Bruce Momjian wrote:
>
> Attached is a small patch to SQLStatistics() to allow proper information
> about CLUSTER and HASH indexes, as proposed by the ODBC code. Right now
> the CLUSTER information is not maintained, but the HASH is accurate.
> Also, I found that the tests for UNIQUE and CLUSTER were done using
> atoi(). I don't think that works because these queries return 't' or
> 'f', not 1 or 0.
>
Psqlodbc driver automatically coverts t/f to 1/0.
regards,
Hiroshi Inoue