Hibernate/Lucene搜索：通配符查询

Question

我使用这个分析器创建了一个模型：

@Indexed
@AnalyzerDef(name = "abAnalyzer", tokenizer = @TokenizerDef(factory = StandardTokenizerFactory.class), filters =
{
    @TokenFilterDef(factory = StandardFilterFactory.class),
    @TokenFilterDef(factory = LowerCaseFilterFactory.class),
    @TokenFilterDef(factory = StopFilterFactory.class, params =
    {
        @Parameter(name = "ignoreCase", value = "true")
    })
})
public class Foo
{
    ...
    @Field(index = Index.YES, analyze = Analyze.YES, store = Store.NO)
    @Analyzer(definition = "abAnalyzer")
    private String name;

    ...
}

我实现了如下所示的查询。我得到了所有预期的结果，除了像“A.B.C”这样的结果。我做错了什么？我哪里误解了事情？

public List<Foo> getResults(final String searchName)
{
    Session session = this.sessionFactory.openSession();
    FullTextSession fullTextSession = Search.getFullTextSession(session);
    Transaction tx = fullTextSession.beginTransaction();

    BooleanQuery bQuery = new BooleanQuery();
    Analyzer analyzer = fullTextSession.getSearchFactory().getAnalyzer("abAnalyzer");

    QueryParser qp = new QueryParser(Version.LUCENE_36, "name", analyzer);
    String cleanedText = qp.parse(searchName).toString("name");
    String[] tokenized = cleanedText.split(""); // split on each character

    QueryBuilder qBuilder = fullTextSession.getSearchFactory().buildQueryBuilder().forEntity(Foo.class).get();

    org.apache.lucene.search.Query query = qBuilder.keyword().wildcard().onField("name").matching("*" + cleanedText + "*").createQuery();
    bQuery.add(query, BooleanClause.Occur.SHOULD);

    query = qBuilder.keyword().wildcard().onField("name").matching("*" + createSearchString(cleanedText) + "*").createQuery();
    bQuery.add(query, BooleanClause.Occur.SHOULD);

    org.hibernate.Query hibQuery = fullTextSession.createFullTextQuery(bQuery, Company.class);    
    return hibQuery.list();
}

....

private String createSearchString(final String name)
{
    StringBuilder searchName = new StringBuilder("");
    for (int i = 0; i < name.length(); i++)
    {
        if (searchName.length() > 0)
        {
            searchName.append("?");
        }
        searchName.append(name.charAt(i));
    }

    return searchName.toString();
}

我的代码基于以下资源：

Answer 1

我想我找到了解决办法...Lucene索引是基于小写字符串生成的，索引中的“特殊”字符被删除。

于是，我把模型改成：

@Indexed
@AnalyzerDef(name = "abAnalyzer", charFilters =
{
    @CharFilterDef(factory = PatternReplaceCharFilterFactory.class, params =
    {
        @Parameter(name = "pattern", value = Company.PATTERN),
        @Parameter(name = "replacement", value = Company.REPLACEMENT_PATTERN)
    })
 }, tokenizer = @TokenizerDef(factory = StandardTokenizerFactory.class), filters =
{
    @TokenFilterDef(factory = StandardFilterFactory.class), @TokenFilterDef(factory = LowerCaseFilterFactory.class),
})
public class Foo
{
    public static final String PATTERN = "(A-Z)*[\\/\\.\\-'+&, ](A-Z)*";
    public static final String REPLACEMENT_PATTERN = "$1$2";

    @Field(index = Index.YES, analyze = Analyze.YES, store = Store.NO)
    @Analyzer(definition = "abAnalyzer")
    private String name;

    ....
}

为了进行查询,我实现了以下内容：

public List<Foo> getResults(final String searchName)
{
    List<Foo> result = new ArrayList<>();

    // remove "special chars from searchName"
    String searchName = name.replaceAll(Company.PATTERN, Company.REPLACEMENT_PATTERN);

    Session session = this.sessionFactory.openSession();
    try
    {
        FullTextSession fullTextSession = Search.getFullTextSession(session);
        Transaction tx = fullTextSession.beginTransaction();
        Analyzer analyzer = fullTextSession.getSearchFactory().getAnalyzer("abAnalyzer");

        QueryParser qp = new QueryParser(Version.LUCENE_36, "name", analyzer);
        String cleanedText = qp.parse(searchName).toString("name");

        BooleanQuery bQuery = new BooleanQuery();
        bQuery.add(new WildcardQuery(new Term("name", "*" + cleanedText + "*")), BooleanClause.Occur.SHOULD);

        org.hibernate.Query hibQuery = fullTextSession.createFullTextQuery(bQuery, Company.class);

        result = hibQuery.list();
        tx.commit();
    }
    catch (Exception e)
    {
        // TODO Auto-generated catch block
        e.printStackTrace();
    }
    finally
    {
        session.close();
    }

    return result;
}