Ticket #279: remotepercentscale.patch

matcher/multimatch.cc

@@ -779 +779 @@
         vector<Xapian::Internal::MSetItem>::const_iterator best;
         best = min_element(items.begin(), items.end(), mcmp);
 
-        if (termfreqandwts.size() > 1) {
-            Xapian::termcount matching_terms = 0;
-            map<string,
-                Xapian::MSet::Internal::TermFreqAndWeight>::const_iterator i;
+        unsigned int multiplier = db.internal.size();
+        Assert(multiplier != 0);
+        Xapian::doccount n = (best->did - 1) % multiplier; // which actual database
+        // If the top result is from a remote database, then we can
+        // just use the percentage scaling calculated for that
+        // database.
+        if (is_remote[n]) {
+            RemoteSubMatch * rem_match;
+            rem_match = static_cast<RemoteSubMatch*>(leaves[n].get());
+            percent_scale = rem_match->get_percent_factor();
+        } else {
+            if (termfreqandwts.size() > 1) {
+                Xapian::termcount matching_terms = 0;
+                map<string,
+                    Xapian::MSet::Internal::TermFreqAndWeight>::const_iterator i;
 
-            Xapian::TermIterator docterms = db.termlist_begin(best->did);
-            Xapian::TermIterator docterms_end = db.termlist_end(best->did);
-            while (docterms != docterms_end) {
-                i = termfreqandwts.find(*docterms);
+                Xapian::TermIterator docterms = db.termlist_begin(best->did);
+                Xapian::TermIterator docterms_end = db.termlist_end(best->did);
+                while (docterms != docterms_end) {
+                    i = termfreqandwts.find(*docterms);
+                    if (i != termfreqandwts.end()) {
+                        percent_scale += i->second.termweight;
+                        ++matching_terms;
+                        if (matching_terms == termfreqandwts.size()) break;
+                    }
+                    ++docterms;
+                }
+                // Special case for MatchAll queries
+                i = termfreqandwts.find("");
                 if (i != termfreqandwts.end()) {
                     percent_scale += i->second.termweight;
                     ++matching_terms;
-                    if (matching_terms == termfreqandwts.size()) break;
                 }
-                ++docterms;
-            }
-            // Special case for MatchAll queries
-            i = termfreqandwts.find("");
-            if (i != termfreqandwts.end()) {
-                percent_scale += i->second.termweight;
-                ++matching_terms;
-            }
-            if (matching_terms < termfreqandwts.size()) {
-                // OK, work out weight corresponding to 100%
-                double denom = 0;
-                for (i = termfreqandwts.begin(); i != termfreqandwts.end(); ++i)
-                    denom += i->second.termweight;
+                if (matching_terms < termfreqandwts.size()) {
+                    // OK, work out weight corresponding to 100%
+                    double denom = 0;
+                    for (i = termfreqandwts.begin(); i != termfreqandwts.end(); ++i)
+                        denom += i->second.termweight;
 
-                DEBUGLINE(MATCH, "denom = " << denom << " percent_scale = " << percent_scale);
-                Assert(percent_scale <= denom);
-                denom *= greatest_wt;
-                Assert(denom > 0);
-                percent_scale /= denom;
+                    DEBUGLINE(MATCH, "denom = " << denom << " percent_scale = " << percent_scale);
+                    Assert(percent_scale <= denom);
+                    denom *= greatest_wt;
+                    Assert(denom > 0);
+                    percent_scale /= denom;
+                } else {
+                    // If all the terms match, the 2 sums of weights cancel
+                    percent_scale = 1.0 / greatest_wt;
+                }
             } else {
-                // If all the terms match, the 2 sums of weights cancel
+                // If there's only a single term in the query, the top document
+                // must score 100%.
                 percent_scale = 1.0 / greatest_wt;
             }
-        } else {
-            // If there's only a single term in the query, the top document
-            // must score 100%.
-            percent_scale = 1.0 / greatest_wt;
         }
         Assert(percent_scale > 0);
         if (percent_cutoff) {

matcher/remotesubmatch.cc

@@ -63 +63 @@
               "[matcher], " << (void*)termfreqandwts);
     Xapian::MSet mset;
     db->get_mset(mset);
+    percent_factor = mset.internal->percent_factor;
     if (termfreqandwts) *termfreqandwts = mset.internal->termfreqandwts;
     return new MSetPostList(mset, decreasing_relevance);
 }

matcher/remotesubmatch.h

@@ -44 +44 @@
      */
     bool decreasing_relevance;
 
+    /// The factor to use to convert weights to percentages.
+    double percent_factor;
+
   public:
     /// Constructor.
     RemoteSubMatch(RemoteDatabase *db_, bool decreasing_relevance_);
@@ -61 +64 @@
     PostList * get_postlist_and_term_info(MultiMatch *matcher,
         map<string, Xapian::MSet::Internal::TermFreqAndWeight> *termfreqandwts);
 
+    /// Get percentage factor - only valid after get_postlist_and_term_info().
+    double get_percent_factor() const { return percent_factor; }
+
     /// Short-cut for single remote match.
     void get_mset(Xapian::MSet & mset) { db->get_mset(mset); }
 };