join filter: 多表联结的时候，表与表之间联结时候的过滤条件
区分内联结和外联结来说明区别。
2. 建表，插入测试数据
@H_502_13@postgres=# create table tb13(id integer,info character varying); CREATE TABLE postgres=# create table tb14(id integer,info character varying); CREATE TABLE postgres=# postgres=# insert into tb13 values(1,'tb13'); INSERT 0 1 postgres=# insert into tb14 values(1,'tb14'); INSERT 0 1 3. outer join：filter和join filter有区别。①、使用join filter
@H_502_13@postgres=# select * from tb13 left join tb14 on tb13.id=tb14.id and tb13.info='no'; id | info | id | info ----+------+----+------ 1 | tb13 | | (1 row)可以看到过滤条件tb13.info='no'对表tb13来说没有起作用。
查看执行计划：
@H_502_13@postgres=# explain analyze select * from tb13 left join tb14 on tb13.id=tb14.id and tb13.info='no'; QUERY PLAN --------------------------------------------------------------------------------------------------------------- Merge Left Join (cost=170.85..309.37 rows=1230 width=72) (actual time=0.027..0.027 rows=1 loops=1) Merge Cond: (tb13.id = tb14.id) <span style="color:#ff0000;">Join Filter: ((tb13.info)::text = 'no'::text)</span> Rows Removed by Join Filter: 1 -> Sort (cost=85.43..88.50 rows=1230 width=36) (actual time=0.014..0.014 rows=1 loops=1) Sort Key: tb13.id Sort Method: quicksort Memory: 25kB -> Seq Scan on tb13 (cost=0.00..22.30 rows=1230 width=36) (actual time=0.007..0.008 rows=1 loops=1) -> Sort (cost=85.43..88.50 rows=1230 width=36) (actual time=0.004..0.004 rows=1 loops=1) Sort Key: tb14.id Sort Method: quicksort Memory: 25kB -> Seq Scan on tb14 (cost=0.00..22.30 rows=1230 width=36) (actual time=0.002..0.002 rows=1 loops=1) Total runtime: 0.067 ms (13 rows) ②、使用filter
@H_502_13@postgres=# select * from tb13 left join tb14 on tb13.id=tb14.id where tb13.info='no'; id | info | id | info ----+------+----+------ (0 rows)可以看到过滤条件tb13.info='no'已经起作用。
查看执行计划：
@H_502_13@postgres=# explain analyze select * from tb13 left join tb14 on tb13.id=tb14.id where tb13.info='no'; QUERY PLAN ------------------------------------------------------------------------------------------------------------ Hash Right Join (cost=25.45..52.73 rows=37 width=72) (actual time=5.707..5.707 rows=0 loops=1) Hash Cond: (tb14.id = tb13.id) -> Seq Scan on tb14 (cost=0.00..22.30 rows=1230 width=36) (never executed) -> Hash (cost=25.38..25.38 rows=6 width=36) (actual time=0.017..0.017 rows=0 loops=1) Buckets: 1024 Batches: 1 Memory Usage: 0kB -> Seq Scan on tb13 (cost=0.00..25.38 rows=6 width=36) (actual time=0.017..0.017 rows=0 loops=1) <span style="color:#ff0000;">Filter: ((info)::text = 'no'::text)</span> Rows Removed by Filter: 1 Total runtime: 34.031 ms (9 rows) 4. inner join: filter和join filter没有区别。
①、使用 join filter
@H_502_13@postgres=# select * from tb13 inner join tb14 on tb13.id=tb14.id and tb13.info='no'; id | info | id | info ----+------+----+------ (0 rows)查看执行计划：
@H_502_13@postgres=# explain analyze select * from tb13 inner join tb14 on tb13.id=tb14.id and tb13.info='no'; QUERY PLAN ------------------------------------------------------------------------------------------------------------ Hash Join (cost=25.45..52.73 rows=37 width=72) (actual time=0.034..0.034 rows=0 loops=1) Hash Cond: (tb14.id = tb13.id) -> Seq Scan on tb14 (cost=0.00..22.30 rows=1230 width=36) (actual time=0.006..0.006 rows=1 loops=1) -> Hash (cost=25.38..25.38 rows=6 width=36) (actual time=0.022..0.022 rows=0 loops=1) Buckets: 1024 Batches: 1 Memory Usage: 0kB -> Seq Scan on tb13 (cost=0.00..25.38 rows=6 width=36) (actual time=0.022..0.022 rows=0 loops=1) <span style="color:#ff0000;">Filter: ((info)::text = 'no'::text)</span> Rows Removed by Filter: 1 Total runtime: 0.062 ms (9 rows) ②、使用 filter
@H_502_13@postgres=# select * from tb13 inner join tb14 on tb13.id=tb14.id where tb13.info='no'; id | info | id | info ----+------+----+------ (0 rows)查看执行计划：
@H_502_13@postgres=# explain analyze select * from tb13 inner join tb14 on tb13.id=tb14.id where tb13.info='no'; QUERY PLAN ------------------------------------------------------------------------------------------------------------ Hash Join (cost=25.45..52.73 rows=37 width=72) (actual time=0.015..0.015 rows=0 loops=1) Hash Cond: (tb14.id = tb13.id) -> Seq Scan on tb14 (cost=0.00..22.30 rows=1230 width=36) (actual time=0.006..0.006 rows=1 loops=1) -> Hash (cost=25.38..25.38 rows=6 width=36) (actual time=0.004..0.004 rows=0 loops=1) Buckets: 1024 Batches: 1 Memory Usage: 0kB -> Seq Scan on tb13 (cost=0.00..25.38 rows=6 width=36) (actual time=0.004..0.004 rows=0 loops=1) <span style="color:#ff0000;">Filter: ((info)::text = 'no'::text)</span> Rows Removed by Filter: 1 Total runtime: 0.036 ms (9 rows)
可以看到，过滤条件都起作用，在执行计划里面也可以看到在内联结中都是当作filter来操作的。
版权声明：本文内容由互联网用户自发贡献，该文观点与技术仅代表作者本人。本站仅提供信息存储空间服务，不拥有所有权，不承担相关法律责任。如发现本站有涉嫌侵权/违法违规的内容， 请发送邮件至 [email protected] 举报，一经查实，本站将立刻删除。
