Re: BUG #5978: Running postgress in a shell script fails

From: Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us>
To: "Paul Deschamps" <pdescham49(at)gmail(dot)com>
Cc: pgsql-bugs(at)postgresql(dot)org
Subject: Re: BUG #5978: Running postgress in a shell script fails
Date: 2011-04-13 21:50:16
Message-ID: 25680.1302731416@sss.pgh.pa.us
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-bugs

"Paul Deschamps" <pdescham49(at)gmail(dot)com> writes:
> PostgreSQL version: 8.4.6
> Operating system: Ubuntu 10.4
> Description: Running postgress in a shell script fails
> Details:

> When running postgres in a shell using the -c option it looks as though it
> parses the contents of the --command as command line arguments.

> Running Command:psql postgres -c"SELECT tablename FROM PG_TABLES limit 1;"
> TEST 1
> psql: warning: extra command-line argument "FROM" ignored
> psql: warning: extra command-line argument "PG_TABLES" ignored
> psql: warning: extra command-line argument "limit" ignored
> psql: warning: extra command-line argument "1;"" ignored
> psql: FATAL: role "tablename" does not exist

I think this is a bug in Ubuntu's version of getopt_long(); it does not
happen that way for me on any of the platforms I use. I get either

psql: FATAL: role "-cSELECT tablename FROM PG_TABLES limit 1;" does not exist

on platforms where getopt_long does not think it has a charter to try to
rearrange command line arguments, or

$ psql postgres -c"SELECT tablename FROM PG_TABLES limit 1;"
tablename
--------------
pg_statistic
(1 row)

on platforms where getopt_long does move the database name after the
switch+argument. What you've evidently got is a getopt_long that tries
to rearrange the command-line arguments but produces the equivalent of

psql "-cSELECT" "postgres" "tablename" "FROM" "PG_TABLES" "limit" "1;"

which is flat out wrong.

Recommendation is to use the documented command line order, which is

psql [OPTION]... [DBNAME [USERNAME]]

rather than assuming getopt_long will fix it for you.

You can try filing a bug against Ubuntu's glibc, but I dunno whether
you'll get any satisfaction there. I think we've heard of this before
and the misbehavior is of long standing.

regards, tom lane

In response to

Browse pgsql-bugs by date

  From Date Subject
Next Message Kevin Grittner 2011-04-13 22:47:36 Re: BUG #5974: UNION construct type cast gives poor error message
Previous Message Tom Lane 2011-04-13 21:32:49 Re: BUG #5974: UNION construct type cast gives poor error message