列名在 UPSERT 操作中不明确

Question

我正在尝试运行 postgres 数据库上的 UPSERT 查询字符串 -

insert into business_data_test 
(select * from business_data_tmp) 
on conflict(phone_number) 
do update set 
average_expense=(average_expense*expense_count + excluded.average_expense*excluded.expense_count)/(expense_count + excluded.expense_count), expense_count=(expense_count + excluded.expense_count);

如果存在数据冲突，我基本上会尝试更新列 average_expense，但我认为查询有问题，因为我运行遇到以下错误 -

ERROR:  column reference "average_expense" is ambiguous
LINE 1: ...lict(phone_number) do update set average_expense=(average_ex...
                                                             ^
SQL state: 42702
Character: 123

我认为我们必须在某处添加一些 table 名称别名，但我不确定如何解决这个问题。

Answer 1

您的查询已“排除”。但没有提及“排除”是什么？

我的建议是，使用单个列值而不是冗长的计算逻辑来进行更新。确保你让它工作然后添加到该更新逻辑....

这样更容易调试。

Answer 2

您需要完全限定对旧值（即“未排除的”值）的引用。如果您为目标使用别名，这会更容易一些 table

insert into business_data_test as tst
select * 
from business_data_tmp 
on conflict(phone_number) 
do update 
 set average_expense = (tst.average_expense * tst.expense_count + excluded.average_expense * excluded.expense_count)/(tst.expense_count + excluded.expense_count), 
     expense_count = tst.expense_count + excluded.expense_count;

列名在 UPSERT 操作中不明确

Column name is ambiguous on UPSERT action

sql

postgresql

upsert