ES各种查询语法及响应结果

妖狐艹你老母 2024-04-03 11:44 190阅读 0赞

**目录**

查询

创建

删除请求

1. 删除索引

数据类型

一、创建索引

1.1 创建book\_test索引

1.2 查看索引映射

1.3 添加文档记录

1.4 查询文档

1.6 删除文档中某一条数据

二、查询语法

2.1 term&terms查询

2.2 match查询属于高层查询，他会根据你查询的字段类型不一样，采用不同的查询方式，

3.1 分词效果分析

4.1 其他查询

4.1.1 根据单个id查询

4.1.2 \# 查询多个id返回结果

4.2.1 prefix查询

4.3 fuzzy查询 模糊查询，我们输入字符的大概，ES就可以根据输入的内容大概去匹配结果。但是查询结果不太稳定

4.4 wildcard 查询

4.5 range范围查询

4.6 正则查询，通过你编写的正则表达式去匹配内容

4.7 scroll 深分页

4.8 delete-by-query 根据查询删除

4.9 复合查询，

4.9 boosting查询

4.10 filter查询

4.11 高亮查询

4.12 聚合查询，多种多样的聚合

--------------------

## 查询 ##

GET请求 ，参数只能在路径上  
POST请求，也是 查询，参数可以放在请求体中，也可以更新文档，指定type=update

## 创建 ##

PUT请求  
创建索引，需要在请求体中指定分片信息，索引信息  
  
http://ip:port/index/type/\_mapping 创建索引时，指定索引文档的属性

##    删除请求 ##

DELETE  
http://ip:port/index 删除整个文档，==删除库

### 1. 删除索引 ###

DELETE /user\_index

## 数据类型 ##

https://www.elastic.co/guide/en/elasticsearch/reference/current/mapping-types.html  
feild  
String  
text：分词，可以放一段话  
keyword: 不会被分词 ，比如地区--北京，上海  
时间类型  
date :指定格式  
数值类型

布尔类型  
Boolean  
二进制类型  
  
范围类型  
  
经纬度  
"type": "geo\_point"

## 一、创建索引 ##

### 1.1 创建book\_test索引 ###

PUT /book\_test  
\{  
"settings": \{  
"number\_of\_replicas": 1,  
"number\_of\_shards": 1  
\},  
"mappings": \{  
"noval":\{ \# 执行文档类型 在es8以上没有这个了。需要删掉，否则会出错  
"properties":\{ \# 指定文档属性  
"name":\{ \# 指定属性名  
"type":"text",  
"index":true,  
"analyzer": "ik\_smart",  
"store": true  
\},  
"author":\{  
"type":"keyword"  
\},  
"count":\{  
"type":"long"  
\},  
"on\_sale":\{  
"type":"date",  
"format": "yyyy-MM-dd HH:mm:ss||yyyy-MM-dd||epoch\_millis"  
\},  
"desc":\{  
"type":"text",  
"analyzer":"ik\_max\_word"  
\}  
\}  
\}  
\}  
\}

1.1 创建成功响应结果  
\{  
"acknowledged" : true,  
"shards\_acknowledged" : true,  
"index" : "book\_test"  
\}

### 1.2 查看索引映射 ###

GET /book\_test/\_mapping  
1.2 响应结果  
\{  
"book\_test" : \{  
"mappings" : \{  
"properties" : \{  
"author" : \{  
"type" : "keyword"  
\},  
"count" : \{  
"type" : "long"  
\},  
"desc" : \{  
"type" : "text",  
"analyzer" : "ik\_max\_word"  
\},  
"name" : \{  
"type" : "text",  
"store" : true,  
"analyzer" : "ik\_smart"  
\},  
"on\_sale" : \{  
"type" : "date",  
"format" : "yyyy-MM-dd HH:mm:ss||yyyy-MM-dd||epoch\_millis"  
\}  
\}  
\}  
\}  
\}

