一、背景

最近在做基于宴会厅档期的商户搜索推荐时，如果用传统平铺式的mapping结构，无法满足需求场景，于是用到了Elasticsearch支持的Nested(嵌套)查询。

二、普通对象与嵌套对象的索引异同

如果一个对象不是嵌套类型，那么以如下原数据为例：

PUT /my_index/blogpost/1  
{  "title":"Nest eggs",  "body":  "Making your money work...",  "tags":  [ "cash", "shares" ],  "comments":[  {  "name":    "John Smith",  "comment": "Great article",  "age":     28,  "stars":   4,  "date":    "2014-09-01"  },  {  "name":    "Alice White",  "comment": "More like this please",  "age":     31,  "stars":   5,  "date":    "2014-10-22"  }  ]  
}

由于是json格式的结构化文档，es会平整成索引内的一个简单键值格式，如下：

{  "title":  [ eggs, nest ],  "body":  [ making, money, work, your ],  "tags":    [ cash, shares ],  "comments.name":    [ alice, john, smith, white ],  "comments.comment":  [ article, great, like, more, please, this ],  "comments.age":      [ 28, 31 ],  "comments.stars":     [ 4, 5 ],  "comments.date":      [ 2014-09-01, 2014-10-22 ]  
}

这样的话，像这种john/28，Alice/31间的关联性就丢失了，Nested Object就是为了解决这个问题。

将comments指定为Nested类型，如下mapping:

curl -XPUT 'localhost:9200/my_index' -d '  
{  "mappings":{  "blogpost":{  "properties":{  "comments":{  "type":"nested",   //声明为nested类型"properties":{  "name":    {"type":"string"},  "comment": { "type": "string"},  "age":     { "type": "short"},  "stars":   { "type": "short"},  "date":    { "type": "date"}  }  }  }  }  }  
}

这样，每一个nested对象将会作为一个隐藏的单独文本建立索引，进而保持了nested对象的内在关联关系，如下：

{ ①  "comments.name":    [ john, smith ],  "comments.comment": [ article, great ],  "comments.age":     [ 28 ],  "comments.stars":   [ 4 ],  "comments.date":    [ 2014-09-01 ]  
}  
{   "comments.name":    [ alice, white ],  "comments.comment": [ like,more,please,this],  "comments.age":     [ 31 ],"comments.stars":   [ 5 ],  "comments.date":    [ 2014-10-22 ]  
}  
{   "title":          [ eggs, nest ],  "body":         [ making, money, work, your ],  "tags":          [ cash, shares ]  
}  
①nested object

三、嵌套对象的查询

命令查询（输出结果1）：

curl -XGET localhost:9200/yzsshopv1/shop/_search?pretty -d '{"query" : {"bool" : {"filter" : {"nested" : {"path":"hallList","query":{"bool":{"filter":{"term":{"hallList.capacityMin" : "11"}}}}}}}}}'
{"took" : 3,"timed_out" : false,"_shards" : {"total" : 5,"successful" : 5,"failed" : 0},"hits" : {"total" : 1,"max_score" : 0.0,"hits" : [ {"_index" : "yzsshopv1","_type" : "shop","_id" : "89999988","_score" : 0.0,"_source" : {"cityId" : "1","shopName" : "xxxx婚宴(yyyy店)","shopId" : "89999988","categoryId" : [ "55", "165", "2738" ],"hallList" : [ {"hallId" : "20625","schedule" : ["2017-11-10", "2017-11-09"],"capacityMax" : 16,"capacityMin" : 12},  {"hallId" : "21080","schedule" : [ "2017-12-10", "2017-09-09",  "2017-02-25"],"capacityMax" : 20,"capacityMin" : 11} ],"wedHotelTagValue" : [ "12087", "9601", "9603", "9602" ],"regionId" : [ "9", "824" ]}} ]}
}

java api查询封装：

BoolQueryBuilder boolBuilder = new BoolQueryBuilder();
NestedQueryBuilder nestedQuery = new NestedQueryBuilder("hallList", new TermQueryBuilder("hallList.capacityMin","11"));   //注意：除path之外，fieldName也要带上path (hallList)boolBuilder.filter(nestedQuery);
searchRequest.setQuery(boolBuilder); //设置查询条件

java api输出字段封装：

searchRequest.addField("shopId");
searchRequest.addField("hallList. schedule");
searchRequest.addField("hallList.capacityMin");
searchRequest.addField("hallList.capacityMax");

如果输出的outputField为searchRequest.addField("hallList")，则会报错：illegal_argument_exception,reason:field [hallList] isn't a leaf field;

如果输出的outputField为searchRequest.addField("capacityMin")，则不报错，但没有capacityMin字段的值；

正确调用search后的输出结果（输出结果2）:

{"took" : 8,"timed_out" : false,"_shards" : {"total" : 5,"successful" : 5,"failed" : 0},"hits" : {"total" : 1,"max_score" : 0.0,"hits" : [{"_index" : "yzsshopv1","_type" : "shop","_id" : "89999988","_score" : 0.0,"fields" : {"shopId" : [ "89999988" ],"hallList.hallId" : [ "20625", "21080"],"hallList.capacityMin" : [12, 11 ],"hallList.capacityMax" : [16, 20 ],"hallList.schedule" : [ "2017-11-10", "2017-11-09",  "2017-12-10", "2017-09-09",  "2017-02-25"]}}]}
}

对比输出结果1和2发现，命令输出嵌套对象结果1没问题，但通过java api输出结果2时，嵌套对象内部的关系也会打乱，比如hallList.schedule字段，无法区分到底哪些值属于hallList.hallId-20625，哪些属于21080。

//============以下更新20170331===========

经过后续调试，发现要让java api输出正确结果的嵌套对象，不能通过searchRequest.addField的方式，因为嵌套对象并不是叶子节点，需要通过以下的方式添加输出字段:

searchRequest.setFetchSource(new String[]{"shopId","hallList"},new String[]{});

还有一个不足点是： 嵌套查询请求返回的是整个文本，而不仅是匹配的nested文本。

四、参考文档

1. https://www.elastic.co/guide/en/elasticsearch/guide/master/nested-objects.html
2. http://stackoverflow.com/questions/23562192/unable-to-retrieve-nested-objects-using-elasticsearch-java-api
3. http://elasticsearch.cn/book/elasticsearch_definitive_guide_2.x/nested-aggregation.html