Elasticsearch之SearchAPI
Search API
ES 的 Search API 分为两大类，第一类是 URI Search，用 HTTP GET 的方式在 URL 中使用查询参数已达到查询的目的；另一类为 Request Body Search，可以使用 ES 提供的基于 JSON 格式的格式更加完备的查询语言 Query DSL（Domain Specific Language）
| 语法 | 范围 |
|---|---|
| /_search | 集群上所有的索引 |
| /{索引名}/_search | 指定索引名 |
| /a,b/_search | 索引a 和 b |
| /a*/_search | 以 a 开头的索引 |
准备工作
新建索引
PUT student
{
"mappings": {
"properties": {
// 学号
"id":{"type": "long"},
// 学生姓名，多字段类型 支持 text和keyword
"name": {
"type": "text",
"fields": {
"raw": {"type": "keyword"}
}
},
// 年龄
"age":{"type": "integer"},
// 性别,0:男;1:女
"sex":{"type":"integer"},
// 创建时间== 入学时间
"create_deate":{
"type":"date"
},
// 家庭地址
"address":{
// 默认类型是object，可以不显式声明
"type": "object",
"properties": {
// 省份
"province":{"type":"keyword"},
// 城市
"city":{"type":"keyword"},
// 区
"district":{"type":"keyword"},
// 详细地址
"detail_adress":{"type":"text"}
}
},
// 成绩单,嵌套对象
"transcript":{
// 设置类型为嵌套
"type": "nested",
"properties": {
// 科目
"subject": {"type":"keyword"},
// 分数
"score": {"type":"integer"}
}
}
}
}
}
添加数据
PUT _bulk
{"create":{"_index":"student","_id":1}}
{"id":100000001,"name":"张三","age":22,"sex":0,"create_date":"2012-09-15T00:00:00","address":{"province":"浙江","city":"杭州","district":"滨江","detail_adress":"滨康路106号" },"transcript":[{"subject":"语文","score":100},{"subject":"数学","score":59}]}
{"create":{"_index":"student","_id":2}}
{"id":100000002,"name":"李四","age":23,"sex":1,"create_date":"2012-09-15T00:00:00","address":{"province":"浙江","city":"杭州","district":"富阳","detail_adress":"银湖街道106号" },"transcript":[{"subject":"语文","score":60},{"subject":"数学","score":60}]}
{"create":{"_index":"student","_id":3}}
{"id":100000003,"name":"王五","age":21,"sex":1,"create_date":"2012-09-15T00:00:00","address":{"province":"河南","city":"郑州","district":"中原","detail_adress":"中原街道556号" },"transcript":[{"subject":"语文","score":70},{"subject":"数学","score":55}]}
URI Search
GET /student/_search?q=name:张三
URI Search 使用的是 GET 方式，其中 q 指定查询语句，语法为 Query String Syntax，是 KV 键值对的形式；上面的请求表示对 name 字段进行查询，查询包含 张三 的所有文档。
URI Search 有很多参数可以指定，除了 q 还有如下参数：
- df：默认字段，不指定时会对所有字段进行查询
- sort：根据字段名排序
- from：返回的索引匹配结果的开始值，默认为 0
- size：搜索结果返回的条数，默认为 10
- timeout：超时的时间设置
- _source：只返回索引中指定的列，多个列中间用逗号分开
- analyzer：当分析查询字符串的时候使用的分词器
- analyze_wildcard：通配符或者前缀查询是否被分析，默认为 false
- explain：在每个返回结果中，将包含评分机制的解释
- lenient：若设置为 true，字段类型转换失败的时候将被忽略，默认为 false
- default_operator：默认多个条件的关系，AND 或者 OR，默认为 OR
- search_type：搜索的类型，可以为 dfs_query_then_fetch 或 query_then_fetch，默认为 query_then_fetch
实例
只返回姓名、性别字段
GET /student/_search?_source=name,sex
{
"took": 0,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 3,
"relation": "eq"
},
"max_score": 1,
"hits": [
{
"_index": "student",
"_id": "1",
"_score": 1,
"_source": {
"name": "张三",
"sex": 0
}
},
{
"_index": "student",
"_id": "2",
"_score": 1,
"_source": {
"name": "李四",
"sex": 1
}
},
{
"_index": "student",
"_id": "3",
"_score": 1,
"_source": {
"name": "王五",
"sex": 1
}
}
]
}
}
获取性别为女的全部记录
GET /student/_search?_source=name,sex&q=sex:1
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 2,
"relation": "eq"
},
"max_score": 1,
"hits": [
{
"_index": "student",
"_id": "2",
"_score": 1,
"_source": {
"name": "李四",
"sex": 1
}
},
{
"_index": "student",
"_id": "3",
"_score": 1,
"_source": {
"name": "王五",
"sex": 1
}
}
]
}
}
获取性别为女的一条记录
GET /student/_search?_source=name,sex&q=sex:1&size=1
{
"took": 0,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 2,
