java - Hibernate 搜索 - 通配符和空格

Question

我正在将 Hibernate Search 与多个单词/字段一起使用，并且在启用通配符之前它会按预期工作。这是我的实体分析器:

@AnalyzerDef(name = "autocompleteAnalyzer",
        tokenizer = @TokenizerDef(factory = StandardTokenizerFactory.class),
        filters = {
                // remove accents
                @TokenFilterDef(factory = ASCIIFoldingFilterFactory.class),
                // lower case
                @TokenFilterDef(factory = LowerCaseFilterFactory.class),
                // Start with same root
                @TokenFilterDef(
                        factory = SnowballPorterFilterFactory.class,
                        params = { @Parameter(name = "language", value = "English") })
        })

这是我的查询:

    if(criteria.length() > 0) {
        fullTextQuery = queryBuilder
                .keyword()
                // .wildcard() => not in use 
                .onFields("firstName", "lastName", "extraName", "biography")
                .matching(criteria)
                .createQuery();

场景01
如果我搜索“John Smith”之类的内容，我会得到以下查询:

+((firstName:john firstName:smith) 
(lastName:john lastName:smith) 
(extraName:john extraName:smith) 
(biography:john biography:smith))

它会找到任何名为 John 或 Smith 的人。

场景02如果我搜索“John*”，因为我想找到名称以 John(Johnny、Johson)开头的任何人，我需要启用通配符，如下所示:

fullTextQuery = queryBuilder
   .keyword()
   .wildcard()

当我输入“John*”时，它会返回以下查询:

+(firstName:john* lastName:john* extraName:john* biography:john*)

但是当我输入“John* Smith”之类的内容时，它不再起作用，因为它不会拆分单词，我认为这是由于通配符选项所致:

+(firstName:john* smith lastName:john* smith extraName:john* smith biography:john* smith)

Answer 1

关键字查询将始终与包含任何提供的关键字的文档相匹配。通配符查询不应用分析。所以这两种情况都不适合您的情况。

使用 simpleQueryString 查询，并强制默认运算符为“and”:

        fullTextQuery = queryBuilder
                .simpleQueryString()
                .onFields("firstName", "lastName", "extraName", "biography")
                .withAndAsDefaultOperator()
                .matching(criteria)
                .createQuery();

文档:https://docs.jboss.org/hibernate/search/5.11/reference/en-US/html_single/#_simple_query_string_queries