Is there a way to tell pg_dump to just dump the SQL
statements required to create the tables, sequences,
indeces, keys, &c.? I DON'T need to restore or
recreate things like users, or most other kinds of DB
objects. Just routine DDL statements. Looking
through a dump file for a small database, it looks
like pg_dump is serious overkill, dumping a lot of
stuff I don't need (since I'm just using defaults for
them anyway).
I am developing a new DB app, to be deployed on a web
based host on the other side of the planet. There is,
at present, no 'data', and the only information to be
transferred consists of the various tables, indeces,
&c. I am creating.
Obviously, we don't want to put any of my test data on
a server that will in due course be the production
host, when the app goes live (so once my colleagues on
the other side of the planet have had a chance to play
with what I've developed, we'll mostly empty the DB of
test data, except for a small amount of data we've
obtained). I expect that a few tiny csv files I have
here will be ftped to the host and we'd use a simple
script to bulk load that. Another fly in the ointment
is that the hosting company is still using v 8.1.9 and
I am using 8.2.5 on my machine, so I am concerned that
a regular dump and restore may be problematic: it
hasn't worked so far, but then I've spent much of my
time so far wrestling with phppgadmin. :-(
I'm just looking for something that will save me a
little time. I've created the core of the DB already
on my development machine, using pgAdmin, but I can
recreate it in about a day using Emacs to create a SQL
script that preproduces what I did in pgAdmin.
Any information would be appreciated.
Thanks
Ted
