Examples with CannedTokenStream org.apache.lucene.analysis.CannedTokenStream used on opensource projects

Example 26 with CannedTokenStream

use of org.apache.lucene.analysis.CannedTokenStream in project lucene-solr by apache.

the class FastVectorHighlighterTest method matchedFieldsTestCase.

private void matchedFieldsTestCase(boolean useMatchedFields, boolean fieldMatch, String fieldValue, String expected, Query... queryClauses) throws IOException {
    Document doc = new Document();
    FieldType stored = new FieldType(TextField.TYPE_STORED);
    stored.setStoreTermVectorOffsets(true);
    stored.setStoreTermVectorPositions(true);
    stored.setStoreTermVectors(true);
    stored.freeze();
    FieldType matched = new FieldType(TextField.TYPE_NOT_STORED);
    matched.setStoreTermVectorOffsets(true);
    matched.setStoreTermVectorPositions(true);
    matched.setStoreTermVectors(true);
    matched.freeze();
    // Whitespace tokenized with English stop words
    doc.add(new Field("field", fieldValue, stored));
    // Whitespace tokenized without stop words
    doc.add(new Field("field_exact", fieldValue, matched));
    // Whitespace tokenized without toLower
    doc.add(new Field("field_super_exact", fieldValue, matched));
    // Each letter is a token
    doc.add(new Field("field_characters", fieldValue, matched));
    // Every three letters is a token
    doc.add(new Field("field_tripples", fieldValue, matched));
    doc.add(new Field("field_sliced", // Sliced at 10 chars then analyzed just like field
    fieldValue.substring(// Sliced at 10 chars then analyzed just like field
    0, Math.min(fieldValue.length() - 1, 10)), matched));
    doc.add(new Field("field_der_red", new // Hacky field containing "der" and "red" at pos = 0
    CannedTokenStream(token("der", 1, 0, 3), token("red", 0, 0, 3)), matched));
    final Map<String, Analyzer> fieldAnalyzers = new TreeMap<>();
    fieldAnalyzers.put("field", new MockAnalyzer(random(), MockTokenizer.WHITESPACE, true, MockTokenFilter.ENGLISH_STOPSET));
    fieldAnalyzers.put("field_exact", new MockAnalyzer(random()));
    fieldAnalyzers.put("field_super_exact", new MockAnalyzer(random(), MockTokenizer.WHITESPACE, false));
    fieldAnalyzers.put("field_characters", new MockAnalyzer(random(), new CharacterRunAutomaton(new RegExp(".").toAutomaton()), true));
    fieldAnalyzers.put("field_tripples", new MockAnalyzer(random(), new CharacterRunAutomaton(new RegExp("...").toAutomaton()), true));
    fieldAnalyzers.put("field_sliced", fieldAnalyzers.get("field"));
    // This is required even though we provide a token stream
    fieldAnalyzers.put("field_der_red", fieldAnalyzers.get("field"));
    Analyzer analyzer = new DelegatingAnalyzerWrapper(Analyzer.PER_FIELD_REUSE_STRATEGY) {

        public Analyzer getWrappedAnalyzer(String fieldName) {
            return fieldAnalyzers.get(fieldName);
        }
    };
    Directory dir = newDirectory();
    IndexWriter writer = new IndexWriter(dir, newIndexWriterConfig(analyzer));
    writer.addDocument(doc);
    FastVectorHighlighter highlighter = new FastVectorHighlighter();
    FragListBuilder fragListBuilder = new SimpleFragListBuilder();
    FragmentsBuilder fragmentsBuilder = new ScoreOrderFragmentsBuilder();
    IndexReader reader = DirectoryReader.open(writer);
    String[] preTags = new String[] { "<b>" };
    String[] postTags = new String[] { "</b>" };
    Encoder encoder = new DefaultEncoder();
    int docId = 0;
    BooleanQuery.Builder query = new BooleanQuery.Builder();
    for (Query clause : queryClauses) {
        query.add(clause, Occur.MUST);
    }
    FieldQuery fieldQuery = new FieldQuery(query.build(), reader, true, fieldMatch);
    String[] bestFragments;
    if (useMatchedFields) {
        Set<String> matchedFields = new HashSet<>();
        matchedFields.add("field");
        matchedFields.add("field_exact");
        matchedFields.add("field_super_exact");
        matchedFields.add("field_characters");
        matchedFields.add("field_tripples");
        matchedFields.add("field_sliced");
        matchedFields.add("field_der_red");
        bestFragments = highlighter.getBestFragments(fieldQuery, reader, docId, "field", matchedFields, 25, 1, fragListBuilder, fragmentsBuilder, preTags, postTags, encoder);
    } else {
        bestFragments = highlighter.getBestFragments(fieldQuery, reader, docId, "field", 25, 1, fragListBuilder, fragmentsBuilder, preTags, postTags, encoder);
    }
    assertEquals(expected, bestFragments[0]);
    reader.close();
    writer.close();
    dir.close();
}

