pg_dump options and perhaps 'Enhancement request'

On Thu, May 08, 2003 at 10:11:48PM +0200, Terry Yapt wrote:

1-) Are there any way to do a pg_dump without '--' COMMENTS ? For
example, I don't like to have this in my output file:
--
-- TOC entry 19 (OID 26027)
-- Name: TABLE news; Type: COMMENT; Schema: www; Owner: www
--

You could do this with grep (and something similar for blank lines):

pg_dump [options] dbname | grep -v "^--" > dumpfile

A

-- 
----
Andrew Sullivan                         204-4141 Yonge Street
Liberty RMS                           Toronto, Ontario Canada
<andrew@libertyrms.info>                              M2P 2A8
                                         +1 416 646 3304 x110

On Thu, 2003-05-08 at 13:11, Terry Yapt wrote:

I don't know if I can do this stuffs. If not I would like to get it in
the future. Thanks.

1-) Are there any way to do a pg_dump without '--' COMMENTS ? For
example, I don't like to have this in my output file:
--
-- TOC entry 19 (OID 26027)
-- Name: TABLE news; Type: COMMENT; Schema: www; Owner: www
--

2-) The same for the 'blank' lines. Are there a lot of in the output
file.

The plain-text pg_dump output is just that and you can use various
command line tools to strip out stuff you don't like. A simple solution
for most cases:

$ pg_dump [args] | egrep -v '^--|^$' >backup.sql

HOWEVER, if you have the diabolical case of a table whose first column
is a text type AND a row in which that first column's text begins with
'--', then you'll strip those too. I think you're safe otherwise.
(newlines are translated as '\n').

3-) Are there any way to get the output file with schemas in all
table_names ?
I would like to get this:
COMMENT ON TABLE www.fo_temas IS 'xxxx.'; -- instead of:
COMMENT ON TABLE fo_temas IS 'xxxx.';

This is fairly easy with a simple perl script.
--
Reece Hart, Ph.D. rkh@gene.com, http://www.gene.com/
Genentech, Inc. 650/225-6133 (voice), -5389 (fax)
Bioinformatics and Protein Engineering
1 DNA Way, MS-93 http://www.in-machina.com/~reece/
South San Francisco, CA 94080-4990 reece@in-machina.com, GPG: 0x25EC91A0

Ok.. ok...

I know I can do a lot of things with external software or utilities (even with sed, vi or others). I can develop my own software tool to do what I want to do too. But, I think PostgreSQL team is porting this great DB to Win32 now. Ok, I can tell you the windows guys don't have so much utilities like linux/unix guys or worse perhaps they will be users only and don't know how to use perl or something similar.

I only want a 'clear' pg_dump if it is possible and with SCHEMA.object_name ala ORACLE if it is possible too.

Thanks a lot four your answers...

Andrew Sullivan wrote:

This guy raises a good point, albeit indirectly. The transitions from one major version to another need to be smoother if Postgres is going to gain any more acceptance in big database circles.

A dump file shouldn't have to be messed with to reload it into the next version. There should be a 'conversion tool' or script between each version. And if a customer waits 8 versions to update, those tools, used in sequence, should be able to change his old dump to be imported into the latest version.

Terry Yapt wrote: