08、ElasticSearch 实战：简单的搜索
关于版本
| 内容 | 版本 |
|---|---|
| Elasticsearch版本 | 7.2.0 |
| JAVA依赖版本 | 7.2.1 |
Elasticsearch 7.x 和之前版本有相当大的变化，所以本篇内容尤其是JAVA代码的操作对于使用旧版本的同学帮助可能不大。因为本人主要是JAVA开发，在介绍相关操作的时候会附带JAVA代码操作的逻辑。
ES的简单搜索
精确查找和短语匹配
- 精确查找(（)term词条查找)：词条查询不会分析查询条件，只有当词条和查询字符串串完全匹配时，才匹配搜索。
- 短语匹配(match词条查找)：ElasticSearch引擎会先分析查询字符串，将其拆分成多个分词，只要已分析的字段中包含词条的任意一个，或全部包含，就匹配查询条件，返回该文档；如果不包含任意一个分词，表示没有任何文档匹配查询条件。
模拟数据
创建一个新索引
PUT "localhost:9200/city_info"
创建新的映射
url: PUT "localhost:9200/test_city_info/_mapping"
head: Content-Type:application/json
请求参数
{
"properties": {
"name": {
"type": "keyword"
},
"desc": {
"type": "text"
},
"province": {
"type": "keyword"
},
"gdp": {
"type": "long"
},
"area": {
"type": "keyword"
},
"carNumPrefix": {
"type": "keyword"
}
}
}
插入数据
PUT localhost:9200/city_info/_doc/1
请求参数
{
"name": "上海",
"desc": "中国经济、金融、贸易、航运、科技创新中心",
"province": "上海",
"gdp": "3267900000000",
"area": "华东地区",
"carNumPrefix": "沪"
}
{
"name": "北京",
"desc": "中华人民共和国首都",
"province": "北京",
"gdp": "3032000000000",
"area": "华北地区",
"carNumPrefix": "京"
}
// ...
// 数据有点多就不写了，大概就是GDP前十几的城市
term词条查询
单条term
我们现在尝试查出城市名字为“北京”的地区，只需要按照下面的请求
http请求
POST localhost:9200/city_info/_search
请求参数
{
"query": {
"term": {
"name": "北京"
}
}
}
响应内容
可以看到最后返回了北京的相关信息
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 1,
"relation": "eq"
},
"max_score": 2.302585,
"hits": [
{
"_index": "city_info",
"_type": "_doc",
"_id": "2",
"_score": 2.302585,
"_source": {
"name": "北京",
"desc": "中华人民共和国首都",
"province": "北京",
"gdp": "3032000000000",
"area": "华北地区",
"carNumPrefix": "京"
}
}
]
}
}
JAVA代码
public static void term() throws IOException {
SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
sourceBuilder.query(QueryBuilders.termQuery("name", "北京"));
SearchRequest request = new SearchRequest(INDEX);
request.source(sourceBuilder);
SearchResponse searchResponse = RestClientUtils.client.search(request, RequestOptions.DEFAULT);
if (searchResponse.getShardFailures().length == 0) {
System.out.println(searchResponse.getHits().getHits()[0]);
}
}
多条term
但是有的时候我们可能尝试查询不止一个数据的时候呢，可以使用terms的api。下面我们尝试查询名字为北京和上海的城市
http请求
POST localhost:9200/city_info/_search
请求参数
{
"query": {
"terms": {
"name": [
"北京",
"上海"
]
}
}
}
响应内容
{
"took": 14,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 2,
"relation": "eq"
},
"max_score": 1.0,
"hits": [
{
"_index": "city_info",
"_type": "_doc",
"_id": "1",
"_score": 1.0,
"_source": {
"name": "上海",
"desc": "中国经济、金融、贸易、航运、科技创新中心",
"province": "上海",
"gdp": "3267900000000",
"area": "华东地区",
"carNumPrefix": "沪"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "2",
"_score": 1.0,
"_source": {
"name": "北京",
"desc": "中华人民共和国首都",
"province": "北京",
"gdp": "3032000000000",
"area": "华北地区",
"carNumPrefix": "京"
}
}
]
}
}
JAVA代码
public static void terms() throws IOException {
SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
sourceBuilder.query(QueryBuilders.termsQuery("name", "北京","上海"));
SearchRequest request = new SearchRequest(INDEX);
request.source(sourceBuilder);
SearchResponse searchResponse = RestClientUtils.client.search(request, RequestOptions.DEFAULT);
if (searchResponse.getShardFailures().length == 0) {
SearchHit[] hits =
searchResponse.getHits().getHits();
for (int i = 0; i < hits.length; i++) {
System.out.println(searchResponse.getHits().getHits()[i]);
}
}
}
match_all
当我们需要查询某个索引所有数据的时候可以使用match_all,此API实现了查询所有的方法。当然当数据非常大的时候最好一次不要查询太多条数据。下面例子中只查询了3条
http请求
POST localhost:9200/city_info/_search
请求参数
{
"query": {
"match_all": {
}
},
"from": 0,
"size": 3
}
响应内容
{
"took": 1,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 14,
"relation": "eq"
},
"max_score": 1.0,
"hits": [
{
"_index": "city_info",
