PostgreSQL 加入不同的子句

Question

我正在 PostgreSQL 中的两个表之间进行连接，一个表对事件编号 (opentickets) 有主键约束，而另一个表没有约束并且可以有重复的事件编号 (incommingtickets)。尝试过滤掉重复项时会出现问题。查询，

SELECT  incommingtickets.* 
FROM incommingtickets
    LEFT JOIN opentickets
        ON incommingtickets.incidentnumber = opentickets.incidentnumber 
WHERE opentickets.incidentnumber IS NULL
      AND incommingtickets.status NOT IN ('Closed','Cancelled', '')

一直工作到遇到重复项为止，我收到违反主键的消息。如果我添加一个不同的子句，例如，

SELECT  DISTINCT ON (incommingtickets.incidentnumber) incommingtickets.* 
FROM incommingtickets
    LEFT JOIN opentickets
        ON incommingtickets.incidentnumber = opentickets.incidentnumber 
WHERE opentickets.incidentnumber IS NULL
      AND incommingtickets.status NOT IN ('Closed','Cancelled', '')

我得到一个错误，

pg_query(): Query failed: ERROR: missing FROM-clause entry for table "incommingtickets" LINE 30: WHERE opentickets.incidentnumber = incommingtickets.incident...

Answer 1

使用 WHERE 子句过滤掉不需要的重复项，尽管我不太清楚为什么要加入“指标”(例如门票数量)。

SELECT  incommingtickets.* 
FROM incommingtickets
WHERE incommingtickets.incidentnumber not in (
        select 
        distinct 
        incidentnumber 
        FROM opentickets)
AND incommingtickets.status NOT IN ('Closed','Cancelled', '')

通过这种方式，您可以过滤掉两个表之间的重复项。

如果您想要检查或更新 opentickets 表中任何门票的门票状态，请尝试从 incommingtickets 中获取最大状态，如下所示:

WITH ticket_rows AS(
    SELECT
    rank() OVER (PARTITION BY ticket_id ORDER BY ticket_timestamp desc) as row_number,
    ticket_id,
    ticket_status,
    ticket_timestamp
    from incommingtickets
)

SELECT  incommingtickets.*, opentickets_2.* 
FROM opentickets o
    LEFT JOIN ticket_rows ON ticket_rows.ticket_id= opentickets.ticket_id AND ticket_rows.row__number=1

如果这些不是您的目标，请更好地解释您试图通过左连接实现的目标。