Lucene BooleanQuery wrong result

I created a Lucene RAMDirectory to collect data from different sources and make them quickly searchable. I spent many hours to understand the different analyzers and index strategies, but in some cases the query result is not the expected.

Here is a demo class:

class LuceneDemo {

    static final String ANIMAL = "animal";
    static final String PERSON = "person";

    private StandardAnalyzer analyzer = new StandardAnalyzer();

    private IndexSearcher searcher;
    private IndexWriter writer;

    LuceneDemo() {
        Directory ramDirectory = new RAMDirectory();
        IndexWriterConfig config = new IndexWriterConfig(analyzer);
        try {
            writer = new IndexWriter(ramDirectory, config);

            addDocument(createDocument(PERSON, "DR-(frankenstein)"));
            addDocument(createDocument(ANIMAL, "gray fox"));
            addDocument(createDocument(ANIMAL, "island fox"));

            writer.close();
            IndexReader reader = DirectoryReader.open(ramDirectory);
            searcher = new IndexSearcher(reader);
        } catch (IOException e) {
            e.printStackTrace();
        }
    }

    private Document createDocument(String type, String value) {
        Document document = new Document();
        document.add(new TextField("type", type, Field.Store.YES));
        document.add(new TextField("name", value, Field.Store.YES));
        document.add(new StringField("name", value, Field.Store.YES));
        return document;
    }

    private void addDocument(Document document) {
        try {
            writer.addDocument(document);
        } catch (IOException e) {
            e.printStackTrace();
        }
    }

    List<String> getDocuments(String type, String value) {
        value = "*" + QueryParser.escape(value) + "*";
        try {
            QueryParser queryParser = new QueryParser("name", analyzer);
            queryParser.setAllowLeadingWildcard(true);
            queryParser.setDefaultOperator(QueryParser.Operator.AND);

            BooleanQuery.Builder query = new BooleanQuery.Builder();
            query.add(new TermQuery(new Term("type", type)), BooleanClause.Occur.MUST);
            query.add(queryParser.parse(value), BooleanClause.Occur.MUST);

            TopDocs docs = searcher.search(query.build(), 10);

            return Arrays.stream(docs.scoreDocs).map(scoreDoc -> {
                try {
                    return searcher.doc(scoreDoc.doc).get("name");
                } catch (IOException e) {
                    return "";
                }
            }).collect(Collectors.toList());
        } catch (ParseException | IOException e) {
            e.printStackTrace();
        }
        return Collections.emptyList();
    }
}

If I search for "ox", "gray fox" or "-(frankenstein)", the code works pretty well. But I have no result for "DR-(frankenstein)". I have no idea what I did wrong. So any suggestions are welcome.

// OK
luceneDemo.getDocuments(LuceneDemo.ANIMAL, "ox").forEach(System.out::println);
luceneDemo.getDocuments(LuceneDemo.ANIMAL, "gray fox").forEach(System.out::println);
luceneDemo.getDocuments(LuceneDemo.PERSON, "-(frankenstein)").forEach(System.out::println);

// NOT OK
luceneDemo.getDocuments(LuceneDemo.PERSON, "DR-(frankenstein)").forEach(System.out::println);

This is how your documents are indexed -

1. doc#1 type:person name:dr name:frankenstein name:DR-(frankenstein) (Note: StringField is not tokenized and is not converted to lowercase)
2. doc#2 type:animal name:gray name:fox name:gray fox
3. doc#3 type:animal name:island name:fox name:island fox

Basically StringField indexes field irrespective of the analyzer - without tokenizing, and lowering the case. Whereas the reader is using StandardAnalyzer and lowers the case for all search. Hence, searching for "DR-(frankenstein)" searches for "dr-(frankenstein)" which doesn't have a match.

To make your code work using StandardAnalyzer you need to index StringField as lowercase.

document.add(new StringField("name", value.toLowerCase(), Field.Store.YES));