"_type": "_doc",
"_id": "1",
"_score": 1.0,
"_source": {
"name": "上海",
"desc": "中国经济、金融、贸易、航运、科技创新中心",
"province": "上海",
"gdp": "3267900000000",
"area": "华东地区",
"carNumPrefix": "沪"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "2",
"_score": 1.0,
"_source": {
"name": "北京",
"desc": "中华人民共和国首都",
"province": "北京",
"gdp": "3032000000000",
"area": "华北地区",
"carNumPrefix": "京"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "3",
"_score": 1.0,
"_source": {
"name": "深圳",
"desc": "中国经济特区、全国性经济中心城市和国际化城市",
"province": "广东",
"gdp": "2469100000000",
"area": "华南地区",
"carNumPrefix": "粤B"
}
}
]
}
}
JAVA代码
public static void matchAll() throws IOException {
SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
sourceBuilder.query(QueryBuilders.matchAllQuery()).from(0).size(3);
SearchRequest request = new SearchRequest(INDEX);
request.source(sourceBuilder);
SearchResponse searchResponse = RestClientUtils.client.search(request, RequestOptions.DEFAULT);
if (searchResponse.getShardFailures().length == 0) {
SearchHit[] hits =
searchResponse.getHits().getHits();
System.out.println(hits.length);
}
}
match
当我们对keyword的字段进行查询匹配的时候我们可以使用term。但是对text类型的字段进行查询操作的时候就需要使用match，他会查询所有匹配上条件的结果。并不要求其完全匹配。下面的例子我们查询了描述为城市的数据，结果可以看出来返回的数据中desc字段都包含城市内容，但是内容都远远多于查询的条件。可以认为term是精确匹配而match是模糊匹配
http请求
POST localhost:9200/city_info/_search
请求参数
{
"query": {
"match": {
"desc" : "城市"
}
},
"from": 0,
"size": 3
}
响应内容
{
"took": 2,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 12,
"relation": "eq"
},
"max_score": 0.47477022,
"hits": [
{
"_index": "city_info",
"_type": "_doc",
"_id": "7",
"_score": 0.47477022,
"_source": {
"name": "苏州",
"desc": "大城市",
"province": "江苏",
"gdp": "1859700000000",
"area": "华东地区",
"carNumPrefix": "苏E"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "8",
"_score": 0.47477022,
"_source": {
"name": "成都",
"desc": "大城市",
"province": "四川",
"gdp": "1534200000000",
"area": "西南地区",
"carNumPrefix": "川A"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "9",
"_score": 0.47477022,
"_source": {
"name": "武汉",
"desc": "大城市",
"province": "湖北",
"gdp": "1484700000000",
"area": "华中地区",
"carNumPrefix": "鄂A"
}
}
]
}
}
JAVA代码
public static void match() throws IOException {
SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
sourceBuilder.query(QueryBuilders.matchQuery("desc","城市")).from(0).size(3);
SearchRequest request = new SearchRequest(INDEX);
request.source(sourceBuilder);
SearchResponse searchResponse = RestClientUtils.client.search(request, RequestOptions.DEFAULT);
if (searchResponse.getShardFailures().length == 0) {
SearchHit[] hits =
searchResponse.getHits().getHits();
System.out.println(hits.length);
}
}
multi_match
multi_match和match相比，它提供了多种字段匹配的能力，你可以设置在多个字段中存在匹配的内容。比如例子中尝试查询描述和省份都携带广东的数据。（正常desc字段是不携带广东内容了，下面数据为了显示API的作用，专门修改了一个城市的描述）
http请求
POST localhost:9200/city_info/_search
请求参数
{
"query": {
"multi_match": {
"query": "广东",
"fields": [
"province",
"desc"
]
}
}
}
响应内容
{
"took": 991,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 3,
"relation": "eq"
},
"max_score": 4.163451,
"hits": [
{
"_index": "city_info",
"_type": "_doc",
"_id": "14",
"_score": 4.163451,
"_source": {
"name": "无锡",
"desc": "广东，大城市",
"province": "江苏",
"gdp": "1143800000000",
"area": "华南地区",
"carNumPrefix": "苏B"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "4",
"_score": 2.1411116,
"_source": {
"name": "广州",
"desc": "广东省省会、副省级市、国家中心城市、超大城市",
"province": "广东",
"gdp": "2300000000000",
"area": "华南地区",
"carNumPrefix": "粤A"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "3",
"_score": 1.856298,
"_source": {
"name": "深圳",
"desc": "中国经济特区、全国性经济中心城市和国际化城市",
"province": "广东",
"gdp": "2469100000000",