Example 27 with CannedTokenStream

use of org.apache.lucene.analysis.CannedTokenStream in project lucene-solr by apache.

the class FastVectorHighlighterTest method testBooleanPhraseWithSynonym.

public void testBooleanPhraseWithSynonym() throws IOException {
    Directory dir = newDirectory();
    IndexWriter writer = new IndexWriter(dir, newIndexWriterConfig(new MockAnalyzer(random())));
    Document doc = new Document();
    FieldType type = new FieldType(TextField.TYPE_NOT_STORED);
    type.setStoreTermVectorOffsets(true);
    type.setStoreTermVectorPositions(true);
    type.setStoreTermVectors(true);
    type.freeze();
    Token syn = new Token("httpwwwfacebookcom", 6, 29);
    syn.setPositionIncrement(0);
    CannedTokenStream ts = new CannedTokenStream(new Token("test", 0, 4), new Token("http", 6, 10), syn, new Token("www", 13, 16), new Token("facebook", 17, 25), new Token("com", 26, 29));
    Field field = new Field("field", ts, type);
    doc.add(field);
    doc.add(new StoredField("field", "Test: http://www.facebook.com"));
    writer.addDocument(doc);
    FastVectorHighlighter highlighter = new FastVectorHighlighter();
    IndexReader reader = DirectoryReader.open(writer);
    int docId = 0;
    // query1: match
    PhraseQuery pq = new PhraseQuery("field", "test", "http", "www", "facebook", "com");
    FieldQuery fieldQuery = highlighter.getFieldQuery(pq, reader);
    String[] bestFragments = highlighter.getBestFragments(fieldQuery, reader, docId, "field", 54, 1);
    assertEquals("<b>Test: http://www.facebook.com</b>", bestFragments[0]);
    // query2: match
    PhraseQuery pq2 = new PhraseQuery("field", "test", "httpwwwfacebookcom", "www", "facebook", "com");
    fieldQuery = highlighter.getFieldQuery(pq2, reader);
    bestFragments = highlighter.getBestFragments(fieldQuery, reader, docId, "field", 54, 1);
    assertEquals("<b>Test: http://www.facebook.com</b>", bestFragments[0]);
    // query3: OR query1 and query2 together
    BooleanQuery.Builder bq = new BooleanQuery.Builder();
    bq.add(pq, BooleanClause.Occur.SHOULD);
    bq.add(pq2, BooleanClause.Occur.SHOULD);
    fieldQuery = highlighter.getFieldQuery(bq.build(), reader);
    bestFragments = highlighter.getBestFragments(fieldQuery, reader, docId, "field", 54, 1);
    assertEquals("<b>Test: http://www.facebook.com</b>", bestFragments[0]);
    reader.close();
    writer.close();
    dir.close();
}

Example 28 with CannedTokenStream

use of org.apache.lucene.analysis.CannedTokenStream in project lucene-solr by apache.

the class BaseTermVectorsFormatTestCase method testPostingsEnumPayloads.