### 1.3 添加文档记录 ###

POST /book\_test/\_doc/2 \# 这是指定id，没有/2自动生成id。带有id多次操作是修改  
\{  
"name":"西游记",  
"author":"张三2",  
"count":"1565",  
"on\_sale":"2022-08-30"  
,"desc":"了接口直接裂开房价高了几十个"  
\}  
  
响应 带有id去更新操作  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "2",  
"\_version" : 2,  
"result" : "updated",  
"\_shards" : \{  
"total" : 2,  
"successful" : 1,  
"failed" : 0  
\},  
"\_seq\_no" : 2,  
"\_primary\_term" : 1  
\}

###    1.4 查询文档 ###

GET /book\_test/\_search  
  
响应  
  
\{  
"took" : 0, \# 查询耗时  
"timed\_out" : false, \# 是否超时，这里没超时  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 2,  
"relation" : "eq"  
\},  
"max\_score" : 1.0,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "23dxtoMBUPIMnS01cm85",  
"\_score" : 1.0,  
"\_source" : \{  
"name" : "我的小说",  
"author" : "张三",  
"count" : "125",  
"on\_sale" : "2022-09-30",  
"desc" : "家乐鸡粉了垃圾的垃圾费附件覅额计入还让他了客户是"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "2",  
"\_score" : 1.0,  
"\_source" : \{  
"name" : "西游记",  
"author" : "张三2",  
"count" : "1565",  
"on\_sale" : "2022-08-30",  
"desc" : "了接口直接裂开房价高了几十个"  
\}  
\}  
\]  
\}  
\}

### 1.5 修改某个字段的值，根据id查询结果，修改一条 ###

POST /book\_test/\_update\_by\_query  
\{  
"script": \{  
"source": "ctx.\_source.name=params.name",  
"params": \{  
"name":"修改西游记"  
\},  
"lang": "painless"  
\}  
, "query": \{  
"term": \{  
"\_id": 2  
\}  
\}  
\}  
修改某个字段的值，根据字段查询结果，修改多条  
POST /book\_test/\_update\_by\_query  
\{  
"script": \{  
"source": "ctx.\_source.name=params.name",  
"params": \{  
"name":"修改西游记2"  
\},  
"lang": "painless"  
\}  
, "query": \{  
"term": \{  
"author" : "张三2"  
\}  
\}  
\}  
文档方式修改单个字段  
POST /book\_test/\_doc/2/\_update  
\{  
"doc":\{  
"count":"1222"  
\}  
\}

###    1.6 删除文档中某一条数据 ###

删除id=\_update的数据  
DELETE /book\_test/\_doc/\_update  
响应结果  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "\_update",  
"\_version" : 2,  
"result" : "deleted",  
"\_shards" : \{  
"total" : 2,  
"successful" : 1,  
"failed" : 0  
\},  
"\_seq\_no" : 11,  
"\_primary\_term" : 1  
\}

## 二、查询语法 ##

### 2.1 term&terms查询 ###

虽然查询的词不会分词，但是搜索的时候，是用词，到分词库中匹配，匹配到就返回响应的数据，比如：一条数据的长文本里有奋斗的时代，分词后会产生奋斗、时代。我们用奋斗term查询的时候就会匹配到这条数据返回。  
2.1.1 term查询  
term 的查询代表的是完全匹配，搜索之前不会对你搜索的关键字进行分词，对你的关键字去文档分词库中去匹配内容。  
  
查询语法  
POST /book\_test/\_search  
\{  
"from": 0,  
"size": 20,  
"query": \{  
"term": \{  
"author": \{  
"value": "李四"  
\}  
\}  
\}  
\}  
  
