Le mardi 26 février 2008, Tom Lane a écrit :
>
> In short, what I think we need here is just some more conveniently
> defined extraction filter switches than --schema-only and --data-only.
> There's no need for any fundamental change to pg_dump's architecture.
Forgive me if what I will say bellow is completly pointless, but I think at
this point that the base of this discussion might be wrong. If the decision
made here is to keep pg_dump simple, then maybe that it could be left as it
is, and create another tool just to extract some parts of a database, either
schema or data.
As far as I understand what is said here, pg_dump is thought to be a tool used
to make a backup of a database to use it somewhere else. So let it be as it
is.
What I intendeed to mean in my first post, is that it would be great to have a
tool that could let one get a partial dump of a database at one time, so as
to modify (or not) and to alter the database afterward (or not).
I use to work on many databases at a time, and sometime, I have to quickly fix
a function, add a trigger to a table... Sometime I create a sql file and save
my work before passing the command set to psql, but sometimes I don't have
much time and type in the code directly in psql.
So far so good, the code works, until a problem is found, and then, I don't
have any source file to work on...unless I use pg_dump and search the so big
file for the code I want to modify.
I hope you see what I mean. Since the idea whas to dump informations about the
structure of a table, function type or whatever object one could want from
the base, I asked for options for pg_dump, but maybe a new tool (based on
pg_dump ?) could satisfy everyone ?