Use numpy.issubdtype for more robust integer type detection

Co-authored-by: joocer <1688479+joocer@users.noreply.github.com>

Author: Copilot

opteryx/managers/expression/__init__.py

@@ -272,7 +272,7 @@ def _inner_evaluate(root: Node, table: Table):
             if not isinstance(centre, pyarrow.Array):
                 centre = numpy.asarray(centre)
                 # Convert numeric types (e.g., uint8 from list_contains_any) to boolean
-                if centre.dtype in (numpy.uint8, numpy.int8, numpy.int32, numpy.int64, numpy.uint16, numpy.uint32, numpy.uint64):
+                if numpy.issubdtype(centre.dtype, numpy.integer):
                     centre = centre.astype(numpy.bool_)
             centre = pyarrow.array(centre, type=pyarrow.bool_())
             return pyarrow.compute.invert(centre)