honeymoose
/
OpenSearch
mirror of https://github.com/honeymoose/OpenSearch.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Improved XBooleanFilter by adding drive logic for bit based filter impl and adding unit test, which tests all possible XBooleanFilter options.

This commit is contained in:
Martijn van Groningen 2012-12-13 22:59:41 +01:00
parent 8f5bca43fb
commit 826a6ab02a
2 changed files with 261 additions and 12 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
src/main/java/org/elasticsearch/common/lucene/search/XBooleanFilter.java
View File

@ -165,15 +165,30 @@ public class XBooleanFilter extends Filter implements Iterable<FilterClause> {
continue; continue;
} }
if (clause.clause.getOccur() == Occur.SHOULD) { if (clause.clause.getOccur() == Occur.SHOULD) {
// TODO: we should let res drive it, and check on all unset bits on it with Bits if (res == null) {
DocIdSetIterator it = clause.docIdSet.iterator(); DocIdSetIterator it = clause.docIdSet.iterator();
if (it == null) { if (it == null) {
continue; continue;
} }
if (res == null) {
res = new FixedBitSet(reader.maxDoc()); res = new FixedBitSet(reader.maxDoc());
}
res.or(it); res.or(it);
} else {
Bits bits = clause.bits;
// use the "res" to drive the iteration
int lastSetDoc = res.nextSetBit(0);
DocIdSetIterator it = res.iterator();
for (int setDoc = it.nextDoc(); setDoc != DocIdSetIterator.NO_MORE_DOCS; setDoc = it.nextDoc()) {
int diff = setDoc - lastSetDoc;
if (diff > 1) {
for (int unsetDoc = lastSetDoc + 1; unsetDoc < setDoc; unsetDoc++) {
if (bits.get(unsetDoc)) {
res.set(unsetDoc);
}
}
}
lastSetDoc = setDoc;
}
}
} else if (clause.clause.getOccur() == Occur.MUST) { } else if (clause.clause.getOccur() == Occur.MUST) {
if (res == null) { if (res == null) {
// nothing we can do, just or it... // nothing we can do, just or it...
@ -217,12 +232,6 @@ public class XBooleanFilter extends Filter implements Iterable<FilterClause> {
return res; return res;
} }
private static DocIdSetIterator getDISI(Filter filter, AtomicReaderContext context, Bits acceptDocs)
throws IOException {
final DocIdSet set = filter.getDocIdSet(context, acceptDocs);
return (set == null || set == DocIdSet.EMPTY_DOCIDSET) ? null : set.iterator();
}
/** /**
* Adds a new FilterClause to the Boolean Filter container * Adds a new FilterClause to the Boolean Filter container
* *

240
src/test/java/org/elasticsearch/test/unit/common/lucene/search/XBooleanFilterTests.java Normal file
View File

@ -0,0 +1,240 @@
package org.elasticsearch.test.unit.common.lucene.search;
import org.apache.lucene.analysis.core.KeywordAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.document.StringField;
import org.apache.lucene.index.*;
import org.apache.lucene.queries.FilterClause;
import org.apache.lucene.search.BooleanClause;
import org.apache.lucene.search.FieldCacheTermsFilter;
import org.apache.lucene.search.Filter;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.RAMDirectory;
import org.apache.lucene.util.FixedBitSet;
import org.elasticsearch.common.lucene.Lucene;
import org.elasticsearch.common.lucene.search.TermFilter;
import org.elasticsearch.common.lucene.search.XBooleanFilter;
import org.testng.annotations.AfterClass;
import org.testng.annotations.BeforeClass;
import org.testng.annotations.Test;
import java.util.ArrayList;
import java.util.List;
import static org.apache.lucene.search.BooleanClause.Occur.*;
import static org.hamcrest.MatcherAssert.assertThat;
import static org.hamcrest.core.IsEqual.equalTo;
/**
*/
public class XBooleanFilterTests {
private Directory directory;
private AtomicReader reader;
@BeforeClass
public void setup() throws Exception {
char[][] documentMatrix = new char[][] {
{'a', 'b', 'c', 'd'},
{'a', 'b', 'c', 'd'},
{'a', 'a', 'a', 'a'}
};
List<Document> documents = new ArrayList<Document>(documentMatrix.length);
for (char[] fields : documentMatrix) {
Document document = new Document();
for (int i = 0; i < fields.length; i++) {
document.add(new StringField(Integer.toString(i), String.valueOf(fields[i]), Field.Store.NO));
