获取表别名以显示为查询结果中列名的一部分

Question

In Postgres, a query like this:

SELECT a.id, b.id FROM table1 a INNER JOIN table2 b on a.id=b.id;

Will return 'deconflicted' column names like:

a.id, b.id

However, in Oracle, the id columns just keep their original names:

id, id

My problem here is that my follow-on tools are confused by the duplicate column names.

In this example, I've only listed 2 columns, but in reality there are 100's, and I'd prefer to not have to list them all out.

Is there a way to do this without having to specify all of the column names (there are 100's).

Answer 1

Try this:

SELECT a.id as 'table1 a', b.id as 'table2 b' FROM table1 a INNER JOIN table2 b on a.id=b.id group by a.id, b.id;

You can name your results with as followed by the name. If you want to use JOIN , you have to use GROUP BY followed by your selected items if you don't want duplicate values.

Answer 2

You can solve the particular problem with id by fixing the join :

SELECT *
FROM table1 a INNER JOIN
     table2 b 
     USING (id);

If id is the only duplicate column, then this may solve your problem.

Answer 3

I suspect there won't be because table names (or aliases) aren't a requirement. For example the following is valid SQL for Oracle, but there's no viable prefix to include in column naming.

SELECT * FROM 
   (select 1 a, 2 b from dual) 
  join 
   (select 1 a, 3 b from dual) 
  using (a)

But I think Postgres requires those FROM subqueries to be aliased