Examples with Term org.apache.lucene.index.Term used on opensource projects

Example 26 with Term

use of org.apache.lucene.index.Term in project elasticsearch by elastic.

the class BlendedTermQueryTests method testDismaxQuery.

public void testDismaxQuery() throws IOException {
    Directory dir = newDirectory();
    IndexWriter w = new IndexWriter(dir, newIndexWriterConfig(new MockAnalyzer(random())));
    String[] username = new String[] { "foo fighters", "some cool fan", "cover band" };
    String[] song = new String[] { "generator", "foo fighers - generator", "foo fighters generator" };
    final boolean omitNorms = random().nextBoolean();
    FieldType ft = new FieldType(TextField.TYPE_NOT_STORED);
    ft.setIndexOptions(random().nextBoolean() ? IndexOptions.DOCS : IndexOptions.DOCS_AND_FREQS);
    ft.setOmitNorms(omitNorms);
    ft.freeze();
    FieldType ft1 = new FieldType(TextField.TYPE_NOT_STORED);
    ft1.setIndexOptions(random().nextBoolean() ? IndexOptions.DOCS : IndexOptions.DOCS_AND_FREQS);
    ft1.setOmitNorms(omitNorms);
    ft1.freeze();
    for (int i = 0; i < username.length; i++) {
        Document d = new Document();
        d.add(new TextField("id", Integer.toString(i), Field.Store.YES));
        d.add(new Field("username", username[i], ft));
        d.add(new Field("song", song[i], ft));
        w.addDocument(d);
    }
    int iters = scaledRandomIntBetween(25, 100);
    for (int j = 0; j < iters; j++) {
        Document d = new Document();
        d.add(new TextField("id", Integer.toString(username.length + j), Field.Store.YES));
        d.add(new Field("username", "foo fighters", ft1));
        d.add(new Field("song", "some bogus text to bump up IDF", ft1));
        w.addDocument(d);
    }
    w.commit();
    DirectoryReader reader = DirectoryReader.open(w);
    IndexSearcher searcher = setSimilarity(newSearcher(reader));
    {
        String[] fields = new String[] { "username", "song" };
        BooleanQuery.Builder query = new BooleanQuery.Builder();
        query.setDisableCoord(true);
        query.add(BlendedTermQuery.dismaxBlendedQuery(toTerms(fields, "foo"), 0.1f), BooleanClause.Occur.SHOULD);
        query.add(BlendedTermQuery.dismaxBlendedQuery(toTerms(fields, "fighters"), 0.1f), BooleanClause.Occur.SHOULD);
        query.add(BlendedTermQuery.dismaxBlendedQuery(toTerms(fields, "generator"), 0.1f), BooleanClause.Occur.SHOULD);
        TopDocs search = searcher.search(query.build(), 10);
        ScoreDoc[] scoreDocs = search.scoreDocs;
        assertEquals(Integer.toString(0), reader.document(scoreDocs[0].doc).getField("id").stringValue());
    }
    {
        BooleanQuery.Builder query = new BooleanQuery.Builder();
        query.setDisableCoord(true);
        DisjunctionMaxQuery uname = new DisjunctionMaxQuery(Arrays.asList(new TermQuery(new Term("username", "foo")), new TermQuery(new Term("song", "foo"))), 0.0f);
        DisjunctionMaxQuery s = new DisjunctionMaxQuery(Arrays.asList(new TermQuery(new Term("username", "fighers")), new TermQuery(new Term("song", "fighers"))), 0.0f);
        DisjunctionMaxQuery gen = new DisjunctionMaxQuery(Arrays.asList(new TermQuery(new Term("username", "generator")), new TermQuery(new Term("song", "generator"))), 0f);
        query.add(uname, BooleanClause.Occur.SHOULD);
        query.add(s, BooleanClause.Occur.SHOULD);
        query.add(gen, BooleanClause.Occur.SHOULD);
        TopDocs search = searcher.search(query.build(), 4);
        ScoreDoc[] scoreDocs = search.scoreDocs;
        assertEquals(Integer.toString(1), reader.document(scoreDocs[0].doc).getField("id").stringValue());
    }
    reader.close();
    w.close();
    dir.close();
}