}
documents.add(document);
}
directory = new RAMDirectory();
IndexWriter w = new IndexWriter(directory, new IndexWriterConfig(Lucene.VERSION, new KeywordAnalyzer()));
w.addDocuments(documents);
w.close();
reader = new SlowCompositeReaderWrapper(DirectoryReader.open(directory));
}
@AfterClass
public void tearDown() throws Exception {
reader.close();
directory.close();
}
@Test
public void testWithTwoClausesOfEachOccur_allFixedBitsetFilters() throws Exception {
List<XBooleanFilter> booleanFilters = new ArrayList<XBooleanFilter>();
booleanFilters.add(createBooleanFilter(
newFilterClause(0, 'a', MUST, false), newFilterClause(1, 'b', MUST, false),
newFilterClause(2, 'c', SHOULD, false), newFilterClause(3, 'd', SHOULD, false),
newFilterClause(4, 'e', MUST_NOT, false), newFilterClause(5, 'f', MUST_NOT, false)
));
booleanFilters.add(createBooleanFilter(
newFilterClause(4, 'e', MUST_NOT, false), newFilterClause(5, 'f', MUST_NOT, false),
newFilterClause(0, 'a', MUST, false), newFilterClause(1, 'b', MUST, false),
newFilterClause(2, 'c', SHOULD, false), newFilterClause(3, 'd', SHOULD, false)
));
booleanFilters.add(createBooleanFilter(
newFilterClause(2, 'c', SHOULD, false), newFilterClause(3, 'd', SHOULD, false),
newFilterClause(4, 'e', MUST_NOT, false), newFilterClause(5, 'f', MUST_NOT, false),
newFilterClause(0, 'a', MUST, false), newFilterClause(1, 'b', MUST, false)
));
for (XBooleanFilter booleanFilter : booleanFilters) {
FixedBitSet result = new FixedBitSet(reader.maxDoc());
result.or(booleanFilter.getDocIdSet(reader.getContext(), reader.getLiveDocs()).iterator());
assertThat(result.cardinality(), equalTo(2));
assertThat(result.get(0), equalTo(true));
assertThat(result.get(1), equalTo(true));
assertThat(result.get(2), equalTo(false));
}
}
@Test
public void testWithTwoClausesOfEachOccur_allBitsBasedFilters() throws Exception {
List<XBooleanFilter> booleanFilters = new ArrayList<XBooleanFilter>();
booleanFilters.add(createBooleanFilter(
newFilterClause(0, 'a', MUST, true), newFilterClause(1, 'b', MUST, true),
newFilterClause(2, 'c', SHOULD, true), newFilterClause(3, 'd', SHOULD, true),
newFilterClause(4, 'e', MUST_NOT, true), newFilterClause(5, 'f', MUST_NOT, true)
));
booleanFilters.add(createBooleanFilter(
newFilterClause(4, 'e', MUST_NOT, true), newFilterClause(5, 'f', MUST_NOT, true),
newFilterClause(0, 'a', MUST, true), newFilterClause(1, 'b', MUST, true),
newFilterClause(2, 'c', SHOULD, true), newFilterClause(3, 'd', SHOULD, true)
));
booleanFilters.add(createBooleanFilter(
newFilterClause(2, 'c', SHOULD, true), newFilterClause(3, 'd', SHOULD, true),
newFilterClause(4, 'e', MUST_NOT, true), newFilterClause(5, 'f', MUST_NOT, true),
newFilterClause(0, 'a', MUST, true), newFilterClause(1, 'b', MUST, true)
));
for (XBooleanFilter booleanFilter : booleanFilters) {
FixedBitSet result = new FixedBitSet(reader.maxDoc());
result.or(booleanFilter.getDocIdSet(reader.getContext(), reader.getLiveDocs()).iterator());
assertThat(result.cardinality(), equalTo(2));
assertThat(result.get(0), equalTo(true));
assertThat(result.get(1), equalTo(true));
assertThat(result.get(2), equalTo(false));
}
}
@Test
public void testWithTwoClausesOfEachOccur_allFilterTypes() throws Exception {
List<XBooleanFilter> booleanFilters = new ArrayList<XBooleanFilter>();
booleanFilters.add(createBooleanFilter(
newFilterClause(0, 'a', MUST, true), newFilterClause(1, 'b', MUST, false),
newFilterClause(2, 'c', SHOULD, true), newFilterClause(3, 'd', SHOULD, false),
newFilterClause(4, 'e', MUST_NOT, true), newFilterClause(5, 'f', MUST_NOT, false)
));
booleanFilters.add(createBooleanFilter(
newFilterClause(4, 'e', MUST_NOT, true), newFilterClause(5, 'f', MUST_NOT, false),
newFilterClause(0, 'a', MUST, true), newFilterClause(1, 'b', MUST, false),
newFilterClause(2, 'c', SHOULD, true), newFilterClause(3, 'd', SHOULD, false)
));
booleanFilters.add(createBooleanFilter(
newFilterClause(2, 'c', SHOULD, true), newFilterClause(3, 'd', SHOULD, false),
