在Oracle中，和子查询相关的查询转换有哪些？
♣答案部分
（一）子查询推进（Push Subquery）示例
1LHR@orclasm > set serveroutput on 2LHR@orclasm > exec sql_explain('SELECT /*+ no_push_subq(@lhr_ps)*/ * FROM SYS.TAB$ A WHERE A.ANALYZETIME > (SELECT /*+qb_name(lhr_ps)*/ MAX(B.ANALYZETIME) FROM SYS.IND$ B)','outline'); 3Plan hash value: 553156288 4---------------------------------------------------------------------------- 5| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 6---------------------------------------------------------------------------- 7| 0 | SELECT STATEMENT | | 3322 | 454K| 1118 (1)| 00:00:14 | 8|* 1 | FILTER | | | | | | 9| 2 | TABLE ACCESS FULL | TAB$ | 3322 | 454K| 559 (1)| 00:00:07 |10| 3 | SORT AGGREGATE | | 1 | 7 | | |11| 4 | TABLE ACCESS FULL| IND$ | 5545 | 38815 | 559 (1)| 00:00:07 |12----------------------------------------------------------------------------13Outline Data14-------------15/*+16BEGIN_OUTLINE_DATA17FULL(@"LHR_PS" "B"@"LHR_PS")18FULL(@"SEL$1" "A"@"SEL$1")19OUTLINE(@"LHR_PS")20OUTLINE_LEAF(@"SEL$1")21OUTLINE_LEAF(@"LHR_PS")22ALL_ROWS23DB_VERSION('11.2.0.3')24OPTIMIZER_FEATURES_ENABLE('11.2.0.3')25IGNORE_OPTIM_EMBEDDED_HINTS26END_OUTLINE_DATA27*/28Predicate Information (identified by operation id):29---------------------------------------------------301 - filter("A"."ANALYZETIME"> (SELECT /*+ NO_PUSH_SUBQ QB_NAME31("LHR_PS") */ MAX("B"."ANALYZETIME") FROM "SYS"."IND$" "B"))3233PL/SQL procedure successfully completed.3435LHR@orclasm > exec sql_explain('SELECT * FROM SYS.TAB$ A WHERE A.ANALYZETIME > (SELECT MAX(B.ANALYZETIME) FROM SYS.IND$ B)','outline');36Plan hash value: 24338703837----------------------------------------------------------------------------38| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |39----------------------------------------------------------------------------40| 0 | SELECT STATEMENT | | 166 | 23240 | 1118 (1)| 00:00:14 |41|* 1 | TABLE ACCESS FULL | TAB$ | 166 | 23240 | 559 (1)| 00:00:07 |42| 2 | SORT AGGREGATE | | 1 | 7 | | |43| 3 | TABLE ACCESS FULL| IND$ | 5545 | 38815 | 559 (1)| 00:00:07 |44----------------------------------------------------------------------------45Outline Data46-------------47/*+48BEGIN_OUTLINE_DATA49FULL(@"SEL$2" "B"@"SEL$2")50PUSH_SUBQ(@"SEL$2")51FULL(@"SEL$1" "A"@"SEL$1")52OUTLINE_LEAF(@"SEL$1")53OUTLINE_LEAF(@"SEL$2")54ALL_ROWS55DB_VERSION('11.2.0.3')56OPTIMIZER_FEATURES_ENABLE('11.2.0.3')57IGNORE_OPTIM_EMBEDDED_HINTS58END_OUTLINE_DATA59*/60Predicate Information (identified by operation id):61---------------------------------------------------621 - filter("A"."ANALYZETIME"> (SELECT MAX("B"."ANALYZETIME") FROM "SYS"."IND$" "B"))6364PL/SQL procedure successfully completed.set serveroutput on 2LHR@orclasm > exec sql_explain('SELECT /*+ no_push_subq(@lhr_ps)*/ * FROM SYS.TAB$ A WHERE A.ANALYZETIME > (SELECT /*+qb_name(lhr_ps)*/ MAX(B.ANALYZETIME) FROM SYS.IND$ B)','outline'); 3Plan hash value: 553156288 4---------------------------------------------------------------------------- 5| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 6---------------------------------------------------------------------------- 7| 0 | SELECT STATEMENT | | 3322 | 454K| 1118 (1)| 00:00:14 | 8|* 1 | FILTER | | | | | | 9| 2 | TABLE ACCESS FULL | TAB$ | 3322 | 454K| 559 (1)| 00:00:07 | 10| 3 | SORT AGGREGATE | | 1 | 7 | | | 11| 4 | TABLE ACCESS FULL| IND$ | 5545 | 38815 | 559 (1)| 00:00:07 | 12---------------------------------------------------------------------------- 13Outline Data 