public void TestWhichMTQMatched()
{
Directory dir = NewDirectory();
// use simpleanalyzer for more natural tokenization (else "test." is a token)
Analyzer analyzer = new MockAnalyzer(Random, MockTokenizer.SIMPLE, true);
IndexWriterConfig iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, analyzer);
iwc.SetMergePolicy(NewLogMergePolicy());
RandomIndexWriter iw = new RandomIndexWriter(Random, dir, iwc);
FieldType offsetsType = new FieldType(TextField.TYPE_STORED);
offsetsType.IndexOptions = (IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS);
Field body = new Field("body", "", offsetsType);
Document doc = new Document();
doc.Add(body);
body.SetStringValue("Test a one sentence document.");
iw.AddDocument(doc);
IndexReader ir = iw.GetReader();
iw.Dispose();
IndexSearcher searcher = NewSearcher(ir);
ICUPostingsHighlighter highlighter = new PostingsHighlighterAnalyzerHelper(analyzer);
BooleanQuery query = new BooleanQuery();
query.Add(new WildcardQuery(new Term("body", "te*")), Occur.SHOULD);
query.Add(new WildcardQuery(new Term("body", "one")), Occur.SHOULD);
query.Add(new WildcardQuery(new Term("body", "se*")), Occur.SHOULD);
TopDocs topDocs = searcher.Search(query, null, 10, Sort.INDEXORDER);
assertEquals(1, topDocs.TotalHits);
String[] snippets = highlighter.Highlight("body", query, searcher, topDocs);
assertEquals(1, snippets.Length);
// Default formatter just bolds each hit:
assertEquals("<b>Test</b> a <b>one</b> <b>sentence</b> document.", snippets[0]);
// Now use our own formatter, that also stuffs the
// matching term's text into the result:
highlighter = new PostingsHighlighterAnalyzerAndFormatterHelper(analyzer, new PassageFormatterHelper());
assertEquals(1, topDocs.TotalHits);
snippets = highlighter.Highlight("body", query, searcher, topDocs);
assertEquals(1, snippets.Length);
// Default formatter bolds each hit:
assertEquals("<b>Test(body:te*)</b> a <b>one(body:one)</b> <b>sentence(body:se*)</b> document.", snippets[0]);
ir.Dispose();
dir.Dispose();
}
public void TestWhichMTQMatched()
{
Directory dir = NewDirectory();
// use simpleanalyzer for more natural tokenization (else "test." is a token)
Analyzer analyzer = new MockAnalyzer(Random(), MockTokenizer.SIMPLE, true);
IndexWriterConfig iwc = NewIndexWriterConfig(TEST_VERSION_CURRENT, analyzer);
iwc.SetMergePolicy(NewLogMergePolicy());
RandomIndexWriter iw = new RandomIndexWriter(Random(), dir, iwc);
FieldType offsetsType = new FieldType(TextField.TYPE_STORED);
offsetsType.IndexOptions = (IndexOptions.DOCS_AND_FREQS_AND_POSITIONS_AND_OFFSETS);
Field body = new Field("body", "", offsetsType);
Document doc = new Document();
doc.Add(body);
body.SetStringValue("Test a one sentence document.");
iw.AddDocument(doc);
IndexReader ir = iw.Reader;
iw.Dispose();
IndexSearcher searcher = NewSearcher(ir);
PostingsHighlighter highlighter = new PostingsHighlighterAnalyzerHelper(analyzer);
//PostingsHighlighter highlighter = new PostingsHighlighter() {
// @Override
// protected Analyzer getIndexAnalyzer(String field)
//{
// return analyzer;
//}
// };
BooleanQuery query = new BooleanQuery();
query.Add(new WildcardQuery(new Term("body", "te*")), Occur.SHOULD);
query.Add(new WildcardQuery(new Term("body", "one")), Occur.SHOULD);
query.Add(new WildcardQuery(new Term("body", "se*")), Occur.SHOULD);
TopDocs topDocs = searcher.Search(query, null, 10, Sort.INDEXORDER);
assertEquals(1, topDocs.TotalHits);
String[] snippets = highlighter.Highlight("body", query, searcher, topDocs);
assertEquals(1, snippets.Length);
// Default formatter just bolds each hit:
assertEquals("<b>Test</b> a <b>one</b> <b>sentence</b> document.", snippets[0]);
// Now use our own formatter, that also stuffs the
// matching term's text into the result:
highlighter = new PostingsHighlighterAnalyzerAndFormatterHelper(analyzer, new PassageFormatterHelper());
//highlighter = new PostingsHighlighter()
//{
// @Override
// protected Analyzer getIndexAnalyzer(String field)
//{
// return analyzer;
//}
//@Override
// protected PassageFormatter getFormatter(String field)
//{
// return new PassageFormatter() {
// @Override
// public Object format(Passage passages[], String content)
//{
// // Copied from DefaultPassageFormatter, but
// // tweaked to include the matched term:
// StringBuilder sb = new StringBuilder();
// int pos = 0;
// for (Passage passage : passages)
// {
// // don't add ellipsis if its the first one, or if its connected.
// if (passage.startOffset > pos && pos > 0)
// {
// sb.append("... ");
// }
// pos = passage.startOffset;
// for (int i = 0; i < passage.numMatches; i++)
// {
// int start = passage.matchStarts[i];
// int end = passage.matchEnds[i];
// // its possible to have overlapping terms
// if (start > pos)
// {
// sb.append(content, pos, start);
// }
// if (end > pos)
// {
// sb.append("<b>");
// sb.append(content, Math.max(pos, start), end);
// sb.append('(');
// sb.append(passage.getMatchTerms()[i].utf8ToString());
// sb.append(')');
// sb.append("</b>");
// pos = end;
// }
// }
// // its possible a "term" from the analyzer could span a sentence boundary.
// sb.append(content, pos, Math.max(pos, passage.endOffset));
// pos = passage.endOffset;
// }
// return sb.toString();
//}
// };
// }
// };
assertEquals(1, topDocs.TotalHits);
snippets = highlighter.Highlight("body", query, searcher, topDocs);
assertEquals(1, snippets.Length);
// Default formatter bolds each hit:
assertEquals("<b>Test(body:te*)</b> a <b>one(body:one)</b> <b>sentence(body:se*)</b> document.", snippets[0]);
ir.Dispose();
dir.Dispose();
}
