Re: Better support for whole-row operations and composite

Andrew Dunstan <andrew(at)dunslane(dot)net> writes:
> Tom Lane wrote:
>> Only named composite types, not RECORD, will be allowed to be used as
>> table column types.

> Interesting. I'm slightly curious to know if there's an external driver
> for this.

There's noplace to store a permanent record of an anonymous rowtype's
structure. To do otherwise would amount to executing an implicit CREATE
TYPE AS for the user, so we might as well just say up front that you
have to create the type.

> Will this apply recursively (an a has a b which has an array of c's)?

Yup.

> Are there indexing implications? Could one index on a subfield?

Using an expression index, sure. I don't think we need to support it as
a "primitive" index type.

regards, tom lane