public void testPostingsEnumPayloads() throws Exception {
    Directory dir = newDirectory();
    IndexWriterConfig iwc = new IndexWriterConfig(null);
    IndexWriter iw = new IndexWriter(dir, iwc);
    Document doc = new Document();
    Token token1 = new Token("bar", 0, 3);
    token1.setPayload(new BytesRef("pay1"));
    Token token2 = new Token("bar", 4, 7);
    token2.setPayload(new BytesRef("pay2"));
    FieldType ft = new FieldType(TextField.TYPE_NOT_STORED);
    ft.setStoreTermVectors(true);
    ft.setStoreTermVectorPositions(true);
    ft.setStoreTermVectorPayloads(true);
    doc.add(new Field("foo", new CannedTokenStream(token1, token2), ft));
    iw.addDocument(doc);
    DirectoryReader reader = DirectoryReader.open(iw);
    Terms terms = getOnlyLeafReader(reader).getTermVector(0, "foo");
    TermsEnum termsEnum = terms.iterator();
    assertNotNull(termsEnum);
    assertEquals(new BytesRef("bar"), termsEnum.next());
    // sugar method (FREQS)
    PostingsEnum postings = termsEnum.postings(null);
    assertEquals(-1, postings.docID());
    assertEquals(0, postings.nextDoc());
    assertEquals(2, postings.freq());
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, postings.nextDoc());
    // termsenum reuse (FREQS)
    PostingsEnum postings2 = termsEnum.postings(postings);
    assertNotNull(postings2);
    // and it had better work
    assertEquals(-1, postings2.docID());
    assertEquals(0, postings2.nextDoc());
    assertEquals(2, postings2.freq());
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, postings2.nextDoc());
    // asking for docs only: ok
    PostingsEnum docsOnly = termsEnum.postings(null, PostingsEnum.NONE);
    assertEquals(-1, docsOnly.docID());
    assertEquals(0, docsOnly.nextDoc());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsOnly.freq() == 1 || docsOnly.freq() == 2);
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsOnly.nextDoc());
    // reuse that too
    PostingsEnum docsOnly2 = termsEnum.postings(docsOnly, PostingsEnum.NONE);
    assertNotNull(docsOnly2);
    // and it had better work
    assertEquals(-1, docsOnly2.docID());
    assertEquals(0, docsOnly2.nextDoc());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsOnly2.freq() == 1 || docsOnly2.freq() == 2);
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsOnly2.nextDoc());
    // asking for positions, ok
    PostingsEnum docsAndPositionsEnum = termsEnum.postings(null, PostingsEnum.POSITIONS);
    assertEquals(-1, docsAndPositionsEnum.docID());
    assertEquals(0, docsAndPositionsEnum.nextDoc());
    assertEquals(2, docsAndPositionsEnum.freq());
    assertEquals(0, docsAndPositionsEnum.nextPosition());
    assertEquals(-1, docsAndPositionsEnum.startOffset());
    assertEquals(-1, docsAndPositionsEnum.endOffset());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsAndPositionsEnum.getPayload() == null || new BytesRef("pay1").equals(docsAndPositionsEnum.getPayload()));
    assertEquals(1, docsAndPositionsEnum.nextPosition());
    assertEquals(-1, docsAndPositionsEnum.startOffset());
    assertEquals(-1, docsAndPositionsEnum.endOffset());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsAndPositionsEnum.getPayload() == null || new BytesRef("pay2").equals(docsAndPositionsEnum.getPayload()));
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsAndPositionsEnum.nextDoc());
    // now reuse the positions
    PostingsEnum docsAndPositionsEnum2 = termsEnum.postings(docsAndPositionsEnum, PostingsEnum.POSITIONS);
    assertEquals(-1, docsAndPositionsEnum2.docID());
    assertEquals(0, docsAndPositionsEnum2.nextDoc());
    assertEquals(2, docsAndPositionsEnum2.freq());
    assertEquals(0, docsAndPositionsEnum2.nextPosition());
    assertEquals(-1, docsAndPositionsEnum2.startOffset());
    assertEquals(-1, docsAndPositionsEnum2.endOffset());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsAndPositionsEnum2.getPayload() == null || new BytesRef("pay1").equals(docsAndPositionsEnum2.getPayload()));
    assertEquals(1, docsAndPositionsEnum2.nextPosition());
    assertEquals(-1, docsAndPositionsEnum2.startOffset());
    assertEquals(-1, docsAndPositionsEnum2.endOffset());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsAndPositionsEnum2.getPayload() == null || new BytesRef("pay2").equals(docsAndPositionsEnum2.getPayload()));
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsAndPositionsEnum2.nextDoc());
    // payloads
    docsAndPositionsEnum = termsEnum.postings(null, PostingsEnum.PAYLOADS);
    assertNotNull(docsAndPositionsEnum);
    assertEquals(-1, docsAndPositionsEnum.docID());
    assertEquals(0, docsAndPositionsEnum.nextDoc());
    assertEquals(2, docsAndPositionsEnum.freq());
    assertEquals(0, docsAndPositionsEnum.nextPosition());
    assertEquals(-1, docsAndPositionsEnum.startOffset());
    assertEquals(-1, docsAndPositionsEnum.endOffset());
    assertEquals(new BytesRef("pay1"), docsAndPositionsEnum.getPayload());
