[ElasticSearch]Term精确匹配中文字符串短语

1. 给定数据

curl -XPUT 'localhost:9200/test-index/stu/1' -d'
{
"name":"陈泽鹏",
"sex":"boy",
"college":"计算机学院"
}';
curl -XPUT 'localhost:9200/test-index/stu/2' -d'
{
"name":"廖力生",
"sex":"boy",
"college":"通信学院"
}';
curl -XPUT 'localhost:9200/test-index/stu/3' -d'
{
"name":"李源一",
"sex":"girl",
"college":"计算机学院"
}';
curl -XPUT 'localhost:9200/test-index/stu/4' -d'
{
"name":"陈哲超",
"sex":"girl",
"college":"计算机学院"
}';
curl -XPUT 'localhost:9200/test-index/stu/5' -d'
{
"name":"AA",
"sex":"girl",
"college":"计算机学院"
}';
curl -XPUT 'localhost:9200/test-index/stu/6' -d'
{
"name":"bb",
"sex":"girll",
"college":"通信学院"
}';
curl -XPUT 'localhost:9200/test-index/stu/7' -d'
{
"name":"方镜淇",
"sex":"boy",
"college":"电子工程学院"
}';
curl -XPUT 'localhost:9200/test-index/stu/8' -d'
{
"name":"吴兴涵",
"sex":"boy",
"college":"计算机学院"
}';

2. 需求

我们想精确匹配出来自计算机学院的学生，所以我们就实现如下语句：

QueryBuilder queryBuilder = QueryBuilders.termQuery("college", "计算机学院");
// Search
SearchRequestBuilder searchRequestBuilder = client.prepareSearch(index);
searchRequestBuilder.setTypes(type);
searchRequestBuilder.setQuery(queryBuilder);
// 执行
SearchResponse searchResponse = searchRequestBuilder.execute().actionGet();

返回结果：

20:30:46.815 [main] INFO  com.sjf.open.api.TermLevelQuery - ----------termMatch size 0

表示没有找到来自计算机学院的同学，这咋可能。。。。

3. 解决方案

3.1 解决方案一　

把计算机学院拆分成多个汉字，再利用bool查询查找：

// Query
BoolQueryBuilder boolQueryBuilder = QueryBuilders.boolQuery();
boolQueryBuilder.must(QueryBuilders.termQuery("college", "计"));
boolQueryBuilder.must(QueryBuilders.termQuery("college", "算"));
boolQueryBuilder.must(QueryBuilders.termQuery("college", "机"));
boolQueryBuilder.must(QueryBuilders.termQuery("college", "学"));
boolQueryBuilder.must(QueryBuilders.termQuery("college", "院"));
// Search
SearchRequestBuilder searchRequestBuilder = client.prepareSearch(index);
searchRequestBuilder.setTypes(type);
searchRequestBuilder.setQuery(boolQueryBuilder);
// 执行
SearchResponse searchResponse = searchRequestBuilder.execute().actionGet();

返回结果：

20:34:53.022 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 5 source {college=计算机学院, sex=girll, name=AA}
20:34:53.028 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 8 source {college=计算机学院, sex=boy, name=吴兴涵}
20:34:53.029 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 1 source {college=计算机学院, sex=boy, name=陈泽鹏}
20:34:53.029 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 7 source {college=计算机学院, sex=boy, name=陈哲超}
20:34:53.029 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 3 source {college=计算机学院, sex=girll, name=李源一}

3.2 解决方案二

设置字段属性，设置为not_analyzed，再插入上面数据：

curl -XPUT 'localhost:9200/test-index' -d'
{
"mappings":{
"stu":{
"properties":{
"name":{
"type":"string"
},
"sex":{
"type":"string"
},
"college":{
"type":"string",
"index":"not_analyzed"
}
}
}
}
}';

具体分析：https://www.elastic.co/guide/en/elasticsearch/reference/2.3/query-dsl-term-query.html再次查询：

QueryBuilder queryBuilder = QueryBuilders.termQuery("college", "计算机学院");
// Search
SearchRequestBuilder searchRequestBuilder = client.prepareSearch(index);
searchRequestBuilder.setTypes(type);
searchRequestBuilder.setQuery(queryBuilder);
// 执行
SearchResponse searchResponse = searchRequestBuilder.execute().actionGet();

输出结果：

20:34:53.022 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 5 source {college=计算机学院, sex=girll, name=AA}
20:34:53.028 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 8 source {college=计算机学院, sex=boy, name=吴兴涵}
20:34:53.029 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 1 source {college=计算机学院, sex=boy, name=陈泽鹏}
20:34:53.029 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 7 source {college=计算机学院, sex=boy, name=陈哲超}
20:34:53.029 [main] INFO  com.sjf.open.api.Search - ----------hit source: id 3 source {college=计算机学院, sex=girll, name=李源一}