Example 27 with Term

use of org.apache.lucene.index.Term in project elasticsearch by elastic.

the class CustomPostingsHighlighterTests method testCustomPostingsHighlighter.

public void testCustomPostingsHighlighter() throws Exception {
    Directory dir = newDirectory();
    IndexWriterConfig iwc = newIndexWriterConfig(new MockAnalyzer(random()));
    iwc.setMergePolicy(newLogMergePolicy());
    RandomIndexWriter iw = new RandomIndexWriter(random(), dir, iwc);
    FieldType offsetsType = new FieldType(TextField.TYPE_STORED);
    offsetsType.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS);
    //good position but only one match
    final String firstValue = "This is a test. Just a test1 highlighting from postings highlighter.";
    Field body = new Field("body", "", offsetsType);
    Document doc = new Document();
    doc.add(body);
    body.setStringValue(firstValue);
    //two matches, not the best snippet due to its length though
    final String secondValue = "This is the second highlighting value to perform highlighting on a longer text that gets scored lower.";
    Field body2 = new Field("body", "", offsetsType);
    doc.add(body2);
    body2.setStringValue(secondValue);
    //two matches and short, will be scored highest
    final String thirdValue = "This is highlighting the third short highlighting value.";
    Field body3 = new Field("body", "", offsetsType);
    doc.add(body3);
    body3.setStringValue(thirdValue);
    //one match, same as first but at the end, will be scored lower due to its position
    final String fourthValue = "Just a test4 highlighting from postings highlighter.";
    Field body4 = new Field("body", "", offsetsType);
    doc.add(body4);
    body4.setStringValue(fourthValue);
    iw.addDocument(doc);
    IndexReader ir = iw.getReader();
    iw.close();
    String firstHlValue = "Just a test1 <b>highlighting</b> from postings highlighter.";
    String secondHlValue = "This is the second <b>highlighting</b> value to perform <b>highlighting</b> on a longer text that gets scored lower.";
    String thirdHlValue = "This is <b>highlighting</b> the third short <b>highlighting</b> value.";
    String fourthHlValue = "Just a test4 <b>highlighting</b> from postings highlighter.";
    IndexSearcher searcher = newSearcher(ir);
    Query query = new TermQuery(new Term("body", "highlighting"));
    TopDocs topDocs = searcher.search(query, 10, Sort.INDEXORDER);
    assertThat(topDocs.totalHits, equalTo(1));
    int docId = topDocs.scoreDocs[0].doc;
    String fieldValue = firstValue + HighlightUtils.PARAGRAPH_SEPARATOR + secondValue + HighlightUtils.PARAGRAPH_SEPARATOR + thirdValue + HighlightUtils.PARAGRAPH_SEPARATOR + fourthValue;
    CustomPostingsHighlighter highlighter = new CustomPostingsHighlighter(null, new CustomPassageFormatter("<b>", "</b>", new DefaultEncoder()), fieldValue, false);
    Snippet[] snippets = highlighter.highlightField("body", query, searcher, docId, 5);
    assertThat(snippets.length, equalTo(4));
    assertThat(snippets[0].getText(), equalTo(firstHlValue));
    assertThat(snippets[1].getText(), equalTo(secondHlValue));
    assertThat(snippets[2].getText(), equalTo(thirdHlValue));
    assertThat(snippets[3].getText(), equalTo(fourthHlValue));
    ir.close();
    dir.close();
}

Example 28 with Term

use of org.apache.lucene.index.Term in project elasticsearch by elastic.

the class CandidateQueryTests method testDuel.

