曾经一段时间我对oracle的多表查询搞的云里雾里,究其原因:oracle自己的语法和SQL国际标准语法混用。此文章仅适合oracle 菜鸟,老鸟直接飞过…
• Cross joins
• Natural joins
• USING clause
• Full (or two-sided) outer joins
• Arbitrary join conditions for outer joins
SELECT table1.column, table2.column
FROM table1
[CROSS JOIN table2] |
[NATURAL JOIN table2] |
[JOIN table2 USING (column_name)] |
[JOIN table2
ON (table1.column_name = table2.column_name)]|
[LEFT|RIGHT|FULL OUTER JOIN table2
ON (table1.column_name = table2.column_name)]|
[CROSS JOIN table2];
语法解释:
table1.column --指明从中检索数据的表和列
CROSS JOIN --返回两个表的笛卡尔集
NATURAL JOIN --根据相同的列名连接两个表
JOIN table
USING column_name --根据列名执行等值连接
JOIN table ON
table1.column_name --根据ON 子句中的条件执行等值连接
= table2.column_name
LEFT/RIGHT/FULL OUTER
一般来说,从数据显示方式来讲,分为内连接和外连接
内连接:只返回满足连接条件的数据。
外连接:除了返回满足连接条的行以外,还返回左(右)表中,不满足条件的行,
称为左(右)连接
演示,主要以SQL标准为主,oracle 写法作对比。示例用户为scott、HR
解锁这两个用户语句:
alter user scott identified by tiger account unlock;
alter user hr identified by hr account unlock;
--Oracle的写法
select empno,ename,sal,dname,loc from emp,dept
where emp.deptno=dept.deptno;
-- SQL 99标准的写法
select empno,ename,job,sal,dept.deptno,dname,loc
from emp join dept on emp.deptno=dept.deptno;
或把join改为inner join
Oracle 外连接语法:
SELECT table1.column, table2.column --右外连接
FROM table1, table2
WHERE table1.column(+) = table2.column;
SELECT table1.column, table2.column --左外连接
FROM table1, table2
WHERE table1.column = table2.column(+);
SQL 1999标准外连接语法见上面SQL1999语法
--Oracle的写法
外连接的符号是(+),(+)要放在字段名后。(+)对面的那个表,会全部显示。
左外连接时,加号在等号的右边
SQL> select d.dname,e.ename,e.deptno
from dept d,emp e
where d.deptno = e.deptno(+)
order by d.deptno;
DNAME ENAME DEPTNO
-------------- ---------- ------
ACCOUNTING CLARK 10
ACCOUNTING KING 10
ACCOUNTING MILLER 10
RESEARCH JONES 20
RESEARCH FORD 20
RESEARCH ADAMS 20
RESEARCH SMITH 20
RESEARCH SCOTT 20
SALES WARD 30
SALES TURNER 30
SALES ALLEN 30
SALES JAMES 30
SALES BLAKE 30
SALES MARTIN 30
OPERATIONS
15 rows selected
-- SQL 99标准的写法接
SQL> select d.dname,e.ename,e.deptno
from dept d
left join emp e
on d.deptno = e.deptno
order by d.deptno;
DNAME ENAME DEPTNO
-------------- ---------- ------
ACCOUNTING CLARK 10
ACCOUNTING KING 10
ACCOUNTING MILLER 10
RESEARCH JONES 20
RESEARCH FORD 20
RESEARCH ADAMS 20
RESEARCH SMITH 20
RESEARCH SCOTT 20
SALES WARD 30
SALES TURNER 30
SALES ALLEN 30
SALES JAMES 30
SALES BLAKE 30
SALES MARTIN 30
OPERATIONS
15 rows selected
--oracle的写法
SQL> select empno, ename, job, sal, dept.deptno, dname, loc
2 from emp, dept
3 where emp.deptno(+) = dept.deptno;
EMPNO ENAME JOB SAL DEPTNO DNAME LOC
----- ---------- --------- --------- ------ -------------- -------------
7782 CLARK MANAGER 2450.00 10 ACCOUNTING NEW YORK
7839 KING PRESIDENT 5000.00 10 ACCOUNTING NEW YORK
7934 MILLER CLERK 1300.00 10 ACCOUNTING NEW YORK
7566 JONES MANAGER 2975.00 20 RESEARCH DALLAS
7902 FORD ANALYST 3000.00 20 RESEARCH DALLAS