newFilterClause(4, 'e', MUST_NOT, true), newFilterClause(5, 'f', MUST_NOT, false),
newFilterClause(0, 'a', MUST, true), newFilterClause(1, 'b', MUST, false)
));
for (XBooleanFilter booleanFilter : booleanFilters) {
FixedBitSet result = new FixedBitSet(reader.maxDoc());
result.or(booleanFilter.getDocIdSet(reader.getContext(), reader.getLiveDocs()).iterator());
assertThat(result.cardinality(), equalTo(2));
assertThat(result.get(0), equalTo(true));
assertThat(result.get(1), equalTo(true));
assertThat(result.get(2), equalTo(false));
}
booleanFilters.clear();
booleanFilters.add(createBooleanFilter(
newFilterClause(0, 'a', MUST, false), newFilterClause(1, 'b', MUST, true),
newFilterClause(2, 'c', SHOULD, false), newFilterClause(3, 'd', SHOULD, true),
newFilterClause(4, 'e', MUST_NOT, false), newFilterClause(5, 'f', MUST_NOT, true)
));
booleanFilters.add(createBooleanFilter(
newFilterClause(4, 'e', MUST_NOT, false), newFilterClause(5, 'f', MUST_NOT, true),
newFilterClause(0, 'a', MUST, false), newFilterClause(1, 'b', MUST, true),
newFilterClause(2, 'c', SHOULD, false), newFilterClause(3, 'd', SHOULD, true)
));
booleanFilters.add(createBooleanFilter(
newFilterClause(2, 'c', SHOULD, false), newFilterClause(3, 'd', SHOULD, true),
newFilterClause(4, 'e', MUST_NOT, false), newFilterClause(5, 'f', MUST_NOT, true),
newFilterClause(0, 'a', MUST, false), newFilterClause(1, 'b', MUST, true)
));
for (XBooleanFilter booleanFilter : booleanFilters) {
FixedBitSet result = new FixedBitSet(reader.maxDoc());
result.or(booleanFilter.getDocIdSet(reader.getContext(), reader.getLiveDocs()).iterator());
assertThat(result.cardinality(), equalTo(2));
assertThat(result.get(0), equalTo(true));
assertThat(result.get(1), equalTo(true));
assertThat(result.get(2), equalTo(false));
}
}
@Test
public void testWithTwoClausesOfEachOccur_singleClauseOptimisation() throws Exception {
List<XBooleanFilter> booleanFilters = new ArrayList<XBooleanFilter>();
booleanFilters.add(createBooleanFilter(
newFilterClause(1, 'b', MUST, true)
));
for (XBooleanFilter booleanFilter : booleanFilters) {
FixedBitSet result = new FixedBitSet(reader.maxDoc());
result.or(booleanFilter.getDocIdSet(reader.getContext(), reader.getLiveDocs()).iterator());
assertThat(result.cardinality(), equalTo(2));
assertThat(result.get(0), equalTo(true));
assertThat(result.get(1), equalTo(true));
assertThat(result.get(2), equalTo(false));
}
booleanFilters.clear();
booleanFilters.add(createBooleanFilter(
newFilterClause(1, 'c', MUST_NOT, true)
));
for (XBooleanFilter booleanFilter : booleanFilters) {
FixedBitSet result = new FixedBitSet(reader.maxDoc());
result.or(booleanFilter.getDocIdSet(reader.getContext(), reader.getLiveDocs()).iterator());
assertThat(result.cardinality(), equalTo(3));
assertThat(result.get(0), equalTo(true));
assertThat(result.get(1), equalTo(true));
assertThat(result.get(2), equalTo(true));
}
booleanFilters.clear();
booleanFilters.add(createBooleanFilter(
newFilterClause(2, 'c', SHOULD, true)
));
for (XBooleanFilter booleanFilter : booleanFilters) {
FixedBitSet result = new FixedBitSet(reader.maxDoc());
result.or(booleanFilter.getDocIdSet(reader.getContext(), reader.getLiveDocs()).iterator());
assertThat(result.cardinality(), equalTo(2));
assertThat(result.get(0), equalTo(true));
assertThat(result.get(1), equalTo(true));
assertThat(result.get(2), equalTo(false));
}
}
private static FilterClause newFilterClause(int field, char character, BooleanClause.Occur occur, boolean slowerBitsBackedFilter) {
Filter filter;
if (slowerBitsBackedFilter) {
filter = new FieldCacheTermsFilter(String.valueOf(field), String.valueOf(character));
} else {
Term term = new Term(String.valueOf(field), String.valueOf(character));
filter = new TermFilter(term);
}
return new FilterClause(filter, occur);
}
private static XBooleanFilter createBooleanFilter(FilterClause... clauses) {
XBooleanFilter booleanFilter = new XBooleanFilter();
for (FilterClause clause : clauses) {
booleanFilter.add(clause);
}
return booleanFilter;
}
}