Re: Making regression tests

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: "Christopher Kings-Lynne" <chriskl(at)familyhealth(dot)com(dot)au>
Cc: "Hackers" <pgsql-hackers(at)postgresql(dot)org>
Subject: Re: Making regression tests
Date: 2001-10-17 05:25:36
Message-ID: 25414.1003296336@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-hackers

"Christopher Kings-Lynne" <chriskl(at)familyhealth(dot)com(dot)au> writes:
> Just wondering if someone could give me an indication as to how to create a
> new regression test, for the ADD UNIQUE stuff I did?

It's not hard. Make a file in src/test/regress/sql containing test
queries, and one in src/test/regress/sql/expected containing the
expected output (which is just the *actual* output, after you've
checked it over). Add the test name to src/test/regress/serial_schedule
and src/test/regress/parallel_schedule.

If your test needs to cater for platform-dependent source or output
then things are a little harder, but I don't see why an ADD UNIQUE
test would need that.

Actually, I don't see why you wouldn't just add some more test queries
to src/test/regress/sql/alter_table.sql and corresponding output to
src/test/regress/expected/alter_table.out. But if you really want
a separate test file, the two _schedule files are the places that
need to know about it.

regards, tom lane

In response to

Browse pgsql-hackers by date

  From Date Subject
Next Message Denis A Ustimenko 2001-10-17 05:55:05 compiling libpq++ on Solaris with Sun SPRO6U2
Previous Message Christopher Kings-Lynne 2001-10-17 03:28:17 Re: Replication