响应结果  
\{  
"took" : 402, \# 耗时  
"timed\_out" : false, \# 是否超时  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 1,  
"relation" : "eq"  
\},  
"max\_score" : 1.3862942,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "4",  
"\_score" : 1.3862942,  
"\_source" : \{  
"name" : "西游记2",  
"author" : "李四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "发根深蒂固"  
\}  
\}  
\]  
\}  
\}  
2.1.2 terms 也不会分词，针对某个字段，多个值进行匹配  
查询语法  
POST /book\_test/\_search  
\{  
"from": 0,  
"size": 20,  
"query": \{  
"terms": \{  
"author": \[  
"张三",  
"李四"  
\]  
\}  
\}  
\}  
  
响应结果  
\{  
"took" : 0,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 2,  
"relation" : "eq"  
\},  
"max\_score" : 1.0,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "23dxtoMBUPIMnS01cm85",  
"\_score" : 1.0,  
"\_source" : \{  
"name" : "我的小说",  
"author" : "张三",  
"count" : "125",  
"on\_sale" : "2022-09-30",  
"desc" : "家乐鸡粉了垃圾的垃圾费附件覅额计入还让他了客户是"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "4",  
"\_score" : 1.0,  
"\_source" : \{  
"name" : "西游记2",  
"author" : "李四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "发根深蒂固"  
\}  
\}  
\]  
\}  
\}

### 2.2 match查询属于高层查询，他会根据你查询的字段类型不一样，采用不同的查询方式， ###

查询日期或者是数据类型，他会将你基于字符串查询的内推转换为日期或者数值  
如果查询的内容是一个不能被分词的内容（keyword），match查询不会对你指定的查询关键字进行分词  
  
如果查询的内容是一个可以被分词的内容（text），match会将你指定的查询内容根据一定的方式去分词，去分词库中匹配指定的内容  
  
总之，match查询，实际底层就是多个term查询，将多个term查询的结果给封装到一起，返回。  
  
ES默认只查询10条数据，要想查询更多，需要指定size。  
  
2.2.1 查询语法 分词匹配  
POST /book\_test/\_search  
\{  
"query": \{  
"match": \{  
"name": "西游记"  
\}  
\}  
\}  
  
响应  
\{  
"took" : 0,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 3,  
"relation" : "eq"  
\},  
"max\_score" : 0.48120394,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "4",  
"\_score" : 0.48120394,  
"\_source" : \{  
"name" : "西游记2",  
"author" : "李四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "发根深蒂固"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "3",  
"\_score" : 0.40841687,  
"\_source" : \{  
"author" : "张三2",  
"name" : "修改西游记2",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "fsgdfgafad"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "2",  
"\_score" : 0.40841687,  
"\_source" : \{  
"author" : "张三2",  
"name" : "修改西游记2",  
"count" : "1222",  
"on\_sale" : "2022-08-30",  
"desc" : "了接口直接裂开房价高了几十个"  
\}  
\}  
\]  
\}  
\}  
  
2.2.2 布尔match查询  
查询一个字段，可以匹配多个字，关系可以是or 或者 and  
POST /book\_test/\_search  
\{  
"query": \{  
"match": \{  
"desc": \{  
"query": "京东 大家",  
"operator": "or"  
\}  
\}  
\}  
\}  
  
响应结果，评分越高，匹配度越高  
\{  
"took" : 275,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 2,  
"relation" : "eq"  
\},  
"max\_score" : 3.179408,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "5",  
"\_score" : 3.179408,  
"\_source" : \{  
"name" : "大话西游",  
"author" : "李大四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "京东卡拉拉LDA； 大家都发"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "6",  
"\_score" : 1.5683775,  
"\_source" : \{  
"name" : "大话济公",  
"author" : "小四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "大家都发财"  
\}  
\}  
\]  
\}  
\}  
  
2.3 multi\_match查询  
match针对一个feild做检索，multi\_match针对多个feild进行检索，多个feild对应一个text。

### 3.1 分词效果分析 ###

POST \_analyze  
\{  
"analyzer": "ik\_max\_word",  
"text": \["笔记本电脑"\]  
\}

### 4.1 其他查询 ###

#### 4.1.1 根据单个id查询 ####

GET /book\_test/\_doc/2  
GET /book\_test/\_search?q=\_id:3.

