Re: Removing pg_migrator limitations

Bruce Momjian wrote:
> Bruce Momjian wrote:
> > I looked at DefineEnum() and basically adding the ability to add enums
> > would put the new enum after the existing ones unless the OID counter
> > has wrapped around and is less than the oid counter at the time the enum
> > type was created, in which case it will be listed as before the existing
> > values.  I wasn't aware enum ordering is something we tried to maintain.
> > One issue is that we are not supporting the addition of enum values even
> > for people who don't care about the ordering of enums (which I bet might
> > be the majority.)
> > 
> > I can think of a few approaches for pg_migrator:
> > 
> >     1)  Create an oid array in a permanent memory context and have
> >         DefineEnum() read from that.
> >     2)  Renumber the enum entries after they are created but before
> >         any of their oids are stored in user tables.
> > 
> > Both can be done by pg_dump with proper server-side functions.  The
> > problem with #2 are cases where the old and new oid ranges overlap,
> > e.g.:
> 
> I now think the easiest solution will be to have pg_dump create the enum
> with a single dummy value, delete the pg_enum dummy row, and then call a
> modified version of EnumValuesCreate() to insert row by row into
> pg_enum, with specified oids.

I thought of a cleaner approach.  CREATE TYPE ENUM will create one enum
with the specified oid, and then a server-side function will call
EnumValuesCreate() be used to add each additional enum with a specified
oid --- no deleting necessary.  I will start working on a patch for
this.

--  Bruce Momjian  <bruce@momjian.us>        http://momjian.us EnterpriseDB
http://enterprisedb.com
 + If your life is a hard drive, Christ can be your backup. +