7876 ADAMS CLERK 1100.00 20 RESEARCH DALLAS
7369 SMITH CLERK 800.00 20 RESEARCH DALLAS
7788 SCOTT ANALYST 3000.00 20 RESEARCH DALLAS
7521 WARD SALESMAN 1250.00 30 SALES CHICAGO
7844 TURNER SALESMAN 1500.00 30 SALES CHICAGO
7499 ALLEN SALESMAN 1600.00 30 SALES CHICAGO
7900 JAMES CLERK 950.00 30 SALES CHICAGO
7698 BLAKE MANAGER 2850.00 30 SALES CHICAGO
7654 MARTIN SALESMAN 1250.00 30 SALES CHICAGO
40 OPERATIONS BOSTON
15 rows selected
--SQL1999标准写法
SQL> select empno, ename, job, sal, dept.deptno, dname, loc
2 from emp
3 right join dept
4 on emp.deptno = dept.deptno;
EMPNO ENAME JOB SAL DEPTNO DNAME LOC
----- ---------- --------- --------- ------ -------------- -------------
7782 CLARK MANAGER 2450.00 10 ACCOUNTING NEW YORK
7839 KING PRESIDENT 5000.00 10 ACCOUNTING NEW YORK
7934 MILLER CLERK 1300.00 10 ACCOUNTING NEW YORK
7566 JONES MANAGER 2975.00 20 RESEARCH DALLAS
7902 FORD ANALYST 3000.00 20 RESEARCH DALLAS
7876 ADAMS CLERK 1100.00 20 RESEARCH DALLAS
7369 SMITH CLERK 800.00 20 RESEARCH DALLAS
7788 SCOTT ANALYST 3000.00 20 RESEARCH DALLAS
7521 WARD SALESMAN 1250.00 30 SALES CHICAGO
7844 TURNER SALESMAN 1500.00 30 SALES CHICAGO
7499 ALLEN SALESMAN 1600.00 30 SALES CHICAGO
7900 JAMES CLERK 950.00 30 SALES CHICAGO
7698 BLAKE MANAGER 2850.00 30 SALES CHICAGO
7654 MARTIN SALESMAN 1250.00 30 SALES CHICAGO
40 OPERATIONS BOSTON
15 rows selected
--SQL1999标准写法
SQL> select empno, ename, job, sal, d.deptno, dname, loc
2 from emp e
3 full join dept d
4 on e.deptno = d.deptno;
EMPNO ENAME JOB SAL DEPTNO DNAME LOC
----- ---------- --------- --------- ------ -------------- -------------
7369 SMITH CLERK 800.00 20 RESEARCH DALLAS
7499 ALLEN SALESMAN 1600.00 30 SALES CHICAGO
7521 WARD SALESMAN 1250.00 30 SALES CHICAGO
7566 JONES MANAGER 2975.00 20 RESEARCH DALLAS
7654 MARTIN SALESMAN 1250.00 30 SALES CHICAGO
7698 BLAKE MANAGER 2850.00 30 SALES CHICAGO
7782 CLARK MANAGER 2450.00 10 ACCOUNTING NEW YORK
7788 SCOTT ANALYST 3000.00 20 RESEARCH DALLAS
7839 KING PRESIDENT 5000.00 10 ACCOUNTING NEW YORK
7844 TURNER SALESMAN 1500.00 30 SALES CHICAGO
7876 ADAMS CLERK 1100.00 20 RESEARCH DALLAS
7900 JAMES CLERK 950.00 30 SALES CHICAGO
7902 FORD ANALYST 3000.00 20 RESEARCH DALLAS
7934 MILLER CLERK 1300.00 10 ACCOUNTING NEW YORK
40 OPERATIONS BOSTON
15 rows selected
把表自身的镜像当成另外一个表
--oracle 的写法
SQL> select e.ename || ' works for ' || m.ename
2 from emp e, emp m
3 where e.empno = m.mgr;
E.ENAME||'WORKSFOR'||M.ENAME
-------------------------------
JONES works for FORD
JONES works for SCOTT
BLAKE works for TURNER
BLAKE works for ALLEN
BLAKE works for WARD
BLAKE works for JAMES
BLAKE works for MARTIN
CLARK works for MILLER
SCOTT works for ADAMS
KING works for BLAKE
KING works for JONES
KING works for CLARK
FORD works for SMITH
13 rows selected
--SQL1999标准的写法
SQL> select e.ename || ' works for ' || m.ename