14------------- 15/*+ 16BEGIN_OUTLINE_DATA 17FULL(@"LHR_PS" "B"@"LHR_PS") 18FULL(@"SEL$1" "A"@"SEL$1") 19OUTLINE(@"LHR_PS") 20OUTLINE_LEAF(@"SEL$1") 21OUTLINE_LEAF(@"LHR_PS") 22ALL_ROWS 23DB_VERSION('11.2.0.3') 24OPTIMIZER_FEATURES_ENABLE('11.2.0.3') 25IGNORE_OPTIM_EMBEDDED_HINTS 26END_OUTLINE_DATA 27*/ 28Predicate Information (identified by operation id): 29--------------------------------------------------- 301 - filter("A"."ANALYZETIME"> (SELECT /*+ NO_PUSH_SUBQ QB_NAME 31("LHR_PS") */ MAX("B"."ANALYZETIME") FROM "SYS"."IND$" "B")) 32 33PL/SQL procedure successfully completed. 34 35LHR@orclasm > exec sql_explain('SELECT * FROM SYS.TAB$ A WHERE A.ANALYZETIME > (SELECT MAX(B.ANALYZETIME) FROM SYS.IND$ B)','outline'); 36Plan hash value: 243387038 37---------------------------------------------------------------------------- 38| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 39---------------------------------------------------------------------------- 40| 0 | SELECT STATEMENT | | 166 | 23240 | 1118 (1)| 00:00:14 | 41|* 1 | TABLE ACCESS FULL | TAB$ | 166 | 23240 | 559 (1)| 00:00:07 | 42| 2 | SORT AGGREGATE | | 1 | 7 | | | 43| 3 | TABLE ACCESS FULL| IND$ | 5545 | 38815 | 559 (1)| 00:00:07 | 44---------------------------------------------------------------------------- 45Outline Data 46------------- 47/*+ 48BEGIN_OUTLINE_DATA 49FULL(@"SEL$2" "B"@"SEL$2") 50PUSH_SUBQ(@"SEL$2") 51FULL(@"SEL$1" "A"@"SEL$1") 52OUTLINE_LEAF(@"SEL$1") 53OUTLINE_LEAF(@"SEL$2") 54ALL_ROWS 55DB_VERSION('11.2.0.3') 56OPTIMIZER_FEATURES_ENABLE('11.2.0.3') 57IGNORE_OPTIM_EMBEDDED_HINTS 58END_OUTLINE_DATA 59*/ 60Predicate Information (identified by operation id): 61--------------------------------------------------- 621 - filter("A"."ANALYZETIME"> (SELECT MAX("B"."ANALYZETIME") FROM "SYS"."IND$" "B")) 63 64PL/SQL procedure successfully completed.
（二）子查询展开（Subquery Unnesting）
1--1）IN和EXISTS转换为半连接（SEMI JOIN）： 2CREATE TABLE EMP_LHR AS SELECT * FROM SCOTT.EMP; 3CREATE TABLE DEPT_LHR AS SELECT * FROM SCOTT.DEPT; 4SELECT * FROM EMP_LHR A WHERE EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO); 5------------------------------------------------------------------------------- 6| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 7------------------------------------------------------------------------------- 8| 0 | SELECT STATEMENT | | 14 | 1400 | 7 (15)| 00:00:01 | 9|* 1 | HASH JOIN SEMI | | 14 | 1400 | 7 (15)| 00:00:01 |10| 2 | TABLE ACCESS FULL| EMP_LHR | 14 | 1218 | 3 (0)| 00:00:01 |11| 3 | TABLE ACCESS FULL| DEPT_LHR | 4 | 52 | 3 (0)| 00:00:01 |12-------------------------------------------------------------------------------1314--子查询引用表DEPT，最终转换为两个表的哈希半连接。也就是说，EXISTS子句中的子查询被展开，其中的对象与主查询中的对象直接进行半关联操作。IN的情况类似，如下：15SELECT * FROM EMP_LHR A WHERE A.DEPTNO IN (SELECT B.DEPTNO FROM DEPT_LHR B);16-------------------------------------------------------------------------------17| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |18-------------------------------------------------------------------------------19| 0 | SELECT STATEMENT | | 14 | 1400 | 7 (15)| 00:00:01 |20|* 1 | HASH JOIN SEMI | | 14 | 1400 | 7 (15)| 00:00:01 |21| 2 | TABLE ACCESS FULL| EMP_LHR | 14 | 1218 | 3 (0)| 00:00:01 |22| 3 | TABLE ACCESS FULL| DEPT_LHR | 4 | 52 | 3 (0)| 00:00:01 |23-------------------------------------------------------------------------------2425--2）NOT IN和NOT EXISTS转换为反连接（ANTI JOIN）：26SELECT * FROM EMP_LHR A WHERE NOT EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO);27-------------------------------------------------------------------------------28| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |29-------------------------------------------------------------------------------30| 0 | SELECT STATEMENT | | 14 | 1400 | 7 (15)| 00:00:01 |31|* 1 | HASH JOIN ANTI | | 14 | 1400 | 7 (15)| 00:00:01 |32| 2 | TABLE ACCESS FULL| EMP_LHR | 14 | 1218 | 3 (0)| 00:00:01 |33| 3 | TABLE ACCESS FULL| DEPT_LHR | 4 | 52 | 3 (0)| 00:00:01 |34-------------------------------------------------------------------------------3536--优化器将NOT EXISTS后的子查询做解嵌套，然后选择了哈希的反连接。这种转换属于基于代价的查询转换。下面看看NOT IN的情况：37SELECT * FROM EMP_LHR A WHERE A.DEPTNO NOT IN (SELECT B.DEPTNO FROM DEPT_LHR B);38-------------------------------------------------------------------------------39| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |40-------------------------------------------------------------------------------41| 0 | SELECT STATEMENT | | 14 | 1400 | 7 (15)| 00:00:01 |42|* 1 | HASH JOIN ANTI NA | | 14 | 1400 | 7 (15)| 00:00:01 |43| 2 | TABLE ACCESS FULL| EMP_LHR | 14 | 1218 | 3 (0)| 00:00:01 |44| 3 | TABLE ACCESS FULL| DEPT_LHR | 4 | 52 | 3 (0)| 00:00:01 |45-------------------------------------------------------------------------------46--1）IN和EXISTS转换为半连接（SEMI JOIN）： 2CREATE TABLE EMP_LHR AS SELECT * FROM SCOTT.EMP; 3CREATE TABLE DEPT_LHR AS SELECT * FROM SCOTT.DEPT; 4SELECT * FROM EMP_LHR A WHERE EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO); 5------------------------------------------------------------------------------- 6| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 7------------------------------------------------------------------------------- 8| 0 | SELECT STATEMENT | | 14 | 1400 | 7 (15)| 00:00:01 | 9|* 1 | HASH JOIN SEMI | | 14 | 1400 | 7 (15)| 00:00:01 | 10| 2 | TABLE ACCESS FULL| EMP_LHR | 14 | 1218 | 3 (0)| 00:00:01 | 11| 3 | TABLE ACCESS FULL| DEPT_LHR | 4 | 52 | 3 (0)| 00:00:01 | 12------------------------------------------------------------------------------- 13 14--子查询引用表DEPT，最终转换为两个表的哈希半连接。也就是说，EXISTS子句中的子查询被展开，其中的对象与主查询中的对象直接进行半关联操作。IN的情况类似，如下： 15SELECT * FROM EMP_LHR A WHERE A.DEPTNO IN (SELECT B.DEPTNO FROM DEPT_LHR B); 16------------------------------------------------------------------------------- 17| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 18------------------------------------------------------------------------------- 19| 0 | SELECT STATEMENT | | 14 | 1400 | 7 (15)| 00:00:01 | 20|* 1 | HASH JOIN SEMI | | 14 | 1400 | 7 (15)| 00:00:01 | 21| 2 | TABLE ACCESS FULL| EMP_LHR | 14 | 1218 | 3 (0)| 00:00:01 | 22| 3 | TABLE ACCESS FULL| DEPT_LHR | 4 | 52 | 3 (0)| 00:00:01 | 23------------------------------------------------------------------------------- 24 25--2）NOT IN和NOT EXISTS转换为反连接（ANTI JOIN）： 26SELECT * FROM EMP_LHR A WHERE NOT EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO); 27------------------------------------------------------------------------------- 28| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 29------------------------------------------------------------------------------- 30| 0 | SELECT STATEMENT | | 14 | 1400 | 7 (15)| 00:00:01 | 31|* 1 | HASH JOIN ANTI | | 14 | 1400 | 7 (15)| 00:00:01 | 32| 2 | TABLE ACCESS FULL| EMP_LHR | 14 | 1218 | 3 (0)| 00:00:01 | 33| 3 | TABLE ACCESS FULL| DEPT_LHR | 4 | 52 | 3 (0)| 00:00:01 | 34------------------------------------------------------------------------------- 35 36--优化器将NOT EXISTS后的子查询做解嵌套，然后选择了哈希的反连接。