Re: Collation-aware comparisons in GIN opclasses

On Tue, Sep 16, 2014 at 06:56:24PM +0400, Alexander Korotkov wrote:
> On Tue, Sep 16, 2014 at 12:14 PM, Emre Hasegeli <emre@hasegeli.com> wrote:
> 
>     > > Changing the default opclasses should work if we make
>     > > pg_dump --binary-upgrade dump the default opclasses with indexes
>     > > and exclusion constraints.  I think it makes sense to do so in
>     > > --binary-upgrade mode.  I can try to come with a patch for this.
>     >
>     > Can you explain it a bit more detail? I didn't get it.
> 
>     pg_upgrade uses pg_dump --binary-upgrade to dump the schema of
>     the old database.  Now, it generates CREATE INDEX statements without
>     explicit opclass if opclass is the default.  We can change pg_dump
>     to generate the statements with opclass even if opclass is the default
>     in --binary-upgrade mode.
> 
> 
> Thanks, I get it. I checked pg_dump implementation. It appears to be not as
> easy as it could be. pg_dump doesn't form index definition by itself. It calls
> pg_get_indexdef function. This function have no option to dump names of default
> opclasses. Since we can't change behaviour of old postgres version, we have to
> make pg_dump form index definition by itself.

Well, the server is also operating in binary-upgrade mode, so you could
have the server-side function pg_get_indexdef() behave differently for
pg_upgrade.

--  Bruce Momjian  <bruce@momjian.us>        http://momjian.us EnterpriseDB
http://enterprisedb.com
 + Everyone has their own god. +