Search in sources :

Example 1 with ScoreCollectorExpression

use of io.crate.expression.reference.doc.lucene.ScoreCollectorExpression in project crate by crate.

the class LuceneOrderedDocCollectorTest method testSearchWithScores.

@Test
public void testSearchWithScores() throws Exception {
    IndexWriter w = new IndexWriter(new ByteBuffersDirectory(), new IndexWriterConfig(new KeywordAnalyzer()));
    FieldType fieldType = KeywordFieldMapper.Defaults.FIELD_TYPE;
    for (int i = 0; i < 3; i++) {
        addDoc(w, "x", fieldType, "Arthur");
    }
    // not "Arthur" to lower score
    addDoc(w, "x", fieldType, "Arthur");
    w.commit();
    IndexSearcher searcher = new IndexSearcher(DirectoryReader.open(w, true, true));
    List<LuceneCollectorExpression<?>> columnReferences = Collections.singletonList(new ScoreCollectorExpression());
    Query query = new ConstantScoreQuery(new TermQuery(new Term("x", new BytesRef("Arthur"))));
    LuceneOrderedDocCollector collector = collector(searcher, columnReferences, query, null, true);
    KeyIterable<ShardId, Row> result = collector.collect();
    assertThat(StreamSupport.stream(result.spliterator(), false).count(), is(2L));
    Iterator<Row> values = result.iterator();
    assertThat(values.next().get(0), Matchers.is(1.0F));
    assertThat(values.next().get(0), Matchers.is(1.0F));
}
Also used : IndexSearcher(org.apache.lucene.search.IndexSearcher) KeywordAnalyzer(org.apache.lucene.analysis.core.KeywordAnalyzer) TermQuery(org.apache.lucene.search.TermQuery) Query(org.apache.lucene.search.Query) FuzzyQuery(org.apache.lucene.search.FuzzyQuery) MatchAllDocsQuery(org.apache.lucene.search.MatchAllDocsQuery) ConstantScoreQuery(org.apache.lucene.search.ConstantScoreQuery) TermQuery(org.apache.lucene.search.TermQuery) Term(org.apache.lucene.index.Term) FieldType(org.apache.lucene.document.FieldType) ShardId(org.elasticsearch.index.shard.ShardId) IndexWriter(org.apache.lucene.index.IndexWriter) ByteBuffersDirectory(org.apache.lucene.store.ByteBuffersDirectory) ConstantScoreQuery(org.apache.lucene.search.ConstantScoreQuery) Row(io.crate.data.Row) ScoreCollectorExpression(io.crate.expression.reference.doc.lucene.ScoreCollectorExpression) LuceneCollectorExpression(io.crate.expression.reference.doc.lucene.LuceneCollectorExpression) BytesRef(org.apache.lucene.util.BytesRef) IndexWriterConfig(org.apache.lucene.index.IndexWriterConfig) RandomizedTest(com.carrotsearch.randomizedtesting.RandomizedTest) Test(org.junit.Test)

Example 2 with ScoreCollectorExpression

use of io.crate.expression.reference.doc.lucene.ScoreCollectorExpression in project crate by crate.

the class LuceneOrderedDocCollectorTest method testSearchMoreAppliesMinScoreFilter.

@Test
public void testSearchMoreAppliesMinScoreFilter() throws Exception {
    IndexWriter w = new IndexWriter(new ByteBuffersDirectory(), new IndexWriterConfig(new KeywordAnalyzer()));
    var keywordFieldType = new KeywordFieldMapper.KeywordFieldType("x");
    var fieldType = KeywordFieldMapper.Defaults.FIELD_TYPE;
    for (int i = 0; i < 3; i++) {
        addDoc(w, "x", fieldType, "Arthur");
    }
    // not "Arthur" to lower score
    addDoc(w, "x", fieldType, "Arthurr");
    w.commit();
    IndexSearcher searcher = new IndexSearcher(DirectoryReader.open(w, true, true));
    List<LuceneCollectorExpression<?>> columnReferences = Collections.singletonList(new ScoreCollectorExpression());
    Query query = new FuzzyQuery(new Term("x", "Arthur"), Fuzziness.AUTO.asDistance("Arthur"), 2, 3, true);
    LuceneOrderedDocCollector collector;
    // without minScore filter we get 2 and 2 docs - this is not necessary for the test but is here
    // to make sure the "FuzzyQuery" matches the right documents
    collector = collector(searcher, columnReferences, query, null, true);
    assertThat(StreamSupport.stream(collector.collect().spliterator(), false).count(), is(2L));
    assertThat(StreamSupport.stream(collector.collect().spliterator(), false).count(), is(2L));
    collector = collector(searcher, columnReferences, query, 0.15f, true);
    int count = 0;
    // initialSearch -> 2 rows
    for (Row row : collector.collect()) {
        assertThat((float) row.get(0), Matchers.greaterThanOrEqualTo(0.15f));
        count++;
    }
    assertThat(count, is(2));
    count = 0;
    // searchMore -> 1 row is below minScore
    for (Row row : collector.collect()) {
        assertThat((float) row.get(0), Matchers.greaterThanOrEqualTo(0.15f));
        count++;
    }
    assertThat(count, is(1));
}
Also used : IndexSearcher(org.apache.lucene.search.IndexSearcher) KeywordAnalyzer(org.apache.lucene.analysis.core.KeywordAnalyzer) Query(org.apache.lucene.search.Query) FuzzyQuery(org.apache.lucene.search.FuzzyQuery) MatchAllDocsQuery(org.apache.lucene.search.MatchAllDocsQuery) ConstantScoreQuery(org.apache.lucene.search.ConstantScoreQuery) TermQuery(org.apache.lucene.search.TermQuery) Term(org.apache.lucene.index.Term) IndexWriter(org.apache.lucene.index.IndexWriter) ByteBuffersDirectory(org.apache.lucene.store.ByteBuffersDirectory) FuzzyQuery(org.apache.lucene.search.FuzzyQuery) Row(io.crate.data.Row) ScoreCollectorExpression(io.crate.expression.reference.doc.lucene.ScoreCollectorExpression) LuceneCollectorExpression(io.crate.expression.reference.doc.lucene.LuceneCollectorExpression) IndexWriterConfig(org.apache.lucene.index.IndexWriterConfig) RandomizedTest(com.carrotsearch.randomizedtesting.RandomizedTest) Test(org.junit.Test)

