When this parameter becomes a bottleneck for a Query one can use a - * Filter. For example instead of a {@link RangeQuery} one can use a + * MatchFilter. For example instead of a {@link RangeQuery} one can use a * {@link RangeFilter}. *
Normally the buffers are allocated by the JVM. When using for example
* {@link org.apache.lucene.store.MMapDirectory} the buffering is left to
Index: src/java/org/apache/lucene/search/ConstantScoreQuery.java
===================================================================
--- src/java/org/apache/lucene/search/ConstantScoreQuery.java (revision 573048)
+++ src/java/org/apache/lucene/search/ConstantScoreQuery.java (working copy)
@@ -20,7 +20,6 @@
import org.apache.lucene.index.IndexReader;
import java.io.IOException;
-import java.util.BitSet;
import java.util.Set;
/**
@@ -31,14 +30,14 @@
* @version $Id$
*/
public class ConstantScoreQuery extends Query {
- protected final Filter filter;
+ protected final MatchFilter filter;
- public ConstantScoreQuery(Filter filter) {
+ public ConstantScoreQuery(MatchFilter filter) {
this.filter=filter;
}
/** Returns the encapsulated filter */
- public Filter getFilter() {
+ public MatchFilter getFilter() {
return filter;
}
@@ -85,7 +84,7 @@
public Explanation explain(IndexReader reader, int doc) throws IOException {
ConstantScorer cs = (ConstantScorer)scorer(reader);
- boolean exists = cs.bits.get(doc);
+ boolean exists = cs.matcher.skipTo(doc) && (cs.matcher.doc() == doc);
ComplexExplanation result = new ComplexExplanation();
@@ -107,23 +106,21 @@
}
protected class ConstantScorer extends Scorer {
- final BitSet bits;
+ final Matcher matcher;
final float theScore;
- int doc=-1;
public ConstantScorer(Similarity similarity, IndexReader reader, Weight w) throws IOException {
super(similarity);
theScore = w.getValue();
- bits = filter.bits(reader);
+ matcher = filter.getMatcher(reader);
}
public boolean next() throws IOException {
- doc = bits.nextSetBit(doc+1);
- return doc >= 0;
+ return matcher.next();
}
public int doc() {
- return doc;
+ return matcher.doc();
}
public float score() throws IOException {
@@ -131,8 +128,7 @@
}
public boolean skipTo(int target) throws IOException {
- doc = bits.nextSetBit(target); // requires JDK 1.4
- return doc >= 0;
+ return matcher.skipTo(target);
}
public Explanation explain(int doc) throws IOException {
Index: src/java/org/apache/lucene/search/FieldDoc.java
===================================================================
--- src/java/org/apache/lucene/search/FieldDoc.java (revision 573048)
+++ src/java/org/apache/lucene/search/FieldDoc.java (working copy)
@@ -47,7 +47,7 @@
* Sort object. Each Object will be either an Integer, Float or String,
* depending on the type of values in the terms of the original field.
* @see Sort
- * @see Searcher#search(Query,Filter,int,Sort)
+ * @see Searcher#search(Query,MatchFilter,int,Sort)
*/
public Comparable[] fields;
Index: src/java/org/apache/lucene/search/FieldSortedHitQueue.java
===================================================================
--- src/java/org/apache/lucene/search/FieldSortedHitQueue.java (revision 573048)
+++ src/java/org/apache/lucene/search/FieldSortedHitQueue.java (working copy)
@@ -33,7 +33,7 @@
* @author Tim Jones (Nacimiento Software)
* @since lucene 1.4
* @version $Id$
- * @see Searcher#search(Query,Filter,int,Sort)
+ * @see Searcher#search(Query,MatchFilter,int,Sort)
* @see FieldCache
*/
public class FieldSortedHitQueue
@@ -127,7 +127,7 @@
* by a MultiSearcher with other search hits.
* @param doc The FieldDoc to store sort values into.
* @return The same FieldDoc passed in.