这种转换属于基于代价的查询转换。下面看看NOT IN的情况： 37SELECT * FROM EMP_LHR A WHERE A.DEPTNO NOT IN (SELECT B.DEPTNO FROM DEPT_LHR B); 38------------------------------------------------------------------------------- 39| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 40------------------------------------------------------------------------------- 41| 0 | SELECT STATEMENT | | 14 | 1400 | 7 (15)| 00:00:01 | 42|* 1 | HASH JOIN ANTI NA | | 14 | 1400 | 7 (15)| 00:00:01 | 43| 2 | TABLE ACCESS FULL| EMP_LHR | 14 | 1218 | 3 (0)| 00:00:01 | 44| 3 | TABLE ACCESS FULL| DEPT_LHR | 4 | 52 | 3 (0)| 00:00:01 | 45------------------------------------------------------------------------------- 46
和NOT EXISTS类似，也选择了哈希连接，只不过是HASH JOIN ANTI NA。这里的NA，实际表示Null-Aware的意思，在11g及以后的版本中，Oracle增加了对空值敏感的反关联的支持。
（三）子查询合并（Subquery Coalesce）
1LHR@orclasm > SELECT * FROM EMP_LHR A WHERE EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO AND B.DEPTNO=10) AND EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO AND B.DEPTNO=20) ; 2 3no rows selected 4 5 6Execution Plan 7---------------------------------------------------------- 8Plan hash value: 3115025369 910---------------------------------------------------------------------------------11| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |12---------------------------------------------------------------------------------13| 0 | SELECT STATEMENT | | 1 | 113 | 0 (0)| |14|* 1 | FILTER | | | | | |15|* 2 | HASH JOIN SEMI | | 1 | 113 | 10 (10)| 00:00:01 |16|* 3 | HASH JOIN SEMI | | 1 | 100 | 7 (15)| 00:00:01 |17|* 4 | TABLE ACCESS FULL| EMP_LHR | 1 | 87 | 3 (0)| 00:00:01 |18|* 5 | TABLE ACCESS FULL| DEPT_LHR | 1 | 13 | 3 (0)| 00:00:01 |19|* 6 | TABLE ACCESS FULL | DEPT_LHR | 1 | 13 | 3 (0)| 00:00:01 |20---------------------------------------------------------------------------------2122Predicate Information (identified by operation id):23---------------------------------------------------2425 1 - filter(NULL IS NOT NULL)26 2 - access("B"."DEPTNO"="A"."DEPTNO")27 3 - access("B"."DEPTNO"="A"."DEPTNO")28 4 - filter("A"."DEPTNO"=10 AND "A"."DEPTNO"=20)29 5 - filter("B"."DEPTNO"=20 AND "B"."DEPTNO"=10)30 6 - filter("B"."DEPTNO"=10 AND "B"."DEPTNO"=20)SELECT * FROM EMP_LHR A WHERE EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO AND B.DEPTNO=10) AND EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO AND B.DEPTNO=20) ; 2 3no rows selected 4 5 6Execution Plan 7---------------------------------------------------------- 8Plan hash value: 3115025369 9 10--------------------------------------------------------------------------------- 11| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 12--------------------------------------------------------------------------------- 13| 0 | SELECT STATEMENT | | 1 | 113 | 0 (0)| | 14|* 1 | FILTER | | | | | | 15|* 2 | HASH JOIN SEMI | | 1 | 113 | 10 (10)| 00:00:01 | 16|* 3 | HASH JOIN SEMI | | 1 | 100 | 7 (15)| 00:00:01 | 17|* 4 | TABLE ACCESS FULL| EMP_LHR | 1 | 87 | 3 (0)| 00:00:01 | 18|* 5 | TABLE ACCESS FULL| DEPT_LHR | 1 | 13 | 3 (0)| 00:00:01 | 19|* 6 | TABLE ACCESS FULL | DEPT_LHR | 1 | 13 | 3 (0)| 00:00:01 | 20--------------------------------------------------------------------------------- 21 22Predicate Information (identified by operation id): 23--------------------------------------------------- 24 25 1 - filter(NULL IS NOT NULL) 26 2 - access("B"."DEPTNO"="A"."DEPTNO") 27 3 - access("B"."DEPTNO"="A"."DEPTNO") 28 4 - filter("A"."DEPTNO"=10 AND "A"."DEPTNO"=20) 29 5 - filter("B"."DEPTNO"=20 AND "B"."DEPTNO"=10) 30 6 - filter("B"."DEPTNO"=10 AND "B"."DEPTNO"=20)
