gistfile1.java

package com.github.tteofili.samples.lucene.benchmarks;

/*
 * 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.util.Random;

import org.apache.lucene.analysis.Analyzer;
import org.apache.lucene.analysis.MockAnalyzer;
import org.apache.lucene.document.BinaryDocValuesField;
import org.apache.lucene.document.Document;
import org.apache.lucene.document.Field;
import org.apache.lucene.document.FieldType;
import org.apache.lucene.document.SortedSetDocValuesField;
import org.apache.lucene.document.TextField;
import org.apache.lucene.index.RandomIndexWriter;
import org.apache.lucene.index.Term;
import org.apache.lucene.search.ConstantScoreQuery;
import org.apache.lucene.search.FieldValueFilter;
import org.apache.lucene.search.Filter;
import org.apache.lucene.search.IndexSearcher;
import org.apache.lucene.search.MatchAllDocsQuery;
import org.apache.lucene.search.Query;
import org.apache.lucene.search.TermQuery;
import org.apache.lucene.search.TermRangeQuery;
import org.apache.lucene.search.TopDocs;
import org.apache.lucene.search.WildcardQuery;
import org.apache.lucene.store.Directory;
import org.apache.lucene.util.BytesRef;
import org.apache.lucene.util.LuceneTestCase;
import org.junit.After;
import org.junit.Before;
import org.junit.Test;

/**
 * benchmark test for "field exists" query
 */
public class FieldExistBenchmarkTest extends LuceneTestCase {

  private String changing = "changing";
  private String constant = "constant";
  private String even = "even";
  private String fields = "fields";

  private String changing_dv = "changing_dv";
  private String constant_dv = "constant_dv";
  private String even_dv = "even_dv";

  private RandomIndexWriter indexWriter;
  private Directory dir;
  private FieldType textFt;

  @Before
  public void setUp() throws Exception {
    super.setUp();
    dir = newDirectory();
    indexWriter = new RandomIndexWriter(new Random(), dir);
    textFt = new FieldType(TextField.TYPE_NOT_STORED);
  }

  @Override
  @After
  public void tearDown() throws Exception {
    super.tearDown();
    indexWriter.close();
    dir.close();
  }

  @Test
  public void testFieldExists() throws Exception {
    populateIndex(new MockAnalyzer(new Random()));

    IndexSearcher indexSearcher = new IndexSearcher(indexWriter.getReader());
    // warm up caches
    indexSearcher.search(new MatchAllDocsQuery(), 100000);

    // queries on changing field
    benchmarkQuery(new TermRangeQuery(changing, null, null, true, true), null, indexSearcher);
    benchmarkQuery(new WildcardQuery(new Term(changing, "*")), null, indexSearcher);
    benchmarkQuery(new TermQuery(new Term(fields, changing)), null, indexSearcher);
    benchmarkQuery(new ConstantScoreQuery(new FieldValueFilter(changing_dv)), null, indexSearcher);
    benchmarkQuery(new MatchAllDocsQuery(), new FieldValueFilter(changing_dv), indexSearcher);

    // queries on constant field
    benchmarkQuery(new TermRangeQuery(constant, null, null, true, true), null, indexSearcher);
    benchmarkQuery(new WildcardQuery(new Term(constant, "*")), null, indexSearcher);
    benchmarkQuery(new TermQuery(new Term(fields, constant)), null, indexSearcher);
    benchmarkQuery(new ConstantScoreQuery(new FieldValueFilter(constant_dv)), null, indexSearcher);
    benchmarkQuery(new MatchAllDocsQuery(), new FieldValueFilter(constant_dv), indexSearcher);

    // queries on even field
    benchmarkQuery(new TermRangeQuery(even, null, null, true, true), null, indexSearcher);
    benchmarkQuery(new WildcardQuery(new Term(even, "*")), null, indexSearcher);
    benchmarkQuery(new TermQuery(new Term(fields, even)), null, indexSearcher);
    benchmarkQuery(new ConstantScoreQuery(new FieldValueFilter(even_dv)), null, indexSearcher);
    benchmarkQuery(new MatchAllDocsQuery(), new FieldValueFilter(even_dv), indexSearcher);
  }

  private void benchmarkQuery(Query query, Filter filter, IndexSearcher indexSearcher) throws Exception {
    System.out.println(query + (filter != null ? " " + filter : ""));
    long start = System.currentTimeMillis();
    TopDocs search;
    if (filter == null) {
      search = indexSearcher.search(query, Integer.MAX_VALUE);
    } else {
      search = indexSearcher.search(query, filter, Integer.MAX_VALUE);
    }
    System.out.println(System.currentTimeMillis() - start + "ms");
    System.out.println(search.totalHits + " hits");
  }

  private String createVectorString(int x, int y, int z) {
    StringBuilder text = new StringBuilder();

    for (int i = 0; i < x; i++)
      text.append("x ");

    for (int i = 0; i < y; i++)
      text.append("y ");

    for (int i = 0; i < z; i++)
      text.append("z ");

    return text.toString();
  }

  private Document createDoc(int x, int y, int z, String token) {
    Document doc = new Document();

    String vectorString = createVectorString(x, y, z);
    String randomValue = vectorString != null ? vectorString : String.valueOf(random().nextInt(1000000));
    doc.add(new Field(changing, randomValue, textFt));
    doc.add(new BinaryDocValuesField(changing_dv, new BytesRef(randomValue)));
    doc.add(new Field(constant, token, textFt));
    doc.add(new BinaryDocValuesField(constant_dv, new BytesRef(token)));
    if (x % 2 == 0) {
      doc.add(new Field(even, token, textFt));
      doc.add(new BinaryDocValuesField(even_dv, new BytesRef(token)));
      doc.add(new Field(fields, even, textFt));
    }
    doc.add(new Field(fields, changing, textFt));
    doc.add(new Field(fields, constant, textFt));

    return doc;
  }

  private void addDoc(int x, int y, int z, String className, Analyzer analyzer) throws IOException {
    Document doc = createDoc(x, y, z, className);
    indexWriter.addDocument(doc, analyzer);
  }

  public void populateIndex(Analyzer analyzer) throws IOException {
    indexWriter.deleteAll();
    indexWriter.commit();

    String constantToken = Character.toString((char) random().nextInt(255));
    for (int i = 0; i < 100000; i++) {
      int seed = (i % 254) + 1;
      addDoc(random().nextInt(seed), random().nextInt(seed), random().nextInt(seed), constantToken, analyzer);
    }
    indexWriter.commit();
  }

}

is this because of the fact that you don't have a common base (e.g. wrt Analyzers, Codecs, etc.) or because there's some other LuceneTestCase specific limitation?
I thought that randomization would have been a good thing also for benchmarks because (on a significant bunch of runs) it allows to find out what's the most performant approach on average, regardless of the other variables.