- * @see Searchable#search(Weight,Filter,int,Sort)
+ * @see Searchable#search(Weight,MatchFilter,int,Sort)
*/
FieldDoc fillFields (final FieldDoc doc) {
final int n = comparators.length;
Index: src/java/org/apache/lucene/search/FilteredQuery.java
===================================================================
--- src/java/org/apache/lucene/search/FilteredQuery.java (revision 573048)
+++ src/java/org/apache/lucene/search/FilteredQuery.java (working copy)
@@ -21,7 +21,6 @@
import org.apache.lucene.util.ToStringUtils;
import java.io.IOException;
-import java.util.BitSet;
import java.util.Set;
@@ -43,15 +42,15 @@
extends Query {
Query query;
- Filter filter;
+ MatchFilter filter;
/**
* Constructs a new query which applies a filter to the results of the original query.
- * Filter.bits() will be called every time this query is used in a search.
+ * MatchFilter.getMatcher() will be called every time this query is used in a search.
* @param query Query to be filtered, cannot be null.
- * @param filter Filter to apply to query results, cannot be null.
+ * @param filter MatchFilter to apply to query results, cannot be null.
*/
- public FilteredQuery (Query query, Filter filter) {
+ public FilteredQuery (Query query, MatchFilter filter) {
this.query = query;
this.filter = filter;
}
@@ -85,14 +84,16 @@
inner.addDetail(new Explanation(getBoost(),"boost"));
inner.addDetail(preBoost);
}
- Filter f = FilteredQuery.this.filter;
- BitSet matches = f.bits(ir);
- if (matches.get(i))
+ MatchFilter f = FilteredQuery.this.filter;
+ Matcher matcher = f.getMatcher(ir);
+ if (matcher.skipTo(i) && (matcher.doc() == i)) {
return inner;
- Explanation result = new Explanation
- (0.0f, "failure to match filter: " + f.toString());
- result.addDetail(inner);
- return result;
+ } else {
+ Explanation result = new Explanation
+ (0.0f, "failure to match filter: " + f.toString());
+ result.addDetail(inner);
+ return result;
+ }
}
// return this query
@@ -100,50 +101,49 @@
// return a filtering scorer
public Scorer scorer (IndexReader indexReader) throws IOException {
- final Scorer scorer = weight.scorer (indexReader);
- final BitSet bitset = filter.bits (indexReader);
- return new Scorer (similarity) {
+ final Scorer scorer = weight.scorer(indexReader);
+ final Matcher matcher = filter.getMatcher(indexReader);
- public boolean next() throws IOException {
- do {
- if (! scorer.next()) {
+ return new Scorer(similarity) {
+
+ private boolean advanceToCommon() throws IOException {
+ while (scorer.doc() != matcher.doc()) {
+ if (scorer.doc() < matcher.doc()) {
+ if (!scorer.skipTo(matcher.doc())) {
+ return false;
+ }
+ } else if (!matcher.skipTo(scorer.doc())) {
return false;
}
- } while (! bitset.get(scorer.doc()));
- /* When skipTo() is allowed on scorer it should be used here
- * in combination with bitset.nextSetBit(...)
- * See the while loop in skipTo() below.
- */
+ }
return true;
}
+
+ public boolean next() throws IOException {
+ return matcher.next() && scorer.next() && advanceToCommon();
+ }
+
public int doc() { return scorer.doc(); }
public boolean skipTo(int i) throws IOException {
- if (! scorer.skipTo(i)) {
- return false;
- }
- while (! bitset.get(scorer.doc())) {
- int nextFiltered = bitset.nextSetBit(scorer.doc() + 1);
- if (nextFiltered == -1) {
- return false;
- } else if (! scorer.skipTo(nextFiltered)) {
- return false;
- }
- }
- return true;
- }
+ return matcher.skipTo(i)
+ && scorer.skipTo(matcher.doc())
+ && advanceToCommon();
+ }
public float score() throws IOException { return getBoost() * scorer.score(); }
// add an explanation about whether the document was filtered
public Explanation explain (int i) throws IOException {
- Explanation exp = scorer.explain (i);
- exp.setValue(getBoost() * exp.getValue());
+ Explanation exp = scorer.explain(i);
- if (bitset.get(i))
+ if (matcher.skipTo(i) && (matcher.doc() == i)) {
exp.setDescription ("allowed by filter: "+exp.getDescription());
- else
+ exp.setValue(getBoost() * exp.getValue());
+ } else {
exp.setDescription ("removed by filter: "+exp.getDescription());
+ exp.setValue(0.0f);
+ }
return exp;
}
};
@@ -167,7 +167,7 @@
