Java Lucene上的模拟弹性搜索索引和查询

Question

我将向你展示我的弹性配置，以及我是如何在Lucene上复制它的。

这是我创建索引的弹性搜索连接器：

"analysis": {
   "analyzer": {
            "default": {           
               "tokenizer": "standard",           
                "filter":  [ "lowercase", "asciifolding" ]
             }
    }     
}

以下是我的疑问：

 "query": {
      "dis_max": {
           "queries": [
                { "match": { "field1": "${term}" }}
           ],
      "tie_breaker": 1
     }
 }

 "query": {
      "dis_max": {
           "queries": [
                { "match": { "field1": "${term}" }},
                { "match": { "field2": "${term}" }},
                { "match": { "field2":  "${term}" }}
           ],
           "tie_breaker": 1
      }
 }

 val CustomStringFieldType: FieldType = {
     val fieldType = new FieldType()
     fieldType.setStoreTermVectors(true)
     fieldType.setStoreTermVectorPositions(true)
     fieldType.setOmitNorms(false)
     fieldType.setTokenized(true)
     fieldType.setStored(false)
     fieldType.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)
     fieldType.freeze()
     fieldType
   }

  def query1(tokens: Seq[String]): Query = {

     var booleanQueryBuilder = new BooleanQuery.Builder()

     for (token <- tokens) {
       booleanQueryBuilder = booleanQueryBuilder.add(
         new TermQuery(new Term("field1", token)),
         BooleanClause.Occur.SHOULD)
     }
     booleanQueryBuilder.build()
   }

 def query2(tokens: Seq[String]): Query = {

     val queries = new java.util.ArrayList[Query]()

     for (field <- ("field1", "field2", "field3") {

       var booleanQueryBuilder = new BooleanQuery.Builder()

       for (token <- tokens) {
         booleanQueryBuilder = booleanQueryBuilder.add(
           new TermQuery(new Term(field, token)),
           BooleanClause.Occur.SHOULD)
       }

       val booleanQuery: BooleanQuery = booleanQueryBuilder.build()
       queries.add(booleanQuery)
     }
     new DisjunctionMaxQuery(queries, broadQueryTieBreaker)
   }

Answer 1

明显的区别在于您的映射设置。

当您在Lucene中选择具有以下内容时

fieldType.setStoreTermVectors(true)
fieldType.setStoreTermVectorPositions(true)
fieldType.setOmitNorms(false)
fieldType.setTokenized(true)
fieldType.setStored(false)
fieldType.setIndexOptions(IndexOptions.DOCS_AND_FREQS_AND_POSITIONS)

默认情况下，在ElasticSearch中：

PUT my_index
{
  "mappings": {
    "_doc": {
      "properties": {
        "text": {
          "type":        "text",
          "term_vector": "with_positions"
        }
      }
    }
  }
}