Обсуждение: multiple SRFs in SELECT clause.

I'm getting some odd results when running two generate_series calls in a
SELECT.  When the two calls return the same number of rows you get that
many rows out:

# SELECT generate_series(1,3), generate_series(1,3);
  generate_series | generate_series
-----------------+-----------------
                1 |               1
                2 |               2
                3 |               3
(3 rows)

When the row counts differ you get the least common multiple number of
rows.

# SELECT generate_series(1,4), generate_series(1,2);
  generate_series | generate_series
-----------------+-----------------
                1 |               1
                2 |               2
                3 |               1
                4 |               2
(4 rows)

I was personally expecting a cross join between them that would be
equivalent to

# SELECT * FROM generate_series(1,4) a, generate_series(1,2) b;
  a | b
---+---
  1 | 1
  1 | 2
  2 | 1
  2 | 2
  3 | 1
  3 | 2
  4 | 1
  4 | 2
(8 rows)

Tested on 8.1.3 and CVS HEAD.

Kris Jurka

Kris Jurka <books@ejurka.com> writes:
> I'm getting some odd results when running two generate_series calls in a
> SELECT ...
> When the row counts differ you get the least common multiple number of
> rows.

Yup, this is the expected or at least historical behavior.  It's not
entirely clear what you *should* get, which is one reason we ought to
deprecate SRFs in SELECT lists.

            regards, tom lane