Lucene TermQuery 和 QueryParser

2022-01-15 00:00:00 lucene java

我有 2 个 lucene 查询:

I have 2 lucene queries:

1)

Term term = new Term(Properties.LUCENE_APPARTMENT_ADDRESS,address);
Query termQuery = new TermQuery(term);
TopDocs topDocs = indexSearcher.search(termQuery, 20); 

和2)

QueryParser queryParser = new QueryParser(Version.LUCENE_36, Properties.LUCENE_APPARTMENT_ADDRESS, analyzer);
Query query = queryParser.parse(address);
ScoreDoc[] queryResults = indexSearcher.search(query, 20).scoreDocs; 

为什么第一个返回 1 个结果,而第二个什么也不返回?(地址为一个字以上)

Why the first one returns 1 result where the 2nd returns nothing? (the address is one word or more)

推荐答案

当您使用 QueryParser 时,它会使用分析器执行与索引期间相同的操作序列(标记化、小写、停用词等).

When you use QueryParser, it uses analyzer which does the same sequence of actions as when during the indexing (tokenization, lowercasing, stopwords, etc.).

当您使用原始 TermQuery 时,您需要自己完成所有这些步骤.

When you use raw TermQuery, you need to do all these steps yourself.

我猜你的分析器对 Properties.LUCENE_APPARTMENT_ADDRESS 做了一些特别的事情,而你在将地址传递给 TermQuery 时没有这样做,因此搜索结果存在差异.

I guess your analyzer does something special about Properties.LUCENE_APPARTMENT_ADDRESS and you are not when passing the address to TermQuery hence the search results discrepancy.

相关文章