2 from emp e
3 join emp m
4 on e.empno = m.mgr;
E.ENAME||'WORKSFOR'||M.ENAME
-------------------------------
JONES works for FORD
JONES works for SCOTT
BLAKE works for TURNER
BLAKE works for ALLEN
BLAKE works for WARD
BLAKE works for JAMES
BLAKE works for MARTIN
CLARK works for MILLER
SCOTT works for ADAMS
KING works for BLAKE
KING works for JONES
KING works for CLARK
FORD works for SMITH
13 rows selected
以两个表具有相同的字段的所有列为基础,可采用自然连接(natural join)
它将选择两个表中那些在所有匹配的列中值相等的行。
如果列具有相同的名称,但数据类型能够不同,则会报错。
--SQL1999标准写法
SQL> select empno, ename, job, sal, deptno, dname, loc
2 from emp natural
3 join dept;
EMPNO ENAME JOB SAL DEPTNO DNAME LOC
----- ---------- --------- --------- ------ -------------- -------------
7782 CLARK MANAGER 2450.00 10 ACCOUNTING NEW YORK
7839 KING PRESIDENT 5000.00 10 ACCOUNTING NEW YORK
7934 MILLER CLERK 1300.00 10 ACCOUNTING NEW YORK
7566 JONES MANAGER 2975.00 20 RESEARCH DALLAS
7902 FORD ANALYST 3000.00 20 RESEARCH DALLAS
7876 ADAMS CLERK 1100.00 20 RESEARCH DALLAS
7369 SMITH CLERK 800.00 20 RESEARCH DALLAS
7788 SCOTT ANALYST 3000.00 20 RESEARCH DALLAS
7521 WARD SALESMAN 1250.00 30 SALES CHICAGO
7844 TURNER SALESMAN 1500.00 30 SALES CHICAGO
7499 ALLEN SALESMAN 1600.00 30 SALES CHICAGO
7900 JAMES CLERK 950.00 30 SALES CHICAGO
7698 BLAKE MANAGER 2850.00 30 SALES CHICAGO
7654 MARTIN SALESMAN 1250.00 30 SALES CHICAGO
14 rows selected
如果几个列具有相同的名称,但是数据类型不匹配,那么可以使用using子句来修改natural join子句以指定要用于等值连接的列,在多个列匹配时,使用using 子句只匹配一个列,在引用列中不要使用表名或别名。natural join 和using 子句是互不相容的。
SELECT l.city, d.department_name
FROM locations l JOIN departments d USING (location_id)
WHERE location_id = 1400;
--下面的语句无效,因为where 子句中限定了location_id为d. location_id
SELECT l.city, d.department_name
FROM locations l JOIN departments d USING (location_id)
WHERE d.location_id = 1400;
ORA-25154: column part of USING clause cannot have qualifier
注意:两个表中名称相同的列在使用时不能有任何限定符,这一限制同样适用natural join
cross join 实现交叉连接
交叉连接产生的结果集也就是笛卡尔乘积
--oracle 的写法
select empno, ename, dname
from emp,dept;
在oracle语法中直接from两张表就是cross join
--SQL1999标准写法
select empno, ename, dname
from emp
cross join dept;
上述的多表连接都是等值连接(最常用),等值连接后面还可以添加限定条件。
--oracle 的写法
select e.empno, e.ename, d.dname
from emp e, dept d
where d.deptno = e.deptno
and e.ename = 'SCOTT';
--SQL1999标准的写法
select e.empno, e.ename, d.dname
from emp e
inner join dept d
on d.deptno = e.deptno
and e.ename = 'SCOTT';
--oracle的写法
select ename, sal, grade
from emp, salgrade
where sal between losal and hisal
and emp.deptno = 20;
--SQL1999标准写法
select e.ename, e.sal, s.grade
from emp e
join salgrade s
on e.sal between losal and hisal
and e.deptno = 20;
Oracle的语法还是比较人性化的,除了左外连接和右外连接。如果想使代码简单明了,使用oracle的语法,如果想使写出的代码兼容性比较好(可能会移植到DB2等),尽量使用SQL标准语法。
详细信息请参考oracle 官方文档
Oracle® Database
SQL Language Reference
11g Release 2 (11.2)
E17118-04
9 SQL Queries and Subqueries 部分