Examples with TopDocs org.apache.lucene.search.TopDocs used on opensource projects

Example 71 with TopDocs

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

the class TestUnifiedHighlighter method testBasics.

//
//  Tests below were ported from the PostingsHighlighter. Possibly augmented.  Far below are newer tests.
//
public void testBasics() throws Exception {
    RandomIndexWriter iw = new RandomIndexWriter(random(), dir, indexAnalyzer);
    Field body = new Field("body", "", fieldType);
    Document doc = new Document();
    doc.add(body);
    body.setStringValue("This is a test. Just a test highlighting from postings. Feel free to ignore.");
    iw.addDocument(doc);
    body.setStringValue("Highlighting the first term. Hope it works.");
    iw.addDocument(doc);
    IndexReader ir = iw.getReader();
    iw.close();
    IndexSearcher searcher = newSearcher(ir);
    UnifiedHighlighter highlighter = new UnifiedHighlighter(searcher, indexAnalyzer);
    Query query = new TermQuery(new Term("body", "highlighting"));
    TopDocs topDocs = searcher.search(query, 10, Sort.INDEXORDER);
    assertEquals(2, topDocs.totalHits);
    String[] snippets = highlighter.highlight("body", query, topDocs);
    assertEquals(2, snippets.length);
    assertEquals("Just a test <b>highlighting</b> from postings. ", snippets[0]);
    assertEquals("<b>Highlighting</b> the first term. ", snippets[1]);
    ir.close();
}

Example 72 with TopDocs

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

the class TestUnifiedHighlighter method testMultipleTerms.

public void testMultipleTerms() throws Exception {
    RandomIndexWriter iw = new RandomIndexWriter(random(), dir, indexAnalyzer);
    Field body = new Field("body", "", fieldType);
    Document doc = new Document();
    doc.add(body);
    body.setStringValue("This is a test. Just a test highlighting from postings. Feel free to ignore.");
    iw.addDocument(doc);
    body.setStringValue("Highlighting the first term. Hope it works.");
    iw.addDocument(doc);
    IndexReader ir = iw.getReader();
    iw.close();
    IndexSearcher searcher = newSearcher(ir);
    UnifiedHighlighter highlighter = new UnifiedHighlighter(searcher, indexAnalyzer);
    BooleanQuery query = new BooleanQuery.Builder().add(new TermQuery(new Term("body", "highlighting")), BooleanClause.Occur.SHOULD).add(new TermQuery(new Term("body", "just")), BooleanClause.Occur.SHOULD).add(new TermQuery(new Term("body", "first")), BooleanClause.Occur.SHOULD).build();
    TopDocs topDocs = searcher.search(query, 10, Sort.INDEXORDER);
    assertEquals(2, topDocs.totalHits);
    String[] snippets = highlighter.highlight("body", query, topDocs);
    assertEquals(2, snippets.length);
    assertEquals("<b>Just</b> a test <b>highlighting</b> from postings. ", snippets[0]);
    assertEquals("<b>Highlighting</b> the <b>first</b> term. ", snippets[1]);
    ir.close();
}

Example 73 with TopDocs

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

the class TokenSourcesTest method testOverlapWithPositionsAndOffset.

public void testOverlapWithPositionsAndOffset() throws IOException, InvalidTokenOffsetsException {
    final String TEXT = "the fox did not jump";
    final Directory directory = newDirectory();
    final IndexWriter indexWriter = new IndexWriter(directory, newIndexWriterConfig(null));
    try {
        final Document document = new Document();
        FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
        customType.setStoreTermVectors(true);
        customType.setStoreTermVectorOffsets(true);
        customType.setStoreTermVectorPositions(true);
        document.add(new Field(FIELD, new OverlappingTokenStream(), customType));
        indexWriter.addDocument(document);
    } finally {
        indexWriter.close();
    }
    final IndexReader indexReader = DirectoryReader.open(directory);
    try {
        assertEquals(1, indexReader.numDocs());
        final IndexSearcher indexSearcher = newSearcher(indexReader);
        final DisjunctionMaxQuery query = new DisjunctionMaxQuery(Arrays.asList(new SpanTermQuery(new Term(FIELD, "{fox}")), new SpanTermQuery(new Term(FIELD, "fox"))), 1);
        // final Query phraseQuery = new SpanNearQuery(new SpanQuery[] {
        // new SpanTermQuery(new Term(FIELD, "{fox}")),
        // new SpanTermQuery(new Term(FIELD, "fox")) }, 0, true);
        TopDocs hits = indexSearcher.search(query, 1);
        assertEquals(1, hits.totalHits);
        final Highlighter highlighter = new Highlighter(new SimpleHTMLFormatter(), new SimpleHTMLEncoder(), new QueryScorer(query));
        final TokenStream tokenStream = TokenSources.getTermVectorTokenStreamOrNull(FIELD, indexReader.getTermVectors(0), -1);
        assertEquals("<B>the fox</B> did not jump", highlighter.getBestFragment(tokenStream, TEXT));
    } finally {
        indexReader.close();
        directory.close();
    }
}

