CreateTime--2017年6月1日14:36:37
Author:Marydon
一、SQL语句
（二）提升篇
1.2.1 左连接+分页查询
情景：班级表和学生表两表关联，查询学生信息及所在班级
方法一：使用左连接
SELECT T.*
FROM (SELECT S.*, C.CLAZZNAME, ROWNUM ROWNO
FROM STUDENT3 S
LEFT JOIN CLAZZ3 C
ON S.CLAZZID = C.CLAZZID
WHERE ROWNUM <= ?) T
WHERE T.ROWNO >= ?
UpdateTime--2017年1月22日14:06:25
方法二：使用(+)
SELECT T.*
FROM (SELECT S.*, C.CLAZZNAME, ROWNUM ROWNO
FROM STUDENT3 S, CLAZZ3 C
WHERE S.CLAZZID = C.CLAZZID(+)
AND ROWNUM <= ?) T
WHERE T.ROWNO >= ?
注意：
a.只有Oracle数据库可以使用(+)来代替左连接和右连接；
b."(+)"在"="右侧，表示的是：左连接，如：S.CLAZZID = C.CLAZZID(+)；左表（学生表）为主表；
c."(+)"在"="左侧，表示的是：右连接，如：S.CLAZZID(+) = C.CLAZZID；右表（班级表）为主表。
方法三：from后面跟两张表
SELECT T.*
FROM (SELECT S.*, C.CLAZZNAME, ROWNUM ROWNO
FROM STUDENT3 S, CLAZZ3 C
WHERE S.CLAZZID = C.CLAZZID
AND ROWNUM <= ?) T
WHERE T.ROWNO >= ?
1.2.2 左连接实现三表关联
表A---------------------------------关联第一张表B-----------------------关联第二张表c
语法：
select * from 表名A left join 表B on A.columnX=B.columnM and A.columnY=B.columnN left join 表c on 表A=表c的id
情景：
四张表 GJPT_BASY、GJZY_BASY、GJPT_BASY_ERROR、GJZY_BASY_ERROR
根据四张表，要求返回：医疗机构名称，医疗机构编号，总数，合格数和问题数
SQL实现：
SELECT TEMP1.*, TEMP2.HEGESUM, TEMP3.TROUBLESUM
FROM (SELECT T1.YLNAME, T1.YLCODE, SUM(RS1) AS YLSUM--总数
FROM (SELECT COUNT(1) AS RS1,
HDSD00_11_118 AS YLNAME,
HDSD00_11_119 AS YLCODE
FROM GJPT_BASY
GROUP BY HDSD00_11_119, HDSD00_11_118
UNION ALL
SELECT COUNT(1) AS RS1,
HDSD00_12_133 AS YLNAME,
HDSD00_12_134 AS YLCODE
FROM GJZY_BASY
GROUP BY HDSD00_12_133, HDSD00_12_134
UNION ALL
SELECT COUNT(1) AS RS1,
HDSD00_11_118 AS YLNAME,
HDSD00_11_119 AS YLCODE
FROM GJPT_BASY_ERROR
GROUP BY HDSD00_11_119, HDSD00_11_118
UNION ALL
SELECT COUNT(1) AS RS1,
HDSD00_12_133 AS YLNAME,
HDSD00_12_134 AS YLCODE
FROM GJZY_BASY_ERROR
GROUP BY HDSD00_12_133, HDSD00_12_134) T1
GROUP BY T1.YLNAME, T1.YLCODE) TEMP1
LEFT JOIN (SELECT *
FROM (SELECT T2.YLNAME, T2.YLCODE, SUM(RS2) AS HEGESUM--合格数
FROM (SELECT COUNT(1) AS RS2,
HDSD00_11_118 AS YLNAME,
HDSD00_11_119 AS YLCODE
FROM GJPT_BASY
GROUP BY HDSD00_11_119, HDSD00_11_118
UNION ALL
SELECT COUNT(1) AS RS2,
HDSD00_12_133 AS YLNAME,
HDSD00_12_134 AS YLCODE
FROM GJZY_BASY
GROUP BY HDSD00_12_133, HDSD00_12_134) T2
GROUP BY T2.YLNAME, T2.YLCODE)) TEMP2
ON TEMP2.YLNAME = TEMP1.YLNAME
AND TEMP2.YLCODE = TEMP1.YLCODE
LEFT JOIN (SELECT *
FROM (SELECT T3.YLNAME, T3.YLCODE, SUM(RS3) TROUBLESUM--问题数
FROM (SELECT COUNT(1) AS RS3,
HDSD00_11_118 AS YLNAME,
HDSD00_11_119 AS YLCODE
FROM GJPT_BASY_ERROR
GROUP BY HDSD00_11_119, HDSD00_11_118
UNION ALL
SELECT COUNT(1) AS RS3,
HDSD00_12_133 AS YLNAME,
HDSD00_12_134 AS YLCODE
FROM GJZY_BASY_ERROR
GROUP BY HDSD00_12_133, HDSD00_12_134) T3
GROUP BY T3.YLNAME, T3.YLCODE)) TEMP3
ON TEMP3.YLNAME = TEMP1.YLNAME
AND TEMP3.YLCODE = TEMP1.YLCODE
WHERE TEMP3.YLCODE=‘41580781841010511A1001‘;
1.2.3 分页，分组，计数，排序
/**
* 返回数据
* FORGID 医疗机构ID
* FRCODE 县区编码
* FORGNAME 医疗机构名称
* FCENNAME_TEM_COUNT 名称不对称计数
* ISNULLCENCODE_COUNT 未对照计数
* TOTALCOUNT L_DIAITEM_ERROR表中共有多少条数据
*/
SELECT T3.*
FROM (SELECT T2.*, ROWNUM ROWNO
FROM (SELECT T.FORGID,
T1.FRCODE,
T1.FORGNAME,
SUM(NVL(T.NCCW, 0)) FCENNAME_TEM_COUNT,
SUM(NVL(T.WDZ, 0)) ISNULLCENCODE_COUNT,
TOTALCOUNT
FROM L_DIAITEM_ERROR T, TORGANIZATION T1
WHERE T.FORGID = T1.FORGID
GROUP BY T.FORGID, T1.FORGNAME, T1.FORGSEQ, T1.FRCODE
ORDER BY T1.FORGSEQ) T2
WHERE ROWNUM <= V_END) T3
WHERE T3.ROWNO >= V_START;
注意：
a.TOTALCOUNT是已存在的一个变量，这里不作为关注重点；
b.被group by的数据，要想查询表中的某个字段，有且只有两种方式：
方式一：
在group by 后面加上想要查询出来的字段
方式二：
对于数字列，使用求和函数sum()实现查询
否则，会报错：不是group by 表达式