Gregory Stark <stark@enterprisedb.com> writes:
> "Martijn van Oosterhout" <kleptog@svana.org> writes:
>> Will this change break my code?
> Well, DBI already has to deal with this anyways because it tries to provide a
> database-independent interface. So you can instruct DBI to upcase, downcase,
> or leave the identifiers as the database provides them by setting this
> property on your database connection:
That's not a solution, that's a kluge with very obvious failure modes.
Now admittedly it's probably not *likely* that someone would use
identifiers differing only in case in a single table definition.
But it's legal, and in fact we're required by spec to support it.
regards, tom lane