#### 4.1.2 \# 查询多个id返回结果 ####

GET /user\_index/\_doc/\_mget  
\{  
"ids":\[1,2\]  
\}  
  
POST /book\_test/\_search  
\{  
"query": \{  
"ids": \{  
"values": \["3","2"\]  
\}  
\}  
\}

#### 4.2.1 prefix查询 ####

4.2.1.1 前缀查询，下面author类型是keyword，如果用match或者term匹配不到，用前缀查询就可以匹配到  
POST /book\_test/\_search  
\{  
"query": \{  
"prefix": \{  
"author": \{  
"value": "李"  
\}  
\}  
\}  
\}  
  
前缀查询响应  
\{  
"took" : 0,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 2,  
"relation" : "eq"  
\},  
"max\_score" : 1.0,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "4",  
"\_score" : 1.0,  
"\_source" : \{  
"name" : "西游记2",  
"author" : "李四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "发根深蒂固"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "5",  
"\_score" : 1.0,  
"\_source" : \{  
"name" : "大话西游",  
"author" : "李大四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "京东卡拉拉LDA； 大家都发"  
\}  
\}  
\]  
\}  
\}  
  
4.2.1.2 前缀匹配text类型的字段查询，匹配的是分词库--比如name  
POST /book\_test/\_search  
\{  
"query": \{  
"prefix": \{  
"name": \{  
"value": "西游"  
\}  
\}  
\}  
\}  
  
分词前缀响应  
\{  
"took" : 0,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 3,  
"relation" : "eq"  
\},  
"max\_score" : 1.0,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "3",  
"\_score" : 1.0,  
"\_source" : \{  
"author" : "张三2",  
"name" : "修改西游记2",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "fsgdfgafad"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "2",  
"\_score" : 1.0,  
"\_source" : \{  
"author" : "张三2",  
"name" : "修改西游记2",  
"count" : "1222",  
"on\_sale" : "2022-08-30",  
"desc" : "了接口直接裂开房价高了几十个"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "4",  
"\_score" : 1.0,  
"\_source" : \{  
"name" : "西游记2",  
"author" : "李四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "发根深蒂固"  
\}  
\}  
\]  
\}  
\}

#### 4.3 fuzzy查询 模糊查询，我们输入字符的大概，ES就可以根据输入的内容大概去匹配结果。但是查询结果不太稳定 ####

4.3.1 模糊查询也是---去分词库查  
POST /book\_test/\_search  
\{  
"query": \{  
"fuzzy": \{ \# 模糊查询  
"desc": \{ \# 模糊字段  
"value": "大家",  
"prefix\_length": 1 \# 前边几个字不可以错的  
\}  
\}  
\}  
\}  
  
模糊响应  
\{  
"took" : 0,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 2,  
"relation" : "eq"  
\},  
"max\_score" : 1.691406,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "6",  
"\_score" : 1.691406,  
"\_source" : \{  
"name" : "大话济公",  
"author" : "小四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "大家都发财"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "5",  
"\_score" : 1.386699,  
"\_source" : \{  
"name" : "大话西游",  
"author" : "李大四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "京东卡拉拉LDA； 大家都发"  
\}  
\}  
\]  
\}  
\}

### 4.4 wildcard 查询 ###

通配符查询，和mysql中的like是一样，可以在查询时，在字符串中指定通配符\*和占位符？  
POST /book\_test/\_search  
\{  
"query": \{  
"wildcard": \{  
"desc": \{  
"value": "\*f\*"  
\}  
\}  
\}  
\}  
  
响应  
\{  
"took" : 0,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 1,  
"relation" : "eq"  
\},  
"max\_score" : 1.0,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "3",  
"\_score" : 1.0,  
"\_source" : \{  
"author" : "张三2",  
"name" : "修改西游记2",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "fsgdfgafad"  
\}  
\}  
\]  
\}  
\}  
  