Example 74 with TopDocs

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

the class TokenSourcesTest method testOverlapWithPositionsAndOffsetExactPhrase.

public void testOverlapWithPositionsAndOffsetExactPhrase() throws IOException, InvalidTokenOffsetsException {
    final String TEXT = "the fox did not jump";
    final Directory directory = newDirectory();
    final IndexWriter indexWriter = new IndexWriter(directory, newIndexWriterConfig(null));
    try {
        final Document document = new Document();
        FieldType customType = new FieldType(TextField.TYPE_NOT_STORED);
        customType.setStoreTermVectors(true);
        customType.setStoreTermVectorPositions(true);
        customType.setStoreTermVectorOffsets(true);
        document.add(new Field(FIELD, new OverlappingTokenStream(), customType));
        indexWriter.addDocument(document);
    } finally {
        indexWriter.close();
    }
    final IndexReader indexReader = DirectoryReader.open(directory);
    try {
        assertEquals(1, indexReader.numDocs());
        final IndexSearcher indexSearcher = newSearcher(indexReader);
        // final DisjunctionMaxQuery query = new DisjunctionMaxQuery(1);
        // query.add(new SpanTermQuery(new Term(FIELD, "the")));
        // query.add(new SpanTermQuery(new Term(FIELD, "fox")));
        final Query phraseQuery = new SpanNearQuery(new SpanQuery[] { new SpanTermQuery(new Term(FIELD, "the")), new SpanTermQuery(new Term(FIELD, "fox")) }, 0, true);
        TopDocs hits = indexSearcher.search(phraseQuery, 1);
        assertEquals(1, hits.totalHits);
        final Highlighter highlighter = new Highlighter(new SimpleHTMLFormatter(), new SimpleHTMLEncoder(), new QueryScorer(phraseQuery));
        final TokenStream tokenStream = TokenSources.getTermVectorTokenStreamOrNull(FIELD, indexReader.getTermVectors(0), -1);
        assertEquals("<B>the fox</B> did not jump", highlighter.getBestFragment(tokenStream, TEXT));
    } finally {
        indexReader.close();
        directory.close();
    }
}

Example 75 with TopDocs

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

the class SynonymTokenizer method testHighlightingCommonTermsQuery.

public void testHighlightingCommonTermsQuery() throws Exception {
    CommonTermsQuery query = new CommonTermsQuery(Occur.MUST, Occur.SHOULD, 3);
    //stop-word
    query.add(new Term(FIELD_NAME, "this"));
    query.add(new Term(FIELD_NAME, "long"));
    query.add(new Term(FIELD_NAME, "very"));
    searcher = newSearcher(reader);
    TopDocs hits = searcher.search(query, 10, new Sort(SortField.FIELD_DOC, SortField.FIELD_SCORE));
    assertEquals(2, hits.totalHits);
    QueryScorer scorer = new QueryScorer(query, FIELD_NAME);
    Highlighter highlighter = new Highlighter(scorer);
    final int docId0 = hits.scoreDocs[0].doc;
    Document doc = searcher.doc(docId0);
    String storedField = doc.get(FIELD_NAME);
    TokenStream stream = getAnyTokenStream(FIELD_NAME, docId0);
    Fragmenter fragmenter = new SimpleSpanFragmenter(scorer);
    highlighter.setTextFragmenter(fragmenter);
    String fragment = highlighter.getBestFragment(stream, storedField);
    assertEquals("Hello this is a piece of text that is <B>very</B> <B>long</B> and contains too much preamble and the meat is really here which says kennedy has been shot", fragment);
    final int docId1 = hits.scoreDocs[1].doc;
    doc = searcher.doc(docId1);
    storedField = doc.get(FIELD_NAME);
    stream = getAnyTokenStream(FIELD_NAME, docId1);
    highlighter.setTextFragmenter(new SimpleSpanFragmenter(scorer));
    fragment = highlighter.getBestFragment(stream, storedField);
    assertEquals("This piece of text refers to Kennedy at the beginning then has a longer piece of text that is <B>very</B>", fragment);
}