Add more tests for modulo pushdown in PostgreSQL#12104
Add more tests for modulo pushdown in PostgreSQL#12104hashhar merged 1 commit intotrinodb:masterfrom
Conversation
hashhar
added
test
no-release-notes
There was a problem hiding this comment.
That could be BCT.
Also, what if we have % (regionkey-1) (a variable on the RHS) ?
There was a problem hiding this comment.
Good point, I'll add a test.
There was a problem hiding this comment.
Added, but kept in TestPostgreSqlConnectorTest for now since most connectors don't support connector expression pushdown yet.
There was a problem hiding this comment.
We can have correctness-only coverage in BCT.
or guard this with SUPPORTS_PREDICATE_EXPRESSION_PUSHDOWN
hashhar
force-pushed
the
hashhar/postgres-modulo-rewrite-test
branch
from
cc5d7a5 to
e1fd849
Compare
|
AC @findepi PTAL. |
|
can you please rebase? |
|
I'd prefer if @vlad-lyutenko takes over this in #14570 since that PR will be more useful to have same test run in two connectors and hence come with more general test. |
|
Rebased either way. But did not move test to BCT. Movement to BCT/BJCT should be done anyway in #14570. |
hashhar
force-pushed
the
hashhar/postgres-modulo-rewrite-test
branch
from
e1fd849 to
228e1db
Compare
Some databases return remainder instead of modulus when one of the values is negative. Since Postgres tests are used as template for other tests let's make sure we have a test covering this case even if Postgres itself has the correct behaviour. This is to ensure that any databases which calculate remainder instead can be caught in tests.
hashhar
force-pushed
the
hashhar/postgres-modulo-rewrite-test
branch
from
a43dc57 to
b14e309
Compare
2 participants
Some databases return remainder instead of modulus when one of the
values is negative. Since Postgres tests are used as template for other
tests let's make sure we have a test covering this case even if Postgres
itself has the correct behaviour. This is to ensure that any databases
which calculate remainder instead can be caught in tests.