第二个测试  
POST /book\_test/\_search  
\{  
"query": \{  
"wildcard": \{  
"author": \{  
"value": "李??"  
\}  
\}  
\}  
\}

### 4.5 range范围查询 ###

范围查询，只针对数值类型，对某一个feild进行大于或者小于的范围指定  
大于gt； 小于lt  
  
POST /book\_test/\_search  
\{  
"query": \{  
"range": \{  
"count": \{  
"gte": 10,  
"lte": 200  
\}  
\}  
\}  
\}

### 4.6 正则查询，通过你编写的正则表达式去匹配内容 ###

ps：prefix，fuzzy，wildcard，regexp查询效率相对比较低，要求效率比较高时，避免去使用。  
POST /book\_test/\_search  
\{  
"query": \{  
"regexp": \{  
"name": "\*\[0-9\]\{1\}"  
\}  
\}  
\}

### 4.7 scroll 深分页 ###

ES对from+size是有限制的，from和size二者纸盒不能超过1万  
原理：  
from+size 的ES查询数据的方式  
1、 先将用户指定的关键字进行分词  
2、 将词汇去分词库中进行检索，得到多个文档的id  
3、 去各个分片中去拉取指定的数据 \# 这个过程有点慢的  
4、 将数据分局score进行排序 \# 这个慢  
5、 根据from的值，将查询到的数据舍弃一部分  
6、 返回结果  
  
scroll 的ES查询数据的方式  
1、 先将用户指定的关键字进行分词  
2、 将词汇去分词库中进行检索，得到多个文档的id  
3、 将文档的id存放在一个ES的上下文中  
4、 根据你指定的size去ES中检索指定的数据，拿完数据的文档id，会从上下文移除。  
5、 如果需要下一页的数据，直接去ES的上下文中，找后续内容。  
6、 返回结果  
ps：不适合做实时查询，因为是从内存中查的。  
查询结果没有评分了，按照id排序或者是自己定义排序  
  
POST /book\_test/\_search?scroll=1m \# 指定scroll查询的id在内存中存1分钟  
\{  
  
"query": \{  
"match": \{  
"name": "西游记"  
\}  
\}  
, "size": 2 \# 查询条数  
, "sort": \[ \# 自定义排序  
\{  
"count": \{  
"order": "desc"  
\}  
\}  
\]  
\}  
  
深分页响应  
\{  
"\_scroll\_id" : "FGluY2x1ZGVfY29udGV4dF91dWlkDXF1ZXJ5QW5kRmV0Y2gBFlZ4ZGplLVZ4VGFlcVdWT1ljQmstTHcAAAAAAIMYSRZDMWpuSWFZa1JFQ2ZReGRRRHdzcm53",  
"took" : 0,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 3,  
"relation" : "eq"  
\},  
"max\_score" : null,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "3",  
"\_score" : null,  
"\_source" : \{  
"author" : "张三2",  
"name" : "修改西游记2",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "fsgdfgafad"  
\},  
"sort" : \[  
1565  
\]  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "4",  
"\_score" : null,  
"\_source" : \{  
"name" : "西游记2",  
"author" : "李四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "发根深蒂固"  
\},  
"sort" : \[  
1565  
\]  
\}  
\]  
\}  
\}  
下一次scroll查询  
POST /\_search/scroll  
\{  
"scroll\_id":"FGluY2x1ZGVfY29udGV4dF91dWlkDXF1ZXJ5QW5kRmV0Y2gBFlZ4ZGplLVZ4VGFlcVdWT1ljQmstTHcAAAAAAIMZghZDMWpuSWFZa1JFQ2ZReGRRRHdzcm53"  
,"scroll":"1m"  
\}  
  