return query;
}
- public Filter getFilter() {
+ public MatchFilter getFilter() {
return filter;
}
Index: src/java/org/apache/lucene/search/Hits.java
===================================================================
--- src/java/org/apache/lucene/search/Hits.java (revision 573048)
+++ src/java/org/apache/lucene/search/Hits.java (working copy)
@@ -35,7 +35,7 @@
public final class Hits {
private Weight weight;
private Searcher searcher;
- private Filter filter = null;
+ private MatchFilter filter = null;
private Sort sort = null;
private int length; // the total number of hits
@@ -46,14 +46,14 @@
private int numDocs = 0; // number cached
private int maxDocs = 200; // max to cache
- Hits(Searcher s, Query q, Filter f) throws IOException {
+ Hits(Searcher s, Query q, MatchFilter f) throws IOException {
weight = q.weight(s);
searcher = s;
filter = f;
getMoreDocs(50); // retrieve 100 initially
}
- Hits(Searcher s, Query q, Filter f, Sort o) throws IOException {
+ Hits(Searcher s, Query q, MatchFilter f, Sort o) throws IOException {
weight = q.weight(s);
searcher = s;
filter = f;
Index: src/java/org/apache/lucene/search/QueryWrapperFilter.java
===================================================================
--- src/java/org/apache/lucene/search/QueryWrapperFilter.java (revision 573048)
+++ src/java/org/apache/lucene/search/QueryWrapperFilter.java (working copy)
@@ -31,8 +31,6 @@
* QueryFilter that matches, e.g., only documents modified within the last
* week. The QueryFilter and RangeQuery would only need to be reconstructed
* once per day.
- *
- * @version $Id:$
*/
public class QueryWrapperFilter extends Filter {
private Query query;
@@ -43,13 +41,12 @@
public QueryWrapperFilter(Query query) {
this.query = query;
}
-
+
public BitSet bits(IndexReader reader) throws IOException {
final BitSet bits = new BitSet(reader.maxDoc());
-
- new IndexSearcher(reader).search(query, new HitCollector() {
- public final void collect(int doc, float score) {
- bits.set(doc); // set bit for hit
+ new IndexSearcher(reader).match(query, new MatchCollector() {
+ public final void collect(int doc) {
+ bits.set(doc);
}
});
return bits;
Index: src/java/org/apache/lucene/search/CachingWrapperFilter.java
===================================================================
--- src/java/org/apache/lucene/search/CachingWrapperFilter.java (revision 573048)
+++ src/java/org/apache/lucene/search/CachingWrapperFilter.java (working copy)
@@ -17,12 +17,15 @@
* limitations under the License.
*/
-import org.apache.lucene.index.IndexReader;
import java.util.BitSet;
import java.util.WeakHashMap;
import java.util.Map;
import java.io.IOException;
+import org.apache.lucene.index.IndexReader;
+import org.apache.lucene.util.DefaultMatcher;
+import org.apache.lucene.util.MatcherProvider;
+
/**
* Wraps another filter's result and caches it. The purpose is to allow
* filters to simply filter, and then wrap with this class to add caching.
@@ -44,24 +47,31 @@
}
public BitSet bits(IndexReader reader) throws IOException {
- if (cache == null) {
- cache = new WeakHashMap();
- }
+ return filter.bits(reader);
+ }
+
+ public Matcher getMatcher(IndexReader reader) throws IOException {
+ MatcherProvider cachedMP = null;
- synchronized (cache) { // check cache
- BitSet cached = (BitSet) cache.get(reader);
- if (cached != null) {
- return cached;
+ synchronized (this) {
+ if (cache == null) {
+ cache = new WeakHashMap();
+ } else {
+ cachedMP = (MatcherProvider) cache.get(reader);
}
}
- final BitSet bits = filter.bits(reader);
+ if (cachedMP != null) {
+ return cachedMP.getMatcher();
+ }
- synchronized (cache) { // update cache
- cache.put(reader, bits);
+ cachedMP = DefaultMatcher.makeMatcherProvider(bits(reader));
+
+ synchronized (this) { // update cache
+ cache.put(reader, cachedMP);
}
- return bits;
+ return cachedMP.getMatcher();
}
public String toString() {
Index: src/test/org/apache/lucene/search/CheckHits.java
===================================================================
--- src/test/org/apache/lucene/search/CheckHits.java (revision 573048)
+++ src/test/org/apache/lucene/search/CheckHits.java (working copy)
