Mark Thornton wrote:
Sai Pullabhotla wrote:
Rick,
Thanks for the advice on using a CHECK constraint. It surely helps
ensuring
the data integrity.
Otherwise, I think this is very basic feature that every one would
like to
see, and it would help if Derby can handle this automatically using the
Computed/Generated columns or allowing UNIQUE index creation using
UPPER/LOWER.
Another alternative would be to allow a collation to be specified when
creating an index.
Mark Thornton
Something like this:
create unique index roadnamesindex on roadnames(name collate
'en_UK_primary', featureid)
(Borrowed from an in house database which supports that statement.)
Mark Thornton