    assertEquals(1, docsAndPositionsEnum.nextPosition());
    assertEquals(-1, docsAndPositionsEnum.startOffset());
    assertEquals(-1, docsAndPositionsEnum.endOffset());
    assertEquals(new BytesRef("pay2"), docsAndPositionsEnum.getPayload());
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsAndPositionsEnum.nextDoc());
    // reuse
    docsAndPositionsEnum2 = termsEnum.postings(docsAndPositionsEnum, PostingsEnum.PAYLOADS);
    assertEquals(-1, docsAndPositionsEnum2.docID());
    assertEquals(0, docsAndPositionsEnum2.nextDoc());
    assertEquals(2, docsAndPositionsEnum2.freq());
    assertEquals(0, docsAndPositionsEnum2.nextPosition());
    assertEquals(-1, docsAndPositionsEnum2.startOffset());
    assertEquals(-1, docsAndPositionsEnum2.endOffset());
    assertEquals(new BytesRef("pay1"), docsAndPositionsEnum2.getPayload());
    assertEquals(1, docsAndPositionsEnum2.nextPosition());
    assertEquals(-1, docsAndPositionsEnum2.startOffset());
    assertEquals(-1, docsAndPositionsEnum2.endOffset());
    assertEquals(new BytesRef("pay2"), docsAndPositionsEnum2.getPayload());
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsAndPositionsEnum2.nextDoc());
    docsAndPositionsEnum = termsEnum.postings(null, PostingsEnum.OFFSETS);
    assertNotNull(docsAndPositionsEnum);
    assertEquals(-1, docsAndPositionsEnum.docID());
    assertEquals(0, docsAndPositionsEnum.nextDoc());
    assertEquals(2, docsAndPositionsEnum.freq());
    assertEquals(0, docsAndPositionsEnum.nextPosition());
    assertEquals(-1, docsAndPositionsEnum.startOffset());
    assertEquals(-1, docsAndPositionsEnum.endOffset());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsAndPositionsEnum.getPayload() == null || new BytesRef("pay1").equals(docsAndPositionsEnum.getPayload()));
    assertEquals(1, docsAndPositionsEnum.nextPosition());
    assertEquals(-1, docsAndPositionsEnum.startOffset());
    assertEquals(-1, docsAndPositionsEnum.endOffset());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsAndPositionsEnum.getPayload() == null || new BytesRef("pay2").equals(docsAndPositionsEnum.getPayload()));
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsAndPositionsEnum.nextDoc());
    // reuse
    docsAndPositionsEnum2 = termsEnum.postings(docsAndPositionsEnum, PostingsEnum.OFFSETS);
    assertEquals(-1, docsAndPositionsEnum2.docID());
    assertEquals(0, docsAndPositionsEnum2.nextDoc());
    assertEquals(2, docsAndPositionsEnum2.freq());
    assertEquals(0, docsAndPositionsEnum2.nextPosition());
    assertEquals(-1, docsAndPositionsEnum2.startOffset());
    assertEquals(-1, docsAndPositionsEnum2.endOffset());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsAndPositionsEnum2.getPayload() == null || new BytesRef("pay1").equals(docsAndPositionsEnum2.getPayload()));
    assertEquals(1, docsAndPositionsEnum2.nextPosition());
    assertEquals(-1, docsAndPositionsEnum2.startOffset());
    assertEquals(-1, docsAndPositionsEnum2.endOffset());
    // we don't define what it is, but if its something else, we should look into it?
    assertTrue(docsAndPositionsEnum2.getPayload() == null || new BytesRef("pay2").equals(docsAndPositionsEnum2.getPayload()));
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsAndPositionsEnum2.nextDoc());
    docsAndPositionsEnum = termsEnum.postings(null, PostingsEnum.ALL);
    assertNotNull(docsAndPositionsEnum);
    assertEquals(-1, docsAndPositionsEnum.docID());
    assertEquals(0, docsAndPositionsEnum.nextDoc());
    assertEquals(2, docsAndPositionsEnum.freq());
    assertEquals(0, docsAndPositionsEnum.nextPosition());
    assertEquals(-1, docsAndPositionsEnum.startOffset());
    assertEquals(-1, docsAndPositionsEnum.endOffset());
    assertEquals(new BytesRef("pay1"), docsAndPositionsEnum.getPayload());
    assertEquals(1, docsAndPositionsEnum.nextPosition());
    assertEquals(-1, docsAndPositionsEnum.startOffset());
    assertEquals(-1, docsAndPositionsEnum.endOffset());
    assertEquals(new BytesRef("pay2"), docsAndPositionsEnum.getPayload());
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsAndPositionsEnum.nextDoc());
    docsAndPositionsEnum2 = termsEnum.postings(docsAndPositionsEnum, PostingsEnum.ALL);
    assertEquals(-1, docsAndPositionsEnum2.docID());
    assertEquals(0, docsAndPositionsEnum2.nextDoc());
    assertEquals(2, docsAndPositionsEnum2.freq());
    assertEquals(0, docsAndPositionsEnum2.nextPosition());
    assertEquals(-1, docsAndPositionsEnum2.startOffset());
    assertEquals(-1, docsAndPositionsEnum2.endOffset());
    assertEquals(new BytesRef("pay1"), docsAndPositionsEnum2.getPayload());
    assertEquals(1, docsAndPositionsEnum2.nextPosition());
    assertEquals(-1, docsAndPositionsEnum2.startOffset());
    assertEquals(-1, docsAndPositionsEnum2.endOffset());
    assertEquals(new BytesRef("pay2"), docsAndPositionsEnum2.getPayload());
    assertEquals(DocIdSetIterator.NO_MORE_DOCS, docsAndPositionsEnum2.nextDoc());
    iw.close();
    reader.close();
    dir.close();
}