@@ -377,7 +377,7 @@
new ExplanationAsserter
(q, null, this));
}
- public Hits search(Query query, Filter filter) throws IOException {
+ public Hits search(Query query, MatchFilter filter) throws IOException {
checkExplanations(query);
return super.search(query,filter);
}
@@ -385,13 +385,13 @@
checkExplanations(query);
return super.search(query,sort);
}
- public Hits search(Query query, Filter filter,
+ public Hits search(Query query, MatchFilter filter,
Sort sort) throws IOException {
checkExplanations(query);
return super.search(query,filter,sort);
}
public TopFieldDocs search(Query query,
- Filter filter,
+ MatchFilter filter,
int n,
Sort sort) throws IOException {
@@ -402,12 +402,12 @@
checkExplanations(query);
super.search(query,results);
}
- public void search(Query query, Filter filter,
+ public void search(Query query, MatchFilter filter,
HitCollector results) throws IOException {
checkExplanations(query);
super.search(query,filter, results);
}
- public TopDocs search(Query query, Filter filter,
+ public TopDocs search(Query query, MatchFilter filter,
int n) throws IOException {
checkExplanations(query);
Index: src/test/org/apache/lucene/search/TestCustomSearcherSort.java
===================================================================
--- src/test/org/apache/lucene/search/TestCustomSearcherSort.java (revision 573048)
+++ src/test/org/apache/lucene/search/TestCustomSearcherSort.java (working copy)
@@ -252,9 +252,9 @@
this.switcher = switcher;
}
/* (non-Javadoc)
- * @see org.apache.lucene.search.Searchable#search(org.apache.lucene.search.Query, org.apache.lucene.search.Filter, int, org.apache.lucene.search.Sort)
+ * @see org.apache.lucene.search.Searchable#search(org.apache.lucene.search.Query, org.apache.lucene.search.MatchFilter, int, org.apache.lucene.search.Sort)
*/
- public TopFieldDocs search(Query query, Filter filter, int nDocs,
+ public TopFieldDocs search(Query query, MatchFilter filter, int nDocs,
Sort sort) throws IOException {
BooleanQuery bq = new BooleanQuery();
bq.add(query, BooleanClause.Occur.MUST);
@@ -262,9 +262,9 @@
return super.search(bq, filter, nDocs, sort);
}
/* (non-Javadoc)
- * @see org.apache.lucene.search.Searchable#search(org.apache.lucene.search.Query, org.apache.lucene.search.Filter, int)
+ * @see org.apache.lucene.search.Searchable#search(org.apache.lucene.search.Query, org.apache.lucene.search.MatchFilter, int)
*/
- public TopDocs search(Query query, Filter filter, int nDocs)
+ public TopDocs search(Query query, MatchFilter filter, int nDocs)
throws IOException {
BooleanQuery bq = new BooleanQuery();
bq.add(query, BooleanClause.Occur.MUST);
Index: src/test/org/apache/lucene/search/TestRemoteCachingWrapperFilter.java
===================================================================
--- src/test/org/apache/lucene/search/TestRemoteCachingWrapperFilter.java (revision 573048)
+++ src/test/org/apache/lucene/search/TestRemoteCachingWrapperFilter.java (working copy)
@@ -77,7 +77,7 @@
Naming.rebind("//localhost/Searchable", impl);
}
- private static void search(Query query, Filter filter, int hitNumber, String typeValue) throws Exception {
+ private static void search(Query query, MatchFilter filter, int hitNumber, String typeValue) throws Exception {
Searchable[] searchables = { getRemote() };
Searcher searcher = new MultiSearcher(searchables);
Hits result = searcher.search(query,filter);
Index: src/test/org/apache/lucene/search/TestCachingWrapperFilter.java
===================================================================
--- src/test/org/apache/lucene/search/TestCachingWrapperFilter.java (revision 573048)
+++ src/test/org/apache/lucene/search/TestCachingWrapperFilter.java (working copy)
@@ -36,12 +36,12 @@
CachingWrapperFilter cacher = new CachingWrapperFilter(filter);
// first time, nested filter is called
- cacher.bits(reader);
+ cacher.getMatcher(reader);
assertTrue("first time", filter.wasCalled());
// second time, nested filter should not be called
filter.clear();
- cacher.bits(reader);
+ cacher.getMatcher(reader);
assertFalse("second time", filter.wasCalled());
reader.close();