package org.apache.lucene.index;
/*
* Licensed to the Apache Software Foundation (ASF) under one or more
* contributor license agreements. See the NOTICE file distributed with
* this work for additional information regarding copyright ownership.
* The ASF licenses this file to You under the Apache License, Version 2.0
* (the "License"); you may not use this file except in compliance with
* the License. You may obtain a copy of the License at
*
* http://www.apache.org/licenses/LICENSE-2.0
*
* Unless required by applicable law or agreed to in writing, software
* distributed under the License is distributed on an "AS IS" BASIS,
* WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
* See the License for the specific language governing permissions and
* limitations under the License.
*/
import java.io.IOException;
import java.lang.reflect.Method;
import java.lang.reflect.Modifier;
import org.apache.lucene.analysis.MockAnalyzer;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.search.DocIdSetIterator;
import org.apache.lucene.store.Directory;
import org.apache.lucene.store.BaseDirectoryWrapper;
import org.apache.lucene.util.Bits;
import org.apache.lucene.util.BytesRef;
import org.apache.lucene.util.LuceneTestCase;
public class TestFilterAtomicReader extends LuceneTestCase {
private static class TestReader extends FilterAtomicReader {
/** Filter that only permits terms containing 'e'.*/
private static class TestFields extends FilterFields {
TestFields(Fields in) {
super(in);
}
@Override
public Terms terms(String field) throws IOException {
return new TestTerms(super.terms(field));
}
}
private static class TestTerms extends FilterTerms {
TestTerms(Terms in) {
super(in);
}
@Override
public TermsEnum iterator(TermsEnum reuse) throws IOException {
return new TestTermsEnum(super.iterator(reuse));
}
}
private static class TestTermsEnum extends FilterTermsEnum {
public TestTermsEnum(TermsEnum in) {
super(in);
}
/** Scan for terms containing the letter 'e'.*/
@Override
public BytesRef next() throws IOException {
BytesRef text;
while ((text = in.next()) != null) {
if (text.utf8ToString().indexOf('e') != -1)
return text;
}
return null;
}
@Override
public DocsAndPositionsEnum docsAndPositions(Bits liveDocs, DocsAndPositionsEnum reuse, int flags) throws IOException {
return new TestPositions(super.docsAndPositions(liveDocs, reuse == null ? null : ((FilterDocsAndPositionsEnum) reuse).in, flags));
}
}
/** Filter that only returns odd numbered documents. */
private static class TestPositions extends FilterDocsAndPositionsEnum {
public TestPositions(DocsAndPositionsEnum in) {
super(in);
}
/** Scan for odd numbered documents. */
@Override
public int nextDoc() throws IOException {
int doc;
while ((doc = in.nextDoc()) != NO_MORE_DOCS) {
if ((doc % 2) == 1)
return doc;
}
return NO_MORE_DOCS;
}
}
public TestReader(IndexReader reader) throws IOException {
super(SlowCompositeReaderWrapper.wrap(reader));
}
@Override
public Fields fields() throws IOException {
return new TestFields(super.fields());
}
}
/**
* Tests the IndexReader.getFieldNames implementation
* @throws Exception on error
*/
public void testFilterIndexReader() throws Exception {
Directory directory = newDirectory();
IndexWriter writer = new IndexWriter(directory, newIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(random())));
Document d1 = new Document();
d1.add(newTextField("default", "one two", Field.Store.YES));
writer.addDocument(d1);
Document d2 = new Document();
d2.add(newTextField("default", "one three", Field.Store.YES));
writer.addDocument(d2);
Document d3 = new Document();
d3.add(newTextField("default", "two four", Field.Store.YES));
writer.addDocument(d3);
writer.close();
Directory target = newDirectory();
// We mess with the postings so this can fail:
((BaseDirectoryWrapper) target).setCrossCheckTermVectorsOnClose(false);
writer = new IndexWriter(target, newIndexWriterConfig(TEST_VERSION_CURRENT, new MockAnalyzer(random())));
IndexReader reader = new TestReader(DirectoryReader.open(directory));
writer.addIndexes(reader);
writer.close();
reader.close();
reader = DirectoryReader.open(target);
TermsEnum terms = MultiFields.getTerms(reader, "default").iterator(null);
while (terms.next() != null) {
assertTrue(terms.term().utf8ToString().indexOf('e') != -1);
}
assertEquals(TermsEnum.SeekStatus.FOUND, terms.seekCeil(new BytesRef("one")));
DocsAndPositionsEnum positions = terms.docsAndPositions(MultiFields.getLiveDocs(reader), null);
while (positions.nextDoc() != DocIdSetIterator.NO_MORE_DOCS) {
assertTrue((positions.docID() % 2) == 1);
}
reader.close();
directory.close();
target.close();
}
private void checkOverrideMethods(Class<?> clazz) {
boolean fail = false;
for (Method m : clazz.getMethods()) {
int mods = m.getModifiers();
if (Modifier.isStatic(mods) || Modifier.isFinal(mods) || m.isSynthetic()) {
continue;
}
Class<?> declaringClass = m.getDeclaringClass();
if (declaringClass != clazz && declaringClass != Object.class) {
System.err.println("method is not overridden by "+clazz.getName()+": " + m.toGenericString());
fail = true;
}
}
assertFalse(clazz.getName()+" does not override some methods; see log above", fail);
}
public void testOverrideMethods() throws Exception {
checkOverrideMethods(FilterAtomicReader.class);
checkOverrideMethods(FilterAtomicReader.FilterFields.class);
checkOverrideMethods(FilterAtomicReader.FilterTerms.class);
checkOverrideMethods(FilterAtomicReader.FilterTermsEnum.class);
checkOverrideMethods(FilterAtomicReader.FilterDocsEnum.class);
checkOverrideMethods(FilterAtomicReader.FilterDocsAndPositionsEnum.class);
}
}