在这个查询语句中，外部查询要满足两个子查询—SUB1和SUB2，但两者条件不同，不能简单合并。因此在执行计划中，分别对两者进行了扫描（直观感觉就是对DEPT_LHR进行了两次扫描），然后再做关联查询。
1LHR@orclasm > SELECT * FROM EMP_LHR A WHERE EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO AND B.DEPTNO=10) AND EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO) ; 2 3 EMPNO ENAME JOB MGR HIREDATE SAL COMM DEPTNO 4---------- ---------- --------- ---------- ------------------- ---------- ---------- ---------- 5 7934 MILLER CLERK 7782 1982-01-23 00:00:00 1300 10 6 7839 KING PRESIDENT 1981-11-17 00:00:00 5000 10 7 7782 CLARK MANAGER 7839 1981-06-09 00:00:00 2450 10 8 910Execution Plan11----------------------------------------------------------12Plan hash value: 34036918551314-------------------------------------------------------------------------------15| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time |16-------------------------------------------------------------------------------17| 0 | SELECT STATEMENT | | 1 | 100 | 7 (15)| 00:00:01 |18|* 1 | HASH JOIN SEMI | | 1 | 100 | 7 (15)| 00:00:01 |19|* 2 | TABLE ACCESS FULL| EMP_LHR | 3 | 261 | 3 (0)| 00:00:01 |20|* 3 | TABLE ACCESS FULL| DEPT_LHR | 1 | 13 | 3 (0)| 00:00:01 |21-------------------------------------------------------------------------------2223Predicate Information (identified by operation id):24---------------------------------------------------2526 1 - access("B"."DEPTNO"="A"."DEPTNO")27 2 - filter("A"."DEPTNO"=10)28 3 - filter("B"."DEPTNO"=10)SELECT * FROM EMP_LHR A WHERE EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO AND B.DEPTNO=10) AND EXISTS (SELECT 1 FROM DEPT_LHR B WHERE B.DEPTNO=A.DEPTNO) ; 2 3 EMPNO ENAME JOB MGR HIREDATE SAL COMM DEPTNO 4---------- ---------- --------- ---------- ------------------- ---------- ---------- ---------- 5 7934 MILLER CLERK 7782 1982-01-23 00:00:00 1300 10 6 7839 KING PRESIDENT 1981-11-17 00:00:00 5000 10 7 7782 CLARK MANAGER 7839 1981-06-09 00:00:00 2450 10 8 9 10Execution Plan 11---------------------------------------------------------- 12Plan hash value: 3403691855 13 14------------------------------------------------------------------------------- 15| Id | Operation | Name | Rows | Bytes | Cost (%CPU)| Time | 16------------------------------------------------------------------------------- 17| 0 | SELECT STATEMENT | | 1 | 100 | 7 (15)| 00:00:01 | 18|* 1 | HASH JOIN SEMI | | 1 | 100 | 7 (15)| 00:00:01 | 19|* 2 | TABLE ACCESS FULL| EMP_LHR | 3 | 261 | 3 (0)| 00:00:01 | 20|* 3 | TABLE ACCESS FULL| DEPT_LHR | 1 | 13 | 3 (0)| 00:00:01 | 21------------------------------------------------------------------------------- 22 23Predicate Information (identified by operation id): 24--------------------------------------------------- 25 26 1 - access("B"."DEPTNO"="A"."DEPTNO") 27 2 - filter("A"."DEPTNO"=10) 28 3 - filter("B"."DEPTNO"=10)
在这个查询中，外部对EMP_LHR表的查询要同时满足SUB1和SUB2两个子查询，而SUB1在语义上又是SUB2的子集，因此优化器将两个子查询进行了合并（只进行一次对DEPT_LHR表的扫描），然后与外部表EMP_LHR进行半连接。
本文选自《Oracle程序员面试笔试宝典》，作者：李华荣。
---------------优质麦课------------
详细内容可以添加麦老师微信或QQ私聊。
About Me：小麦苗
● 本文作者：小麦苗，只专注于数据库的技术，更注重技术的运用
● 作者博客地址：http://blog.itpub.net/26736162/abstract/1/
● 本系列题目来源于作者的学习笔记，部分整理自网络，若有侵权或不当之处还请谅解
● 版权所有，欢迎分享本文，转载请保留出处
● QQ：646634621 QQ群：618766405
● 提供OCP、OCM和高可用部分最实用的技能培训
● 题目解答若有不当之处，还望各位朋友批评指正，共同进步
DBA宝典
长按下图识别二维码或微信扫描下图二维码来关注小麦苗的微信公众号：xiaomaimiaolhr，学习最实用的数据库技术。
喜欢就点击“好看”吧