"area": "华南地区",
"carNumPrefix": "粤B"
}
}
]
}
}
JAVA代码
public static void match() throws IOException {
SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
sourceBuilder.query(QueryBuilders.multiMatchQuery("广东","province","desc"));
SearchRequest request = new SearchRequest(INDEX);
request.source(sourceBuilder);
SearchResponse searchResponse = RestClientUtils.client.search(request, RequestOptions.DEFAULT);
if (searchResponse.getShardFailures().length == 0) {
SearchHit[] hits =
searchResponse.getHits().getHits();
System.out.println(hits.length);
}
}
match_phrase
match_phrase在其查询配置中可以添加slop参数，使用此参数可以限制输入的内容被分词后，短语中间还能间隔的词语的数量。
http请求
POST localhost:9200/city_info/_search
请求参数
{
"query": {
"match_phrase": {
"desc": "中心城市",
"slop" : 0
}
}
}
响应内容
{
"took": 2,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 2,
"relation": "eq"
},
"max_score": 1.8040439,
"hits": [
{
"_index": "city_info",
"_type": "_doc",
"_id": "4",
"_score": 1.8040439,
"_source": {
"name": "广州",
"desc": "广东省省会、副省级市、国家中心城市、超大城市",
"province": "广东",
"gdp": "2300000000000",
"area": "华南地区",
"carNumPrefix": "粤A"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "3",
"_score": 1.6869828,
"_source": {
"name": "深圳",
"desc": "中国经济特区、全国性经济中心城市和国际化城市",
"province": "广东",
"gdp": "2469100000000",
"area": "华南地区",
"carNumPrefix": "粤B"
}
}
]
}
}
JAVA代码
public static void matchPhrase() throws IOException {
SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
sourceBuilder.query(QueryBuilders.matchPhraseQuery("desc","中心城市"));
SearchRequest request = new SearchRequest(INDEX);
request.source(sourceBuilder);
SearchResponse searchResponse = RestClientUtils.client.search(request, RequestOptions.DEFAULT);
if (searchResponse.getShardFailures().length == 0) {
SearchHit[] hits =
searchResponse.getHits().getHits();
System.out.println(hits.length);
}
}
match_phrase_prefix
match_phrase_prefix的用法有点类似搜索推荐中的内容补全（ES实现推荐搜索是用另外一个API），但是又不完全一样，他会将最后一个被切分的词条（trem）作为前缀去匹配索引，然后再从匹配的结果中定位包含前面词条的数据。比如我们尝试搜索大城市 北，es会尝试先找出北开头的doc然后再去定位包含北和大城市的doc。
http请求
POST localhost:9200/city_info/_search
请求参数
{
"query": {
"match_phrase_prefix": {
"desc": "大城"
}
}
}
响应内容
{
"took": 2,
"timed_out": false,
"_shards": {
"total": 1,
"successful": 1,
"skipped": 0,
"failed": 0
},
"hits": {
"total": {
"value": 11,
"relation": "eq"
},
"max_score": 0.5402701,
"hits": [
{
"_index": "city_info",
"_type": "_doc",
"_id": "7",
"_score": 0.5402701,
"_source": {
"name": "苏州",
"desc": "大城市",
"province": "江苏",
"gdp": "1859700000000",
"area": "华东地区",
"carNumPrefix": "苏E"
}
},
.....
{
"_index": "city_info",
"_type": "_doc",
"_id": "5",
"_score": 0.50118303,
"_source": {
"name": "重庆",
"desc": "超大城市",
"province": "重庆",
"gdp": "2036300000000",
"area": "西南地区",
"carNumPrefix": "渝"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "6",
"_score": 0.50118303,
"_source": {
"name": "天津",
"desc": "超大城市",
"province": "天津",
"gdp": "1880900000000",
"area": "华北地区",
"carNumPrefix": "津"
}
},
{
"_index": "city_info",
"_type": "_doc",
"_id": "14",
"_score": 0.46737003,
"_source": {
"name": "无锡",
"desc": "广东，大城市",
"province": "江苏",
"gdp": "1143800000000",
"area": "华南地区",
"carNumPrefix": "苏B"
}
}
]
}
}
JAVA代码
public static void matchPhrasePrefix() throws IOException {
SearchSourceBuilder sourceBuilder = new SearchSourceBuilder();
sourceBuilder.query(QueryBuilders.matchPhrasePrefixQuery("desc","大城"));
SearchRequest request = new SearchRequest(INDEX);
request.source(sourceBuilder);
SearchResponse searchResponse = RestClientUtils.client.search(request, RequestOptions.DEFAULT);
if (searchResponse.getShardFailures().length == 0) {
SearchHit[] hits =
searchResponse.getHits().getHits();
System.out.println(hits.length);
}
}