The upgrade is 99% finished

romanchyla · Oct 5, 2016 · d76da1b · d76da1b
1 parent 7ea5cb8
commit d76da1b
Show file tree

Hide file tree

Showing 12 changed files with 189 additions and 239 deletions.
diff --git a/contrib/adsabs/src/java/org/apache/lucene/search/AbstractSecondOrderCollector.java b/contrib/adsabs/src/java/org/apache/lucene/search/AbstractSecondOrderCollector.java
@@ -23,7 +23,7 @@ public abstract class AbstractSecondOrderCollector implements Collector, LeafCol
 	private Integer lastPos = null;
 	protected float ensureCapacityRatio = 0.25f;
 	protected FinalValueType compactingType = FinalValueType.MAX_VALUE;
-  private LeafReaderContext context;
+  protected LeafReaderContext context;
 
 	public AbstractSecondOrderCollector() {
 		lock = new ReentrantLock();

diff --git a/contrib/adsabs/src/java/org/apache/lucene/search/BasedBitSetIterator.java b/contrib/adsabs/src/java/org/apache/lucene/search/BasedBitSetIterator.java
@@ -7,40 +7,55 @@
  * Iterator which works with the knowledge of index segment boundaries.
  * It returns un-based document ids.
  * 
+ * Internal bitset iterator works with bounded ranges. We remove/add 
+ * bases before returning the results.
+ * 
  */
-public class BasedBitSetIterator extends BitSetIterator {
+public class BasedBitSetIterator extends DocIdSetIterator {
   private int base;
-  private int v;
+  private int doc = -1;
+  private int upperBound;
+  private BitSet bits;
+  private int length;
+  private long cost;
 
-  public BasedBitSetIterator(BitSet bits, long cost, int base) {
-    super(bits, cost);
+  public BasedBitSetIterator(BitSet bits, long cost, int base, int size) {
+    if (cost < 0) {
+      throw new IllegalArgumentException("cost must be >= 0, got " + cost);
+    }
+    this.bits = bits;
+    this.length = bits.length();
+    this.cost = cost;
     this.base = base;
+    this.upperBound = base + size;
+    if (upperBound > length)
+      upperBound = length;
   }
 
   @Override
   public int docID() {
-    v = super.docID();
-    if (v != NO_MORE_DOCS)
-      return v - base;
-    return v;
+    return doc;
   }
 
   @Override
   public int nextDoc() {
-    v = super.nextDoc();
-    if (v != NO_MORE_DOCS)
-      return v - base;
-    return v;
+    return advance(doc + 1);
   }
 
   @Override
   public int advance(int target) {
-    v = super.advance(target + base);
-    if (v != NO_MORE_DOCS)
-      return v - base;
-    return v;
+    if (target+base >= upperBound) {
+      return doc = NO_MORE_DOCS;
+    }
+    doc = bits.nextSetBit(target+base);
+    if (doc >= upperBound)
+      return doc = NO_MORE_DOCS;
+    return doc = doc - base;
   }
-
-
-
-}
+
+  @Override
+  public long cost() {
+    return cost;
+  }
+
+}
diff --git a/contrib/adsabs/src/java/org/apache/lucene/search/BitSetQuery.java b/contrib/adsabs/src/java/org/apache/lucene/search/BitSetQuery.java
@@ -24,7 +24,9 @@ public Weight createWeight(IndexSearcher searcher, boolean needsScores) throws I
       return new ConstantScoreWeight(this) {
         @Override
         public Scorer scorer(LeafReaderContext context) throws IOException {
-          return new ConstantScoreScorer(this, score(), new BasedBitSetIterator(docs, docs.approximateCardinality(), context.docBase));
+          return new ConstantScoreScorer(this, score(), 
+              new BasedBitSetIterator(docs, docs.approximateCardinality(), 
+                  context.docBase, context.reader().maxDoc()));
         }
       };
     }

diff --git a/contrib/adsabs/src/java/org/apache/lucene/search/SecondOrderCollectorCitedBy.java b/contrib/adsabs/src/java/org/apache/lucene/search/SecondOrderCollectorCitedBy.java
@@ -21,19 +21,7 @@ public SecondOrderCollectorCitedBy(SolrCacheWrapper cache) {
 
 
 	@Override
-	public boolean searcherInitialization(IndexSearcher searcher, Weight firstOrderWeight) throws IOException {
-		return super.searcherInitialization(searcher, firstOrderWeight);
-	}
-
-
-	@Override
-	public void setScorer(Scorer scorer) throws IOException {
-		this.scorer = scorer;
-
-	}
-
-	@Override
-	public void collect(int doc) throws IOException {
+  public void collect(int doc) throws IOException {
 		int[] v = cache.getLuceneDocIds(doc+docBase);
 		if (v == null) return;
 		float s = scorer.score();

diff --git a/contrib/adsabs/src/java/org/apache/lucene/search/SecondOrderCollectorCites.java b/contrib/adsabs/src/java/org/apache/lucene/search/SecondOrderCollectorCites.java
@@ -45,16 +45,14 @@ public SecondOrderCollectorCites(SolrCacheWrapper cache, String[] referenceField
 	@SuppressWarnings("unchecked")
   @Override
 	public boolean searcherInitialization(IndexSearcher searcher, Weight firstOrderWeight) throws IOException {
-		reader = searcher.getIndexReader();
 		return super.searcherInitialization(searcher, firstOrderWeight);
 	}
 