public void testDuel() throws Exception {
    List<Function<String, Query>> queryFunctions = new ArrayList<>();
    queryFunctions.add((id) -> new PrefixQuery(new Term("field", id)));
    queryFunctions.add((id) -> new WildcardQuery(new Term("field", id + "*")));
    queryFunctions.add((id) -> new CustomQuery(new Term("field", id)));
    queryFunctions.add((id) -> new SpanTermQuery(new Term("field", id)));
    queryFunctions.add((id) -> new TermQuery(new Term("field", id)));
    queryFunctions.add((id) -> {
        BooleanQuery.Builder builder = new BooleanQuery.Builder();
        return builder.build();
    });
    queryFunctions.add((id) -> {
        BooleanQuery.Builder builder = new BooleanQuery.Builder();
        builder.add(new TermQuery(new Term("field", id)), BooleanClause.Occur.MUST);
        if (randomBoolean()) {
            builder.add(new MatchNoDocsQuery("no reason"), BooleanClause.Occur.MUST_NOT);
        }
        if (randomBoolean()) {
            builder.add(new CustomQuery(new Term("field", id)), BooleanClause.Occur.MUST);
        }
        return builder.build();
    });
    queryFunctions.add((id) -> {
        BooleanQuery.Builder builder = new BooleanQuery.Builder();
        builder.add(new TermQuery(new Term("field", id)), BooleanClause.Occur.SHOULD);
        if (randomBoolean()) {
            builder.add(new MatchNoDocsQuery("no reason"), BooleanClause.Occur.MUST_NOT);
        }
        if (randomBoolean()) {
            builder.add(new CustomQuery(new Term("field", id)), BooleanClause.Occur.SHOULD);
        }
        return builder.build();
    });
    queryFunctions.add((id) -> {
        BooleanQuery.Builder builder = new BooleanQuery.Builder();
        builder.add(new MatchAllDocsQuery(), BooleanClause.Occur.MUST);
        builder.add(new MatchAllDocsQuery(), BooleanClause.Occur.MUST);
        if (randomBoolean()) {
            builder.add(new MatchNoDocsQuery("no reason"), BooleanClause.Occur.MUST_NOT);
        }
        return builder.build();
    });
    queryFunctions.add((id) -> {
        BooleanQuery.Builder builder = new BooleanQuery.Builder();
        builder.add(new MatchAllDocsQuery(), BooleanClause.Occur.SHOULD);
        builder.add(new MatchAllDocsQuery(), BooleanClause.Occur.SHOULD);
        if (randomBoolean()) {
            builder.add(new MatchNoDocsQuery("no reason"), BooleanClause.Occur.MUST_NOT);
        }
        return builder.build();
    });
    queryFunctions.add((id) -> {
        BooleanQuery.Builder builder = new BooleanQuery.Builder();
        builder.setMinimumNumberShouldMatch(randomIntBetween(0, 4));
        builder.add(new TermQuery(new Term("field", id)), BooleanClause.Occur.SHOULD);
        builder.add(new CustomQuery(new Term("field", id)), BooleanClause.Occur.SHOULD);
        return builder.build();
    });
    queryFunctions.add((id) -> new MatchAllDocsQuery());
    queryFunctions.add((id) -> new MatchNoDocsQuery("no reason at all"));
    int numDocs = randomIntBetween(queryFunctions.size(), queryFunctions.size() * 3);
    List<ParseContext.Document> documents = new ArrayList<>();
    for (int i = 0; i < numDocs; i++) {
        String id = Integer.toString(i);
        Query query = queryFunctions.get(i % queryFunctions.size()).apply(id);
        addQuery(query, documents);
    }
    indexWriter.addDocuments(documents);
    indexWriter.close();
    directoryReader = DirectoryReader.open(directory);
    IndexSearcher shardSearcher = newSearcher(directoryReader);
    // Disable query cache, because ControlQuery cannot be cached...
    shardSearcher.setQueryCache(null);
    for (int i = 0; i < numDocs; i++) {
        String id = Integer.toString(i);
        Iterable<? extends IndexableField> doc = Collections.singleton(new StringField("field", id, Field.Store.NO));
        MemoryIndex memoryIndex = MemoryIndex.fromDocument(doc, new WhitespaceAnalyzer());
        duelRun(queryStore, memoryIndex, shardSearcher);
    }
    Iterable<? extends IndexableField> doc = Collections.singleton(new StringField("field", "value", Field.Store.NO));
    MemoryIndex memoryIndex = MemoryIndex.fromDocument(doc, new WhitespaceAnalyzer());
    duelRun(queryStore, memoryIndex, shardSearcher);
    // Empty percolator doc:
    memoryIndex = new MemoryIndex();
    duelRun(queryStore, memoryIndex, shardSearcher);
}