删除scroll  
DELETE /\_search/scroll/FGluY2x1ZGVfY29udGV4dF91dWlkDXF1ZXJ5QW5kRmV0Y2gBFlZ4ZGplLVZ4VGFlcVdWT1ljQmstTHcAAAAAAIMZghZDMWpuSWFZa1JFQ2ZReGRRRHdzcm53

### 4.8 delete-by-query 根据查询删除 ###

如果删除文档数量是大部分文档，不建议用这种方式，这种方式，比较耗费性能，一条条删除的  
  
POST /book\_test/\_delete\_by\_query  
\{  
"query": \{  
"match": \{  
"name": "西游记"  
\}  
\}  
\}

### 4.9 复合查询， ###

复合过滤器，将你的多个查询条件，以一定的逻辑组合在一起  
must ：所有的条件，用must组合在一起，表示and意思  
must\_not : 将must\_not中的条件，全部都不能匹配，表示not的意思  
should ： 所有的条件，用should组合在一起，表示or的意思  
  
POST /book\_test/\_search  
\{  
"query": \{  
"bool": \{  
"must": \[  
\{  
"match": \{  
"name": "西游记"  
\}  
\},  
\{  
"range": \{  
"on\_sale": \{  
"gte": "2022-08-31",  
"lte": "2022-08-31"  
\}  
\}  
\}  
\],  
"should": \[  
\{  
"term": \{  
"author": \{  
"value": "李四"  
\}  
\}  
\}  
\],  
"must\_not": \[  
\{  
"range": \{  
"count": \{  
"gte": 10,  
"lte": 20  
\}  
\}  
\}  
\]  
\}  
\}  
\}  
  
响应结果  
\{  
"took" : 0,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 2,  
"relation" : "eq"  
\},  
"max\_score" : 3.5179136,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "4",  
"\_score" : 3.5179136,  
"\_source" : \{  
"name" : "西游记2",  
"author" : "李四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "发根深蒂固"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "3",  
"\_score" : 1.6099696,  
"\_source" : \{  
"author" : "张三2",  
"name" : "修改西游记2",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "fsgdfgafad"  
\}  
\}  
\]  
\}  
\}

#### 4.9 boosting查询 ####

boosting查询可以帮助我们去影响查询后的score，  
positive：只有匹配上positive的查询的内容，才会被放到返回的结果中  
negative ： 如果匹配上和positive并且也匹配上了negative，就可以降低这样的文档score。  
negative\_boost ： 指定系数，必须小于1.0  
  
关键查询时，分数是如何计算的：  
搜索的关键字在文档中出现的频次越高，分数就越高  
指定的文档内容越短，分数就越高  
我们在搜索时，指定的关键字也会分词，这个被分词的内容，被分词的个数越多，分数越高  
  
POST /book\_test/\_search  
\{  
"query": \{  
"boosting": \{  
"positive": \{  
"match": \{  
"name": "西游记"  
\}  
\},  
"negative": \{  
"match": \{  
"author": "李四"  
\}  
\},  
"negative\_boost": 0.5  
\}  
\}  
\}  
  
响应  
\{  
"took" : 6,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 3,  
"relation" : "eq"  
\},  
"max\_score" : 0.60996956,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "3",  
"\_score" : 0.60996956,  
"\_source" : \{  
"author" : "张三2",  
"name" : "修改西游记2",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "fsgdfgafad"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "2",  
"\_score" : 0.60996956,  
"\_source" : \{  
"author" : "张三2",  
"name" : "修改西游记2",  
"count" : "1222",  
"on\_sale" : "2022-08-30",  
"desc" : "了接口直接裂开房价高了几十个"  
\}  
\},  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "4",  
"\_score" : 0.36307707,  
"\_source" : \{  
"name" : "西游记2",  
"author" : "李四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "发根深蒂固"  
\}  
\}  
\]  
\}  
\}