 
 	@Override
 	public void collect(int doc) throws IOException {
 		//if (reader.isDeleted(doc)) return;
-
-		Document document = reader.document(doc, fieldsToLoad);
+		Document document = this.context.reader().document(doc, fieldsToLoad);
 		float s = scorer.score();
 
 		for (String f: fieldsToLoad) {

diff --git a/contrib/adsabs/src/java/org/apache/lucene/search/SecondOrderCollectorCitesRAM.java b/contrib/adsabs/src/java/org/apache/lucene/search/SecondOrderCollectorCitesRAM.java
@@ -17,13 +17,6 @@ public SecondOrderCollectorCitesRAM(SolrCacheWrapper cache) {
   }
 
 
-	@SuppressWarnings("unchecked")
-  @Override
-	public boolean searcherInitialization(IndexSearcher searcher, Weight firstOrderWeight) throws IOException {
-		return super.searcherInitialization(searcher, firstOrderWeight);
-	}
-
-
 	@Override
 	public void collect(int doc) throws IOException {
 		int[] citations = cache.getLuceneDocIds(doc+docBase);

diff --git a/contrib/adsabs/src/java/org/apache/solr/search/BitSetQParserPlugin.java b/contrib/adsabs/src/java/org/apache/solr/search/BitSetQParserPlugin.java
@@ -20,6 +20,7 @@
 import org.apache.lucene.index.NumericDocValues;
 import org.apache.lucene.queryparser.classic.ParseException;
 import org.apache.lucene.search.BitSetQuery;
+import org.apache.lucene.search.DocIdSetIterator;
 import org.apache.lucene.search.MatchNoDocsQuery;
 import org.apache.lucene.search.Query;
 import org.apache.lucene.search.SolrCacheWrapper;
@@ -285,7 +286,7 @@ else if (c.isAssignableFrom(TrieIntField.class)) {
 
     					SolrCacheWrapper<SolrCache<Object,Integer>> cacheWrapper = super.getCache(fieldName);
     					if (cacheWrapper != null) { // we are lucky and we have a cache that can translate values for us
-    						for (int i = bits.nextSetBit(0); i >= 0; i = bits.nextSetBit(i+1)) {
+    						for (int i = bits.nextSetBit(0); i >= 0 && i < DocIdSetIterator.NO_MORE_DOCS; i = bits.nextSetBit(i+1)) {
     					     if (fieldIsInt) {
     					    	 int v = cacheWrapper.getLuceneDocId(0, i);
     					    	 if (v == -1)
@@ -542,7 +543,7 @@ private byte[] unZip(byte[] data) throws IOException {
 
 	protected byte[] toByteArray(BitSet bitSet) {
 		byte[] bytes = new byte[(bitSet.length() + 7) / 8];
-		for ( int i = bitSet.nextSetBit(0); i >= 0 && i < Integer.MAX_VALUE; i = bitSet.nextSetBit(i+1) ) {
+		for ( int i = bitSet.nextSetBit(0); i >= 0 && i < DocIdSetIterator.NO_MORE_DOCS; i = bitSet.nextSetBit(i+1) ) {
 			bytes[i / 8] |= 128 >> (i % 8);
 			if (i+1 >= bitSet.length())
 			  break;

diff --git a/contrib/adsabs/src/java/org/apache/solr/search/CitationLRUCache.java b/contrib/adsabs/src/java/org/apache/solr/search/CitationLRUCache.java
@@ -630,9 +630,8 @@ private void unInvertedTheDamnThing(
       List<String> fields, 
       KVSetter setter) throws IOException {
 
-    LeafReader reader = searcher.getLeafReader();
   	IndexSchema schema = searcher.getCore().getLatestSchema();
-  	List<LeafReaderContext> leaves = reader.getContext().leaves();
+  	List<LeafReaderContext> leaves = searcher.getIndexReader().getContext().leaves();
 
   	Bits liveDocs;
   	LeafReader lr;
@@ -642,7 +641,6 @@ private void unInvertedTheDamnThing(
   	  liveDocs = leave.reader().getLiveDocs();
   	  lr = leave.reader();
   	  FieldInfos fInfo = lr.getFieldInfos();
-
   	  for (String field: fields) {
 
   	    FieldInfo fi = fInfo.fieldInfo(field);
@@ -994,7 +992,7 @@ public void addReference(int sourceDocid, Object value) {
     		addReference(sourceDocid, (Integer) this.get(value));
     	}
     	else {
-    		addReference(sourceDocid, -1);
+    		//addReference(sourceDocid, -1);
     	}
     }
     public void addReference(int sourceDocid, Integer targetDocid) {
@@ -1007,7 +1005,7 @@ public void addCitation(int sourceDocid, Object value) {
     		addCitation(sourceDocid, (Integer) this.get(value));
     	}
     	else {
-    		addCitation(sourceDocid, -1);
+    		//addCitation(sourceDocid, -1);
     	}
     }
 

diff --git a/contrib/adsabs/src/test/org/adsabs/lucene/BenchmarkAuthorSearch.java b/contrib/adsabs/src/test/org/adsabs/lucene/BenchmarkAuthorSearch.java
@@ -458,6 +458,8 @@ private void verifySearch(int[] randomIds) throws IOException {
 	}
 
 	private Query[] buildQueries(String[] parts) throws UnsupportedEncodingException {
+	  if (parts.length - 1 < 3)
+	    return null;
 		int howMany = TestUtil.nextInt(random(), 2, parts.length-1); // how many initials
 		if (howMany < 2)
 		  return null;