Example 29 with Term

use of org.apache.lucene.index.Term in project elasticsearch by elastic.

the class PercolateQueryTests method testPercolateQuery.

public void testPercolateQuery() throws Exception {
    List<Iterable<? extends IndexableField>> docs = new ArrayList<>();
    List<Query> queries = new ArrayList<>();
    PercolateQuery.QueryStore queryStore = ctx -> queries::get;
    queries.add(new TermQuery(new Term("field", "fox")));
    docs.add(Collections.singleton(new StringField("select", "a", Field.Store.NO)));
    SpanNearQuery.Builder snp = new SpanNearQuery.Builder("field", true);
    snp.addClause(new SpanTermQuery(new Term("field", "jumps")));
    snp.addClause(new SpanTermQuery(new Term("field", "lazy")));
    snp.addClause(new SpanTermQuery(new Term("field", "dog")));
    snp.setSlop(2);
    queries.add(snp.build());
    docs.add(Collections.singleton(new StringField("select", "b", Field.Store.NO)));
    PhraseQuery.Builder pq1 = new PhraseQuery.Builder();
    pq1.add(new Term("field", "quick"));
    pq1.add(new Term("field", "brown"));
    pq1.add(new Term("field", "jumps"));
    pq1.setSlop(1);
    queries.add(pq1.build());
    docs.add(Collections.singleton(new StringField("select", "b", Field.Store.NO)));
    BooleanQuery.Builder bq1 = new BooleanQuery.Builder();
    bq1.add(new TermQuery(new Term("field", "quick")), BooleanClause.Occur.MUST);
    bq1.add(new TermQuery(new Term("field", "brown")), BooleanClause.Occur.MUST);
    bq1.add(new TermQuery(new Term("field", "fox")), BooleanClause.Occur.MUST);
    queries.add(bq1.build());
    docs.add(Collections.singleton(new StringField("select", "b", Field.Store.NO)));
    indexWriter.addDocuments(docs);
    indexWriter.close();
    directoryReader = DirectoryReader.open(directory);
    IndexSearcher shardSearcher = newSearcher(directoryReader);
    MemoryIndex memoryIndex = new MemoryIndex();
    memoryIndex.addField("field", "the quick brown fox jumps over the lazy dog", new WhitespaceAnalyzer());
    IndexSearcher percolateSearcher = memoryIndex.createSearcher();
    // no scoring, wrapping it in a constant score query:
    Query query = new ConstantScoreQuery(new PercolateQuery("type", queryStore, new BytesArray("a"), new TermQuery(new Term("select", "a")), percolateSearcher, new MatchNoDocsQuery("")));
    TopDocs topDocs = shardSearcher.search(query, 10);
    assertThat(topDocs.totalHits, equalTo(1));
    assertThat(topDocs.scoreDocs.length, equalTo(1));
    assertThat(topDocs.scoreDocs[0].doc, equalTo(0));
    Explanation explanation = shardSearcher.explain(query, 0);
    assertThat(explanation.isMatch(), is(true));
    assertThat(explanation.getValue(), equalTo(topDocs.scoreDocs[0].score));
    query = new ConstantScoreQuery(new PercolateQuery("type", queryStore, new BytesArray("b"), new TermQuery(new Term("select", "b")), percolateSearcher, new MatchNoDocsQuery("")));
    topDocs = shardSearcher.search(query, 10);
    assertThat(topDocs.totalHits, equalTo(3));
    assertThat(topDocs.scoreDocs.length, equalTo(3));
    assertThat(topDocs.scoreDocs[0].doc, equalTo(1));
    explanation = shardSearcher.explain(query, 1);
    assertThat(explanation.isMatch(), is(true));
    assertThat(explanation.getValue(), equalTo(topDocs.scoreDocs[0].score));
    assertThat(topDocs.scoreDocs[1].doc, equalTo(2));
    explanation = shardSearcher.explain(query, 2);
    assertThat(explanation.isMatch(), is(true));
    assertThat(explanation.getValue(), equalTo(topDocs.scoreDocs[1].score));
    assertThat(topDocs.scoreDocs[2].doc, equalTo(3));
    explanation = shardSearcher.explain(query, 2);
    assertThat(explanation.isMatch(), is(true));
    assertThat(explanation.getValue(), equalTo(topDocs.scoreDocs[2].score));
    query = new ConstantScoreQuery(new PercolateQuery("type", queryStore, new BytesArray("c"), new MatchAllDocsQuery(), percolateSearcher, new MatchAllDocsQuery()));
    topDocs = shardSearcher.search(query, 10);
    assertThat(topDocs.totalHits, equalTo(4));
    query = new PercolateQuery("type", queryStore, new BytesArray("{}"), new TermQuery(new Term("select", "b")), percolateSearcher, new MatchNoDocsQuery(""));
    topDocs = shardSearcher.search(query, 10);
    assertThat(topDocs.totalHits, equalTo(3));
    assertThat(topDocs.scoreDocs.length, equalTo(3));
    assertThat(topDocs.scoreDocs[0].doc, equalTo(3));
    explanation = shardSearcher.explain(query, 3);
    assertThat(explanation.isMatch(), is(true));
    assertThat(explanation.getValue(), equalTo(topDocs.scoreDocs[0].score));
    assertThat(explanation.getDetails(), arrayWithSize(1));
    assertThat(topDocs.scoreDocs[1].doc, equalTo(2));
    explanation = shardSearcher.explain(query, 2);
    assertThat(explanation.isMatch(), is(true));
    assertThat(explanation.getValue(), equalTo(topDocs.scoreDocs[1].score));
    assertThat(explanation.getDetails(), arrayWithSize(1));
    assertThat(topDocs.scoreDocs[2].doc, equalTo(1));
    explanation = shardSearcher.explain(query, 1);
    assertThat(explanation.isMatch(), is(true));
    assertThat(explanation.getValue(), equalTo(topDocs.scoreDocs[2].score));
    assertThat(explanation.getDetails(), arrayWithSize(1));
}

