Thread: pgsql-server/src/test/regress sql/union.sql ex ...

pgsql-server/src/test/regress sql/union.sql ex ...

From
tgl@svr1.postgresql.org (Tom Lane)
Date:
CVSROOT:    /cvsroot
Module name:    pgsql-server
Changes by:    tgl@svr1.postgresql.org    03/11/02 18:35:26

Modified files:
    src/test/regress/sql: union.sql
    src/test/regress/expected: union.out

Log message:
    Adjust data types in some of the UNION tests to avoid potentially
    platform-dependent results, as per example from Larry Rosenman.


Re: pgsql-server/src/test/regress sql/union.sql ex ...

From
Peter Eisentraut
Date:
Tom Lane writes:

> Log message:
>     Adjust data types in some of the UNION tests to avoid potentially
>     platform-dependent results, as per example from Larry Rosenman.

The join test still fails on UnixWare:

*** ./expected/join.out Thu Sep 25 01:58:06 2003
--- ./results/join.out  Tue Nov  4 13:12:17 2003
***************
*** 1732,1739 ****
       | 6 | 6 | six   |
       | 7 | 7 | seven |
       | 8 | 8 | eight |
-      |   |   | null  |
       |   | 0 | zero  |
  (13 rows)

  SELECT '' AS "xxx", *
--- 1732,1739 ----
       | 6 | 6 | six   |
       | 7 | 7 | seven |
       | 8 | 8 | eight |
       |   | 0 | zero  |
+      |   |   | null  |
  (13 rows)

  SELECT '' AS "xxx", *
***************
*** 1752,1759 ****
       | 6 | 6 | six   |
       | 7 | 7 | seven |
       | 8 | 8 | eight |
-      |   |   | null  |
       |   | 0 | zero  |
  (13 rows)

  SELECT '' AS "xxx", *
--- 1752,1759 ----
       | 6 | 6 | six   |
       | 7 | 7 | seven |
       | 8 | 8 | eight |
       |   | 0 | zero  |
+      |   |   | null  |
  (13 rows)

  SELECT '' AS "xxx", *
***************
*** 1793,1800 ****
  -----+---+---+-------+----
       | 0 |   | zero  |
       | 1 | 4 | one   | -1
-      | 2 | 3 | two   |  2
       | 2 | 3 | two   |  4
       | 3 | 2 | three | -3
       | 4 | 1 | four  |
       | 5 | 0 | five  | -5
--- 1793,1800 ----
  -----+---+---+-------+----
       | 0 |   | zero  |
       | 1 | 4 | one   | -1
       | 2 | 3 | two   |  4
+      | 2 | 3 | two   |  2
       | 3 | 2 | three | -3
       | 4 | 1 | four  |
       | 5 | 0 | five  | -5
***************
*** 1815,1822 ****
  -----+---+---+-------+----
       | 0 |   | zero  |
       | 1 | 4 | one   | -1
-      | 2 | 3 | two   |  2
       | 2 | 3 | two   |  4
       | 3 | 2 | three | -3
       | 4 | 1 | four  |
       | 5 | 0 | five  | -5
--- 1815,1822 ----
  -----+---+---+-------+----
       | 0 |   | zero  |
       | 1 | 4 | one   | -1
       | 2 | 3 | two   |  4
+      | 2 | 3 | two   |  2
       | 3 | 2 | three | -3
       | 4 | 1 | four  |
       | 5 | 0 | five  | -5


--
Peter Eisentraut   peter_e@gmx.net


Re: pgsql-server/src/test/regress sql/union.sql ex ...

From
Tom Lane
Date:
Peter Eisentraut <peter_e@gmx.net> writes:
> Tom Lane writes:
>> Log message:
>> Adjust data types in some of the UNION tests to avoid potentially
>> platform-dependent results, as per example from Larry Rosenman.

> The join test still fails on UnixWare:

Yeah.  I intended to leave that unfixed, since row-ordering differences
are specifically documented as a legitimate platform-specific
difference.

Our alternatives would be to add enough ORDER BY clauses to constrain
the results (which I don't want to do because it would limit the set
of join plans testable by the tests), or to add another result file
for this test.  I'd be in favor of the latter if there were more than
one platform showing this behavior, or if I thought that UnixWare's
behavior wouldn't change again next time SCO decides to tweak qsort.

            regards, tom lane

Re: pgsql-server/src/test/regress sql/union.sql ex ...

From
Peter Eisentraut
Date:
Tom Lane writes:

> Our alternatives would be to add enough ORDER BY clauses to constrain
> the results (which I don't want to do because it would limit the set
> of join plans testable by the tests), or to add another result file
> for this test.  I'd be in favor of the latter if there were more than
> one platform showing this behavior, or if I thought that UnixWare's
> behavior wouldn't change again next time SCO decides to tweak qsort.

Fair enough.  I just thought, reading your commit message, that you had
set out to completely fix it.  I say we leave it as is for now.

--
Peter Eisentraut   peter_e@gmx.net