背景：附近功能在很多生活类的App或软件中经常出现？那他们是怎么实现的呢？如果数据量不是很大，且功能比较简单，基于MySQL就可以实现。然而很多时候数据量很大且功能复杂，那么我们就需要使用Elasticsearch这种数据库了，不仅功能丰富，而且性能强大，大数据量情况下性能不再是问题。
一、基于MySQL实现（8.0.28）
1、建表
DROP TABLE IF EXISTS `t_city`;CREATE TABLE `t_city` (`id` bigint(20) unsigned NOT NULL AUTO_INCREMENT COMMENT 'id',`city_code` varchar(20) NOT NULL DEFAULT '' COMMENT '城市编码',`city_name` varchar(20) NOT NULL DEFAULT '' COMMENT '城市名称',`addr` varchar(255) NOT NULL DEFAULT '' COMMENT '详细地址',`lat` decimal(12,8) DEFAULT NULL COMMENT '纬度',`lon` decimal(12,7) DEFAULT NULL COMMENT '经度',`efficiency_level` varchar(2) DEFAULT '' COMMENT '效率评级',`credit_level` varchar(2) DEFAULT NULL COMMENT '信用评级',PRIMARY KEY (`id`)) ENGINE=InnoDB AUTO_INCREMENT=937 DEFAULT CHARSET=utf8mb4;
2、插入测试数据
INSERT INTO `hb-test`.`t_city` (`city_code`, `city_name`, `addr`, `lat`, `lon`, `efficiency_level`, `credit_level`) VALUES ('1', '北京', '北海公园', null, null, 'A', '1');INSERT INTO `hb-test`.`t_city` (`city_code`, `city_name`, `addr`, `lat`, `lon`, `efficiency_level`, `credit_level`) VALUES ('2', '西安', '西安火车站', null, null, 'A', '2');
INSERT INTO `hb-test`.`t_city` (`city_code`, `city_name`, `addr`, `lat`, `lon`, `efficiency_level`, `credit_level`) VALUES ('1', '北京', '北海公园', '39.93463300', '116.3927710', 'A', '1');INSERT INTO `hb-test`.`t_city` (`city_code`, `city_name`, `addr`, `lat`, `lon`, `efficiency_level`, `credit_level`) VALUES ('2', '西安', '西安火车站', '34.28405900', '108.9688710', 'A', '2');INSERT INTO `hb-test`.`t_city` (`city_code`, `city_name`, `addr`, `lat`, `lon`, `efficiency_level`, `credit_level`) VALUES ('2', '西安', '西安北站', '34.38171600', '108.9452810', 'B', '3');INSERT INTO `hb-test`.`t_city` (`city_code`, `city_name`, `addr`, `lat`, `lon`, `efficiency_level`, `credit_level`) VALUES ('2', '西安', '电子正街', '34.21045200', '108.9257140', 'C', '2');INSERT INTO `hb-test`.`t_city` (`city_code`, `city_name`, `addr`, `lat`, `lon`, `efficiency_level`, `credit_level`) VALUES ('2', '西安', '零壹广场', '34.23076900', '108.8875360', 'D', '1');INSERT INTO `hb-test`.`t_city` (`city_code`, `city_name`, `addr`, `lat`, `lon`, `efficiency_level`, `credit_level`) VALUES ('2', '西安', '软件新城', '34.21228800', '108.8411110', 'D', '2');
3、第一种查询sql：基于球面距离公式
纬度最大是90度，大于90度的一定是经度。百度地图上的坐标拾取。精度在前，和经纬度的字面意思一致
地球平均半径约：6371km 最大半径（赤道半径）约：6378km
#以西安北站（108.945281,34.381716）为原点，计算出有经纬度的城市距离当前远点的距离并正序排列
SELECT id, city_name, addr, lat, lon, efficiency_level, credit_level, ROUND( 6378 * 2 * ASIN( SQRT( POW( SIN( ( 34.381716 * PI() / 180 - lat * PI() / 180 ) / 2 ), 2 ) + COS(34.381716 * PI() / 180) * COS(lat * PI() / 180) * POW( SIN( ( 108.945281 * PI() / 180 - lon * PI() / 180 ) / 2 ), 2 ) ) ) * 1000 ) AS distanceFROM t_cityWHERE lat IS NOT NULLORDER BY distance ASC, efficiency_level DESC, credit_level ASCLIMIT 2000
4、第二种查询sql：基于mysql自带函数st_distance
#以西安北站（108.945281,34.381716）为原点，计算出有经纬度的城市距离当前远点的距离并正序排列
SELECT *, ( st_distance ( point (lon, lat), point (108.945281, 34.381716) ) * 6378*1000*PI()/180 ) AS distanceFROM t_citywhere lon IS NOT NULLORDER BY distance ASCLIMIT 2000
二、通过Elasticsearch（7.15.1）的geo_point实现
1、创建索引mapping
PUT t_city{ "settings": { "number_of_shards": 1, "number_of_replicas": 1 }, "mappings": { "properties": { "addr": { "type": "keyword" }, "city_code": { "type": "keyword" }, "city_name": { "type": "keyword" }, "id": { "type": "keyword" }, "location": { "type": "geo_point" }, "time": { "type": "date" } } }}
2、插入测试数据
POST /t_city/_bulk
{"index":{}}
{"city_code":"1","city_name":"北京","addr":"北海公园","time":"2022-01-23","location":{"lat":39.934633,"lon":39.934633}}
{"index":{}}
{"city_code":"2","city_name":"西安","addr":"西安火车站","time":"2022-01-23","location":{"lat":34.284059,"lon":108.96887}}
{"index":{}}
{"city_code":"2","city_name":"上海","addr":"上海虹桥机场","time":"2022-01-23","location":{"lat":31.203347,"lon":121.346817}}
3、geo_distance 询问：查找在中心点指定距离内的地理点（更多用法参考官方文档）
#以西安北站（108.94528,34.381716）为原点，查询距离西安北站100km以内的10个点
GET /t_city/_search{ "size": 10, "query": { "bool": { "filter": [ { "geo_distance": { "distance": "100km", "location": { "lat": 34.381716, "lon": 108.945281 } } } ] } }}
4、距离范围内的桶聚合（更多功能参考官方文档）
#以西安北站（108.94528,34.381716）为原点
#查询落在(*-100000),(100000-300000),(300000-*)的点
GET /t_city/_search{ "size": 1, "track_total_hits": true, "query": { "match_all": {} }, "aggs": { "distance": { "geo_distance": { "field": "location", "unit": "m", "origin": { "lat": 34.381716, "lon": 108.945281 }, "ranges": [ { "to": 100000 }, { "from": 100000, "to": 300000 }, { "from": 300000 } ] } } }}
注意：Note that this aggregation includes thefromvalue and excludes thetovalue for each range.
后记：很多数据库都有对GEO地理位置数据的处理，除了文中说到的MySQL、Elasticsearch；还有Redis、MongoDB等。