Example 30 with Term

use of org.apache.lucene.index.Term in project elasticsearch by elastic.

the class PercolatorFieldMapperTests method testCreateCandidateQuery.

public void testCreateCandidateQuery() throws Exception {
    addQueryMapping();
    MemoryIndex memoryIndex = new MemoryIndex(false);
    memoryIndex.addField("field1", "the quick brown fox jumps over the lazy dog", new WhitespaceAnalyzer());
    memoryIndex.addField("field2", "some more text", new WhitespaceAnalyzer());
    memoryIndex.addField("_field3", "unhide me", new WhitespaceAnalyzer());
    memoryIndex.addField("field4", "123", new WhitespaceAnalyzer());
    memoryIndex.addField(new LongPoint("number_field", 10L), new WhitespaceAnalyzer());
    IndexReader indexReader = memoryIndex.createSearcher().getIndexReader();
    BooleanQuery candidateQuery = (BooleanQuery) fieldType.createCandidateQuery(indexReader);
    assertEquals(2, candidateQuery.clauses().size());
    assertEquals(Occur.SHOULD, candidateQuery.clauses().get(0).getOccur());
    TermInSetQuery termsQuery = (TermInSetQuery) candidateQuery.clauses().get(0).getQuery();
    PrefixCodedTerms terms = termsQuery.getTermData();
    assertThat(terms.size(), equalTo(14L));
    PrefixCodedTerms.TermIterator termIterator = terms.iterator();
    assertTermIterator(termIterator, "_field3me", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "_field3unhide", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field1brown", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field1dog", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field1fox", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field1jumps", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field1lazy", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field1over", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field1quick", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field1the", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field2more", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field2some", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field2text", fieldType.queryTermsField.name());
    assertTermIterator(termIterator, "field4123", fieldType.queryTermsField.name());
    assertEquals(Occur.SHOULD, candidateQuery.clauses().get(1).getOccur());
    assertEquals(new TermQuery(new Term(fieldType.extractionResultField.name(), EXTRACTION_FAILED)), candidateQuery.clauses().get(1).getQuery());
}