Sql 如何实现合并两个表的联接—完全外部联接并避免重复？

我正在尝试将

表A

和

表B

加入一个新表。我希望避免重复，但我希望保留表中的所有记录

我已经尝试了完全外部连接，左外部连接，但我无法达到预期的结果

TABLE A
======================
CLIENT_ID   SUPPLIER_NO  INVOICE_COUNT VOLUME
ABC         1234         10            456
GEF             4567         20        657      
ERT         8912         30        567
THE         6872         42        781

TABLE B
========================
CLIENT_ID   SUPPLIER_NO REJECTED_COUNT  REJECTED_VOLUME
ERT     8912        56          102
HJI     7823        34          781
REW     6721        45          632

FINAL TABLE(WHAT IT SHOULD LOOK LIKE)
========================
CLIENT_ID SUPPLIER_NO INVOICE_COUNT VOLUME REJECTED_COUNT REJECTED_VOLUME
ABC   1234        10         456   NULL or 0      NULL or 0 
GEF   4567        20         657   NULL or 0      NULL or 0     
ERT   8912        30         567   56         102
THE   6872        42         781   NULL or 0      NULL or 0     
HJI   7823        NULL       NULL  34         781
REW   6721        NULL       NULL  45         632

我的问题是

select   A.client_id
      ,  A.supplier_no
      ,  count(*)        as Rejected_Count 
      , sum(TOTAL_AMT)   as Rejected_Volume
      , count(*)         as InvoiceCount 
      , sum(INVOICE_AMT) as Volume  
from TABLEA A 
     FULL OUTER JOIN TABLEB B 
         ON A.client_id=B.client_id 
        AND A.SUPPLIER_NO=B.SUPPLIER_NO
group by  A.client_id,  A.supplier_no
order by A.client_id,  A.supplier_no;

无论我是否与

客户id

和

供应商编号

进行连接，我都无法实现此输出。如果遍历后

客户端id

不在表A中，我希望表B中的内容避免重复。表A只是查询的结果，表B是相同的，我想合并这些结果。

将您的连接修复为：

FULL OUTER JOIN 
    TABLEB B ON A.client_id = B.client_id 
             AND A.SUPPLIER_NO = B.SUPPLIER_NO

现在，它正在从同一个表连接到同一列上

将您的连接修复为：

FULL OUTER JOIN 
    TABLEB B ON A.client_id = B.client_id 
             AND A.SUPPLIER_NO = B.SUPPLIER_NO

---

到目前为止，它是在同一个表的同一列上联接的

我认为可以使用

UNION ALL

和

GROUP BY

实现这一点，如下所示：

SELECT CLIENT_ID, SUPPLIER_NO,
MAX(INVOICE_COUNT) AS INVOICE_COUNT,
MAX(VOLUME) AS VOLUME,
MAX(REJECTED_COUNT) AS REJECTED_COUNT,
MAX(REJECTED_VOLUME) AS REJECTED_VOLUME
FROM
(SELECT CLIENT_ID, SUPPLIER_NO, INVOICE_COUNT, VOLUME, NULL AS REJECTED_COUNT, NULL REJECTED_VOLUME
FROM TABLEA
UNION ALL
SELECT CLIENT_ID, SUPPLIER_NO, NULL AS INVOICE_COUNT, NULL AS VOLUME, REJECTED_COUNT, REJECTED_VOLUME
FROM TABLEB)
GROUP BY CLIENT_ID, SUPPLIER_NO;

根据您的要求使用聚合函数。我使用了

max

---

干杯

我认为这可以通过使用

UNION ALL

和

GROUP BY

实现，如下所示：

SELECT CLIENT_ID, SUPPLIER_NO,
MAX(INVOICE_COUNT) AS INVOICE_COUNT,
MAX(VOLUME) AS VOLUME,
MAX(REJECTED_COUNT) AS REJECTED_COUNT,
MAX(REJECTED_VOLUME) AS REJECTED_VOLUME
FROM
(SELECT CLIENT_ID, SUPPLIER_NO, INVOICE_COUNT, VOLUME, NULL AS REJECTED_COUNT, NULL REJECTED_VOLUME
FROM TABLEA
UNION ALL
SELECT CLIENT_ID, SUPPLIER_NO, NULL AS INVOICE_COUNT, NULL AS VOLUME, REJECTED_COUNT, REJECTED_VOLUME
FROM TABLEB)
GROUP BY CLIENT_ID, SUPPLIER_NO;

根据您的要求使用聚合函数。我使用了

max

---

干杯

我认为

使用
子句是最简单的解决方案：

select client_id, supplier_no,
       count(*)        as Rejected_Count,
       sum(TOTAL_AMT)   as Rejected_Volume,
       count(*)         as InvoiceCount ,
       sum(INVOICE_AMT) as Volume  
from TABLEA A full join
     TABLEB B 
     using (client_id, SUPPLIER_NO)
group by client_id, supplier_no
order by client_id, supplier_no;

另一种选择是慷慨地帮助
coalesce（）
：

我认为
使用
子句是最简单的解决方案：

select client_id, supplier_no,
       count(*)        as Rejected_Count,
       sum(TOTAL_AMT)   as Rejected_Volume,
       count(*)         as InvoiceCount ,
       sum(INVOICE_AMT) as Volume  
from TABLEA A full join
     TABLEB B 
     using (client_id, SUPPLIER_NO)
group by client_id, supplier_no
order by client_id, supplier_no;

另一种选择是慷慨地帮助
coalesce（）
：

PL/SQL表明您的数据库是Oracle，不是吗？@BarbarosÖzhan YesTypo alert:拼写为Achive-而不是“acheive”…SQL正在尝试sum（）在两个联接表中都不存在的列上。@GJones这两个表表A和表B来自两个单独的查询，我试图进行完全的外部联接以获得所需的输出pl/SQL表明您的数据库是Oracle，不是吗？@BarbarosÖzhan YesTypo alert:拼写为Achive-不是“Achive”…SQL正在尝试sum（）在两个联接表中都不存在的列上。@GJones这两个表表A和表B来自两个单独的查询，我试图进行完全的外部联接以获得所需的输出。我已修复join我已修复join我使用Union代替Union All以避免重复。谢谢你，我用工会代替工会来避免重复。谢谢