Examples with SpanTermQuery org.apache.lucene.search.spans.SpanTermQuery used on opensource projects

Example 86 with SpanTermQuery

use of org.apache.lucene.search.spans.SpanTermQuery in project lucene-solr by apache.

the class SynonymTokenizer method testPayloadQuery.

/** We can highlight based on payloads. It's supported both via term vectors and MemoryIndex since Lucene 5. */
public void testPayloadQuery() throws IOException, InvalidTokenOffsetsException {
    //"words" at positions 1 & 4
    final String text = "random words and words";
    //sets payload to "pos: X" (where X is position #)
    Analyzer analyzer = new MockPayloadAnalyzer();
    try (IndexWriter writer = new IndexWriter(dir, new IndexWriterConfig(analyzer))) {
        writer.deleteAll();
        Document doc = new Document();
        doc.add(new Field(FIELD_NAME, text, fieldType));
        writer.addDocument(doc);
        writer.commit();
    }
    try (IndexReader reader = DirectoryReader.open(dir)) {
        Query query = new SpanPayloadCheckQuery(new SpanTermQuery(new Term(FIELD_NAME, "words")), //just match the first "word" occurrence
        Collections.singletonList(new BytesRef("pos: 1")));
        IndexSearcher searcher = newSearcher(reader);
        QueryScorer scorer = new QueryScorer(query, searcher.getIndexReader(), FIELD_NAME);
        scorer.setUsePayloads(true);
        Highlighter h = new Highlighter(scorer);
        TopDocs hits = searcher.search(query, 10);
        assertEquals(1, hits.scoreDocs.length);
        TokenStream stream = TokenSources.getAnyTokenStream(searcher.getIndexReader(), 0, FIELD_NAME, analyzer);
        if (random().nextBoolean()) {
            //conceals detection of TokenStreamFromTermVector
            stream = new CachingTokenFilter(stream);
        }
        String result = h.getBestFragment(stream, text);
        //only highlight first "word"
        assertEquals("random <B>words</B> and words", result);
    }
}

Example 87 with SpanTermQuery

use of org.apache.lucene.search.spans.SpanTermQuery in project lucene-solr by apache.

the class TestQueryRescorer method testMissingSecondPassScore.

public void testMissingSecondPassScore() throws Exception {
    Directory dir = newDirectory();
    RandomIndexWriter w = new RandomIndexWriter(random(), dir, newIndexWriterConfig());
    Document doc = new Document();
    doc.add(newStringField("id", "0", Field.Store.YES));
    doc.add(newTextField("field", "wizard the the the the the oz", Field.Store.NO));
    w.addDocument(doc);
    doc = new Document();
    doc.add(newStringField("id", "1", Field.Store.YES));
    // 1 extra token, but wizard and oz are close;
    doc.add(newTextField("field", "wizard oz the the the the the the", Field.Store.NO));
    w.addDocument(doc);
    IndexReader r = w.getReader();
    w.close();
    // Do ordinary BooleanQuery:
    BooleanQuery.Builder bq = new BooleanQuery.Builder();
    bq.add(new TermQuery(new Term("field", "wizard")), Occur.SHOULD);
    bq.add(new TermQuery(new Term("field", "oz")), Occur.SHOULD);
    IndexSearcher searcher = getSearcher(r);
    TopDocs hits = searcher.search(bq.build(), 10);
    assertEquals(2, hits.totalHits);
    assertEquals("0", searcher.doc(hits.scoreDocs[0].doc).get("id"));
    assertEquals("1", searcher.doc(hits.scoreDocs[1].doc).get("id"));
    // Now, resort using PhraseQuery, no slop:
    PhraseQuery pq = new PhraseQuery("field", "wizard", "oz");
    TopDocs hits2 = QueryRescorer.rescore(searcher, hits, pq, 2.0, 10);
    // Resorting changed the order:
    assertEquals(2, hits2.totalHits);
    assertEquals("1", searcher.doc(hits2.scoreDocs[0].doc).get("id"));
    assertEquals("0", searcher.doc(hits2.scoreDocs[1].doc).get("id"));
    // Resort using SpanNearQuery:
    SpanTermQuery t1 = new SpanTermQuery(new Term("field", "wizard"));
    SpanTermQuery t2 = new SpanTermQuery(new Term("field", "oz"));
    SpanNearQuery snq = new SpanNearQuery(new SpanQuery[] { t1, t2 }, 0, true);
    TopDocs hits3 = QueryRescorer.rescore(searcher, hits, snq, 2.0, 10);
    // Resorting changed the order:
    assertEquals(2, hits3.totalHits);
    assertEquals("1", searcher.doc(hits3.scoreDocs[0].doc).get("id"));
    assertEquals("0", searcher.doc(hits3.scoreDocs[1].doc).get("id"));
    r.close();
    dir.close();
}