Example 29 with CannedTokenStream

use of org.apache.lucene.analysis.CannedTokenStream in project lucene-solr by apache.

the class TestMemoryIndexAgainstRAMDir method testEmptyString.

// LUCENE-4880
public void testEmptyString() throws IOException {
    MemoryIndex memory = new MemoryIndex();
    memory.addField("foo", new CannedTokenStream(new Token("", 0, 5)));
    IndexSearcher searcher = memory.createSearcher();
    TopDocs docs = searcher.search(new TermQuery(new Term("foo", "")), 10);
    assertEquals(1, docs.totalHits);
    TestUtil.checkReader(searcher.getIndexReader());
}

Example 30 with CannedTokenStream

use of org.apache.lucene.analysis.CannedTokenStream in project lucene-solr by apache.

the class TestTermAutomatonQuery method testAnyFromTokenStream.

public void testAnyFromTokenStream() throws Exception {
    Directory dir = newDirectory();
    RandomIndexWriter w = new RandomIndexWriter(random(), dir);
    Document doc = new Document();
    doc.add(newTextField("field", "here comes the sun", Field.Store.NO));
    w.addDocument(doc);
    doc = new Document();
    doc.add(newTextField("field", "here comes the moon", Field.Store.NO));
    w.addDocument(doc);
    doc = new Document();
    doc.add(newTextField("field", "here comes sun", Field.Store.NO));
    w.addDocument(doc);
    // Should not match:
    doc = new Document();
    doc.add(newTextField("field", "here comes the other sun", Field.Store.NO));
    w.addDocument(doc);
    IndexReader r = w.getReader();
    IndexSearcher s = newSearcher(r);
    TokenStream ts = new CannedTokenStream(new Token[] { token("comes", 1, 1), token("comes", 0, 2), token("*", 1, 1), token("sun", 1, 1), token("moon", 0, 1) });
    TermAutomatonQuery q = new TokenStreamToTermAutomatonQuery().toQuery("field", ts);
    // System.out.println("DOT: " + q.toDot());
    assertEquals(3, s.search(q, 1).totalHits);
    w.close();
    r.close();
    dir.close();
}