### 4.10 filter查询 ###

query查询，根据你的查询条件，去计算文档的匹配度，得到一个分数，并且根据分数进行排序，不会做缓存的。  
filter ，根据你的查询条件去查询文档，不去计算分数，不做排序，而且filter会对经常被过滤的数据进行缓存，查询效率相对query高  
POST /book\_test/\_search  
\{  
"query": \{  
"bool": \{  
"filter": \[  
\{  
"term": \{ \# 第一个过滤条件  
"author": "李四"  
\}  
\},  
\{  
"range": \{ \# 第二个过滤条件  
"count": \{  
"gte": 10,  
"lte": 20000  
\}  
\}  
\}  
\]  
\}  
\}  
\}  
响应结果没有分数  
\{  
"took" : 0,  
"timed\_out" : false,  
"\_shards" : \{  
"total" : 1,  
"successful" : 1,  
"skipped" : 0,  
"failed" : 0  
\},  
"hits" : \{  
"total" : \{  
"value" : 1,  
"relation" : "eq"  
\},  
"max\_score" : 0.0,  
"hits" : \[  
\{  
"\_index" : "book\_test",  
"\_type" : "\_doc",  
"\_id" : "4",  
"\_score" : 0.0,  
"\_source" : \{  
"name" : "西游记2",  
"author" : "李四",  
"count" : "1565",  
"on\_sale" : "2022-08-31",  
"desc" : "发根深蒂固"  
\}  
\}  
\]  
\}  
\}

### 4.11 高亮查询 ###

高亮查询就是你用户输入的关键字，以一定的特殊样式展示给用户，让用户知道为什么这个结果被检索出来。 
高亮展示的数据，本身就是稳定中的一个feild，单独将feild以highlight的形式返回给你。 
ES提供了一个highlight属性和query同级别的。 
fragment\_size：指定高亮数据展示多少个字符回来。 
pre\_tags：指定前缀标签 
post\_tags : 指定后缀标签 
fields：指定哪个feild以高亮形式返回。 
 
POST /book\_test/\_search 
\{ 
"query": \{ 
"bool": \{ 
"filter": \[ 
\{ 
"term": \{ 
"desc": "大家" 
\} 
\}, 
\{ 
"range": \{ 
"count": \{ 
"gte": 10, 
"lte": 20000 
\} 
\} 
\} 
\] 
\} 
\}, 
"highlight": \{ 
"fields": \{ 
"desc": \{\} 
\}, 
"pre\_tags": "", 
"post\_tags": "" 
, "fragment\_size":10 
\} 
\} 
 
高亮响应 
\{ 
"took" : 86, 
"timed\_out" : false, 
"\_shards" : \{ 
"total" : 1, 
"successful" : 1, 
"skipped" : 0, 
"failed" : 0 
\}, 
"hits" : \{ 
"total" : \{ 
"value" : 2, 
"relation" : "eq" 
\}, 
"max\_score" : 0.0, 
"hits" : \[ 
\{ 
"\_index" : "book\_test", 
"\_type" : "\_doc", 
"\_id" : "5", 
"\_score" : 0.0, 
"\_source" : \{ 
"name" : "大话西游", 
"author" : "李大四", 
"count" : "1565", 
"on\_sale" : "2022-08-31", 
"desc" : "京东卡拉拉LDA； 大家都发" 
\}, 
"highlight" : \{ 
"desc" : \[ 
"京东卡拉拉LDA； 大家都发" 
\] 
\} 
\}, 
\{ 
"\_index" : "book\_test", 
"\_type" : "\_doc", 
"\_id" : "6", 
"\_score" : 0.0, 
"\_source" : \{ 
"name" : "大话济公", 
"author" : "小四", 
"count" : "1565", 
"on\_sale" : "2022-08-31", 
"desc" : "大家都发财" 
\}, 
"highlight" : \{ 
"desc" : \[ 
"大家都发财" 
\] 
\} 
\} 
\] 
\} 
\}