Example 3 with ScoreCollectorExpression

use of io.crate.expression.reference.doc.lucene.ScoreCollectorExpression in project crate by crate.

the class LuceneOrderedDocCollectorTest method testSearchNoScores.

@Test
public void testSearchNoScores() throws Exception {
    IndexWriter w = new IndexWriter(new ByteBuffersDirectory(), new IndexWriterConfig(new KeywordAnalyzer()));
    String name = "x";
    var keywordFieldType = new KeywordFieldMapper.KeywordFieldType(name);
    var fieldType = KeywordFieldMapper.Defaults.FIELD_TYPE;
    for (int i = 0; i < 3; i++) {
        addDoc(w, name, fieldType, "Arthur");
    }
    // not "Arthur" to lower score
    addDoc(w, name, fieldType, "Arthur");
    w.commit();
    IndexSearcher searcher = new IndexSearcher(DirectoryReader.open(w, true, true));
    List<LuceneCollectorExpression<?>> columnReferences = Collections.singletonList(new ScoreCollectorExpression());
    Query query = new TermQuery(new Term(name, new BytesRef("Arthur")));
    LuceneOrderedDocCollector collector = collector(searcher, columnReferences, query, null, false);
    KeyIterable<ShardId, Row> result = collector.collect();
    assertThat(StreamSupport.stream(result.spliterator(), false).count(), is(2L));
    Iterator<Row> values = result.iterator();
    assertThat(values.next().get(0), Matchers.is(Float.NaN));
    assertThat(values.next().get(0), Matchers.is(Float.NaN));
}
Also used : IndexSearcher(org.apache.lucene.search.IndexSearcher) KeywordAnalyzer(org.apache.lucene.analysis.core.KeywordAnalyzer) TermQuery(org.apache.lucene.search.TermQuery) Query(org.apache.lucene.search.Query) FuzzyQuery(org.apache.lucene.search.FuzzyQuery) MatchAllDocsQuery(org.apache.lucene.search.MatchAllDocsQuery) ConstantScoreQuery(org.apache.lucene.search.ConstantScoreQuery) TermQuery(org.apache.lucene.search.TermQuery) Term(org.apache.lucene.index.Term) ShardId(org.elasticsearch.index.shard.ShardId) IndexWriter(org.apache.lucene.index.IndexWriter) ByteBuffersDirectory(org.apache.lucene.store.ByteBuffersDirectory) Row(io.crate.data.Row) ScoreCollectorExpression(io.crate.expression.reference.doc.lucene.ScoreCollectorExpression) LuceneCollectorExpression(io.crate.expression.reference.doc.lucene.LuceneCollectorExpression) BytesRef(org.apache.lucene.util.BytesRef) IndexWriterConfig(org.apache.lucene.index.IndexWriterConfig) RandomizedTest(com.carrotsearch.randomizedtesting.RandomizedTest) Test(org.junit.Test)

Aggregations

RandomizedTest (com.carrotsearch.randomizedtesting.RandomizedTest)3 Row (io.crate.data.Row)3 LuceneCollectorExpression (io.crate.expression.reference.doc.lucene.LuceneCollectorExpression)3 ScoreCollectorExpression (io.crate.expression.reference.doc.lucene.ScoreCollectorExpression)3 KeywordAnalyzer (org.apache.lucene.analysis.core.KeywordAnalyzer)3 IndexWriter (org.apache.lucene.index.IndexWriter)3 IndexWriterConfig (org.apache.lucene.index.IndexWriterConfig)3 Term (org.apache.lucene.index.Term)3 ConstantScoreQuery (org.apache.lucene.search.ConstantScoreQuery)3 FuzzyQuery (org.apache.lucene.search.FuzzyQuery)3 IndexSearcher (org.apache.lucene.search.IndexSearcher)3 MatchAllDocsQuery (org.apache.lucene.search.MatchAllDocsQuery)3 Query (org.apache.lucene.search.Query)3 TermQuery (org.apache.lucene.search.TermQuery)3 ByteBuffersDirectory (org.apache.lucene.store.ByteBuffersDirectory)3 Test (org.junit.Test)3 BytesRef (org.apache.lucene.util.BytesRef)2 ShardId (org.elasticsearch.index.shard.ShardId)2 FieldType (org.apache.lucene.document.FieldType)1