Hi!
On 09.02.2025 18:38, Alexander Korotkov wrote:
Also, aren't we too restrictive while requiring is_simple_values_sequence()?
For instance, I believe cases like this (containing Var) could be transformed too.
select * from t t1, lateral (select * from t t2 where t2.i in (values (t1.i), (1)));
I'm still working on it.
Also, I think there is quite a code duplication about construction of
SAOP between match_orclause_to_indexcol() and convert_VALUES_to_ANY()
functions. I would like to see a refactoring as a separate first
patch, which extracts the common part into a function.
Done.
I have attached a patch. In addition to the transfer, I added the process of searching for a suitable operator and type for the left expression for input expressions: const and left expression, since they may differ from the declared types. Additionally, we convert the left expr to a type suitable for the found operator.
--
Regards,
Alena Rybakina
Postgres Professional
