PSQL：ROW_NUMBER 连续递增

Question

您好，有以下表格 T1：

 field1 | field3
--------+--------
 A1     | foo
 A2     | v1
 A3     | v2
 A4     | bar

和 T2：

 field2 | field3 
--------+--------
 B1     | foo 
 B2     | bar

如果我执行以下请求：

SELECT DISTINCT ON (T2.field2, T2.field3)
  T2.field2 AS F2,
  T2.field3 AS F3,
  ROW_NUMBER () OVER (ORDER BY T2.field3) AS F4
FROM T2
JOIN T1 ON T2.field3=T1.field3

...我得到以下结果：

F2:B1, F3:foo, F4:1
F2:B2, F3:bar, F4:4 // I would like F4:2

但我希望 F4 一个一个递增...我认为这是因为与 T1 的连接，但我不知道如何隔离 ROW_NUMBER...

Answer 1

有效

SELECT DISTINCT ON (T2.field2, T2.field3)
  T2.field2 AS F2,
  T2.field3 AS F3,
  ROW_NUMBER () OVER (ORDER BY T2.field3 DESC) AS F4
FROM T2
JOIN T1 ON T2.field3=T1.field3;

也有效和 DENSE_RANK() 不会重复计算

在第 302 行，我回顾了为什么我使用 DENSE_RANK 而不是 ROW_NUMBER 或 RANK https://github.com/pavankat/fantasy-football/blob/master/db/queries.sql

SELECT DISTINCT ON (T2.field2, T2.field3)
  T2.field2 AS F2,
  T2.field3 AS F3,
  DENSE_RANK() OVER (ORDER BY T2.field3 DESC) AS F4
FROM T2
JOIN T1 ON T2.field3=T1.field3;

PSQL：ROW_NUMBER 连续递增

PSQL: ROW_NUMBER incremented continuously

row-number

psql