added another peer selection rule:

- select also non-robinson (dht-) peers if their peer tags match with search words
- the peer tag '*' can now act as catch-all rule: shall be selected always

git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@4963 6c8d7289-2bf4-0310-a012-ef5d649a1542

build.properties

@@ -3,7 +3,7 @@ javacSource=1.5
 javacTarget=1.5
 
 # Release Configuration
-releaseVersion=0.591
+releaseVersion=0.592
 stdReleaseFile=yacy_v${releaseVersion}_${DSTAMP}_${releaseNr}.tar.gz
 embReleaseFile=yacy_emb_v${releaseVersion}_${DSTAMP}_${releaseNr}.tar.gz
 proReleaseFile=yacy_pro_v${releaseVersion}_${DSTAMP}_${releaseNr}.tar.gz

source/de/anomic/yacy/yacySearch.java

@@ -177,6 +177,7 @@ public class yacySearch extends Thread {
         // put in seeds according to dht
         final kelondroMScoreCluster<String> ranking = new kelondroMScoreCluster<String>();
         final HashMap<String, yacySeed> regularSeeds = new HashMap<String, yacySeed>();
+        final HashMap<String, yacySeed> matchingSeeds = new HashMap<String, yacySeed>();
         final HashMap<String, yacySeed> robinsonSeeds = new HashMap<String, yacySeed>();
         yacySeed seed;
         Iterator<yacySeed> dhtEnum;         
@@ -225,40 +226,49 @@ public class yacySearch extends Thread {
                 // enhance ranking for regular peers
                 if (seed.matchPeerTags(wordhashes)) { // access robinson peers with matching tag
                     serverLog.logInfo("PLASMA", "selectPeers/PeerTags: " + seed.hash + ":" + seed.getName() + ", is specialized peer for " + seed.getPeerTags().toString());
-                    ranking.addScore(seed.hash, seedcount);
-                    regularSeeds.put(seed.hash, seed);
+                    regularSeeds.remove(seed.hash);
+                    ranking.deleteScore(seed.hash);
+                    matchingSeeds.put(seed.hash, seed);
                 }
                 if (seed.getAge() < 1) { // the 'workshop feature'
                     serverLog.logInfo("PLASMA", "selectPeers/Age: " + seed.hash + ":" + seed.getName() + ", is newbie, age = " + seed.getAge());
-                    ranking.addScore(seed.hash, seedcount);
-                    regularSeeds.put(seed.hash, seed);
+                    regularSeeds.remove(seed.hash);
+                    ranking.deleteScore(seed.hash);
+                    matchingSeeds.put(seed.hash, seed);
                 }
             } else {
                 // this is a robinson peer
                 // in case the peer has more than a million urls, take it as search target
-                if (seed.getLinkCount() > 1000000) {
+                if (seed.getLinkCount() > 1000000 || seed.matchPeerTags(wordhashes)) {
                     regularSeeds.remove(seed.hash);
                     ranking.deleteScore(seed.hash);
-                    robinsonSeeds.put(seed.hash, seed);
+                    if (seed.matchPeerTags(wordhashes))
+                        matchingSeeds.put(seed.hash, seed);
+                    else
+                        robinsonSeeds.put(seed.hash, seed);
                 }
             }
         }
         
         // evaluate the ranking score and select seeds
         seedcount = Math.min(ranking.size(), seedcount);
-        yacySeed[] result = new yacySeed[seedcount + robinsonSeeds.size()];
+        yacySeed[] result = new yacySeed[seedcount + robinsonSeeds.size() + matchingSeeds.size()];
         c = 0;
         iter = ranking.scores(false); // higher are better
         while (iter.hasNext() && c < seedcount) {
             seed = regularSeeds.get(iter.next());
             seed.selectscore = c;
-            serverLog.logInfo("PLASMA", "selectPeers/_lineup_: " + seed.hash + ":" + seed.getName() + " is choice " + c);
+            serverLog.logInfo("PLASMA", "selectPeers/_dht_: " + seed.hash + ":" + seed.getName() + " is choice " + c);
             result[c++] = seed;
         }
         for (yacySeed s: robinsonSeeds.values()) {
             serverLog.logInfo("PLASMA", "selectPeers/_robinson_: " + s.hash + ":" + s.getName() + " is choice " + c);
             result[c++] = s;
         }
+        for (yacySeed s: matchingSeeds.values()) {
+            serverLog.logInfo("PLASMA", "selectPeers/_match_: " + s.hash + ":" + s.getName() + " is choice " + c);
+            result[c++] = s;
+        }
 
 //      System.out.println("DEBUG yacySearch.selectPeers = " + seedcount + " seeds:"); for (int i = 0; i < seedcount; i++) System.out.println(" #" + i + ":" + result[i]); // debug
         return result;

source/de/anomic/yacy/yacySeed.java

@@ -591,7 +591,9 @@ public class yacySeed {
     }
 
     public boolean matchPeerTags(Set<String> searchHashes) {
-        Set<String> tags = serverCodings.string2set(get(PEERTAGS, ""), "|");
+        String peertags = get(PEERTAGS, "");
+        if (peertags.equals("*")) return true;
+        Set<String> tags = serverCodings.string2set(peertags, "|");
         Iterator<String> i = tags.iterator();
         while (i.hasNext()) {
         	if (searchHashes.contains(indexWord.word2hash(i.next()))) return true;