Re: Declarative partitioning - another take

On 2017/01/31 6:42, Peter Eisentraut wrote:
> On 1/25/17 12:54 AM, Ashutosh Bapat wrote:
>> The documentation available at
>> https://www.postgresql.org/docs/devel/static/sql-createtable.html,
>> does not make it clear that the lower bound of a range partition is
>> always inclusive and the higher one is exclusive. I think a note in
>> section " PARTITION OF parent_table FOR VALUES partition_bound_spec"
>> would be helpful.
>
> Hmm. I see the practical use of that, but I think this is going to be a
> source of endless confusion. Can we make that a bit clearer in the
> syntax, for example by using additional keywords (INCLUSIVE/EXCLUSIVE)?

The decision not to make that configurable with INCLUSIVE/EXCLUSIVE syntax
was deliberate. To summarize, we can start with a default configuration
catering to most practical cases (that is, inclusive lower and exclusive
upper bounds) and documenting so (not done yet, which I will post a doc
patch today for). If it turns out that there is some demand for making
that configurable, we can later add the code to handle that internally
plus the syntax. But *starting* with that syntax means we have to
potentially needlessly carry the code to handle seldom used cases that
could not be made as efficient as it is now with all lower bounds being
inclusive and upper bounds exclusive.

Thanks,
Amit

[1]
https://www.postgresql.org/message-id/CA%2BTgmoZou4ApEvC_nfhOxsi5G4SoD_evwNaiYn60ZcJ4XB_-QQ%40mail.gmail.com