_unknownLangWeight; //backward compatible. questionable behaviour
+			m_baseScoringParameters.m_languageWeights[m_queryLangId] = m_sameLangWeight;
+		}
+		//else: user specified language weights explicitly
 	}
 	
 	int32_t maxQueryTerms = cr->m_maxQueryTerms;