Excerpts from Peter Eisentraut's message of vie may 28 12:27:52 -0400 2010:
> Option 2, invent some new mechanism that accompanies a datum or a type
> whereever it goes. Kind of like typmod, but not really. Then the
> collation information would presumably be made available to functions
> through the fmgr interface. The binary representation of data values
> stays the same.
Is the collation a property of the datum, or one of the comparison?
If the latter, should it be really be made a sidecar of a datum, or
would it make more sense to attach it to the operation being performed?
I wonder if instead of trying to pass it down multiple layers till
bttextcmp and further down, it would make more sense to set a global
variable somewhere in the high levels, and only have it checked in
varstr_cmp.
--
Álvaro Herrera <alvherre@commandprompt.com>
The PostgreSQL Company - Command Prompt, Inc.
PostgreSQL Replication, Consulting, Custom Development, 24x7 support