### 4.12 聚合查询，多种多样的聚合 ###

ES的聚合查询和MySQL的聚合查询类似  
  
聚合查询的语法  
POST /book\_test/\_search  
\{  
"aggs":\{  
"名字(count\_agg)":\{  
"聚合类型：cardinality , avg agg\_type":\{  
"属性":"值"  
\}  
\}  
\}  
\}  
  
4.12.1 去重计数查询  
去重计数，即cardinality，第一步先将返回的文档中的一个指定的feild进行去重，统计一共多少条。 去重的字段不能是text。会报错，可以是keyword、long等  
POST /mall\_index\_spu/\_search  
\{  
"aggs": \{  
"aggs": \{  
"cardinality": \{  
"field": "teamId"  
\}  
\}  
\}  
\}  
  
响应最后  
\{  
"aggregations" : \{  
"aggs" : \{  
"value" : 4  
\}  
\}  
4.12.2 范围统计  
统计一定范围内出现的文档，比如，针对某一个feild的值在0-100,100-200之间  
范围统计可以针对普通的数值，针对时间类型，针对ip类型都可以做相应的统计  
range，date\_range,ip\_range  
  
POST /mall\_index\_spu/\_search  
\{  
"aggs": \{  
"range\_count": \{  
"range": \{  
"field": "saleNum",  
"ranges": \[  
\{  
"to": 10  
\},  
\{  
"from": 10, \# from 是包含的  
"to": 50  
\},  
\{  
"from": 50  
\}  
\]  
\}  
\}  
\}  
\}  
  
4.12.3 时间范围统计  
POST /book\_test/\_search  
\{  
"aggs": \{  
"range\_count": \{  
"date\_range": \{  
"field": "on\_sale",  
"format": "yyyy-MM",  
"ranges": \[  
\{  
"from": "2021-10",  
"to": "2023-10"  
\}  
\]  
\}  
\}  
\}  
\}  
  
时间范围统计结果  
\{  
"aggregations" : \{  
"range\_count" : \{  
"buckets" : \[  
\{  
"key" : "2021-10-2023-10",  
"from" : 1.6330464E12,  
"from\_as\_string" : "2021-10",  
"to" : 1.6961184E12,  
"to\_as\_string" : "2023-10",  
"doc\_count" : 7  
\}  
\]  
\}  
\}  
\}  
4.12.4 extended\_stats统计聚合查询  
他可以帮你查询指定feild的最大值，最小值，平均值，平方和  
POST /book\_test/\_search  
\{  
"aggs": \{  
"range\_count": \{  
"extended\_stats": \{  
"field": "count"  
\}  
\}  
\}  
\}  
  
响应结果  
"aggregations" : \{  
"range\_count" : \{  
"count" : 7,  
"min" : 125.0,  
"max" : 1565.0,  
"avg" : 1310.2857142857142,  
"sum" : 9172.0,  
"sum\_of\_squares" : 1.3755034E7,  
"variance" : 248156.20408163272,  
"variance\_population" : 248156.20408163272,  
"variance\_sampling" : 289515.57142857154,  
"std\_deviation" : 498.15279190388236,  
"std\_deviation\_population" : 498.15279190388236,  
"std\_deviation\_sampling" : 538.0665120861654,  
"std\_deviation\_bounds" : \{  
"upper" : 2306.591298093479,  
"lower" : 313.9801304779495,  
"upper\_population" : 2306.591298093479,  
"lower\_population" : 313.9801304779495,  
"upper\_sampling" : 2386.418738458045,  
"lower\_sampling" : 234.15269011338341  
\}  
\}  
\}  
  
  
4.12.5 ES 的地图检索方式 经纬度  
geo\_distance: 直线距离检索方式  
geo\_bounding\_box: 以两个点确定一个矩形，获取在矩形内的全部数据  
geo\_polygon: 以多个点，确定一个多边形，获取多边形内的全部数据