"relation": "eq"
},
"max_score": 1,
"hits": [
{
"_index": "student",
"_id": "2",
"_score": 1,
"_source": {
"name": "李四",
"sex": 1
}
}
]
}
}
获取性别为女的第二条记录
GET /student/_search?_source=name,sex&q=sex:1&size=1&from=1
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 2,
"relation": "eq"
},
"max_score": 1,
"hits": [
{
"_index": "student",
"_id": "3",
"_score": 1,
"_source": {
"name": "王五",
"sex": 1
}
}
]
}
}
URI Search 好处就是操作简单，只要写个 URI 就可以了，方便测试，但是 URI Search 只包含一部分查询语法，不能覆盖所有 ES 支持的查询语法。
Request Body Search
在 ES 中一些高阶用法只能在 Request Body 里做，所以我们尽量使用 Request Body Search，它支持 GET 和 POST 方式对索引进行查询，需要指定操作的索引名称，同样也要通过 _search 来标明这个请求为搜索请求，我们可以在请求体中使用 ES 提供的 DSL，下面这个例子就是简单的 Query DSL：
GET /student/_search
{
"query": {
"match_all": {}
}
}
实例
只返回姓名、性别字段
GET /student/_search
{
"_source":["name","sex"]
}
获取性别为女的全部记录
GET /student/_search
{
"_source":["name","sex"],
"query": {
"term":{"sex":1}
}
}
获取性别为女的一条记录
GET /student/_search
{
"_source":["name","sex"],
"query": {
"term":{"sex":1}
},
"size":1
}
获取性别为女的第二条记录
GET /student/_search
{
"_source":["name","sex"],
"query": {
"term":{"sex":1}
},
//第几条起 类似 limit 1,1
// from size 组合可以当做分页查询使用，from默认从0开始，size默认为 10
"from":1,
"size":1
}
按年龄倒序排序
GET /student/_search
{
"_source":["name","sex","age"],
"sort": [
{
"age": "desc"
}
]
}
返回结果：
{
"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": "student",
"_id": "2",
"_score": null,
"_source": {
"name": "李四",
"age": 23,
"sex": 1
},
"sort": [
23
]
},
{
"_index": "student",
"_id": "1",
"_score": null,
"_source": {
"name": "张三",
"age": 22,
"sex": 0
},
"sort": [
22
]
},
{
"_index": "student",
"_id": "3",
"_score": null,
"_source": {
"name": "王五",
"age": 21,
"sex": 1
},
"sort": [
21
]
}
]
}
}
返回成绩总分
目前我们文档中是没有成绩总分这个字段，需要通过聚合计算的形式返回
GET /student/_search
{
"size": 0,
"aggs": {
"by_name": {
"terms": {
"field": "name.raw",
"size": 10
},
"aggs": {
"total_score": {
"nested": {
"path": "transcript"
},
"aggs": {
"sum_score": {
"sum": {
"field": "transcript.score"
}
}
}
}
}
}
}
}
返回结果：
{
"took": 0,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 3,
"relation": "eq"
},
"max_score": null,
"hits": []
},
"aggregations": {
"by_name": {
"doc_count_error_upper_bound": 0,
"sum_other_doc_count": 0,
"buckets": [
{
"key": "张三",
"doc_count": 1,
"total_score": {
"doc_count": 2,
"sum_score": {
"value": 159
}
}
},
{
"key": "李四",
"doc_count": 1,
"total_score": {
"doc_count": 2,
"sum_score": {
"value": 120
}
}
},
{
"key": "王五",
"doc_count": 1,
"total_score": {
"doc_count": 2,
"sum_score": {
"value": 125
}
}
}
]
}
}
}
按年龄排序(脚本计算)
王五年龄登记错误，需要+5 ，这个时候需要通过脚本字段的形式排序
GET /student/_search
{
"_source": [
"name",
"sex",
"age"
],
"sort": [
{
"_script": {
// 排序类型 number 或 string
"type": "number",
// 排序方式 desc 或 asc
"order": "desc",
// 脚本定义
"script": {
// 使用的脚本语言，默认值为painless
"lang": "painless",
// 脚本内容
"source": """
if(params['_source']['name'] == '王五'){
params['_source']['age']+params.addValue
} else {
params['_source']['age']
}
""",
// 变量参数
"params": {
"addValue": 5
}
}
}
}
]
}
返回结果：
{
"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": "student",
"_id": "3",
"_score": null,
"_source": {
"name": "王五",
"age": 21,
"sex": 1
},
"sort": [
26
]
},
{
"_index": "student",
"_id": "2",
"_score": null,
"_source": {
"name": "李四",
"age": 23,
"sex": 1
},
"sort": [
23
]
},
{
"_index": "student",
"_id": "1",
"_score": null,
"_source": {
"name": "张三",
"age": 22,
"sex": 0
},
"sort": [
22
]
}
]
}
}
