xinrong-databricks commented on a change in pull request #33911:
URL: https://github.com/apache/spark/pull/33911#discussion_r704796337
##########
File path: python/pyspark/pandas/data_type_ops/boolean_ops.py
##########
@@ -248,6 +249,30 @@ def and_func(left: Column, right: Any) -> Column:
return column_op(and_func)(left, right)
+ def __xor__(self, left: IndexOpsLike, right: Any) -> SeriesOrIndex:
+ _sanitize_list_like(right)
+ if isinstance(right, IndexOpsMixin) and isinstance(right.dtype,
extension_dtypes):
+ return right.__xor__(left)
+ elif (
+ isinstance(right, int)
+ or (hasattr(right, "dtype") and np.issubdtype(right.dtype, int))
+ or isinstance(right, bool)
+ or (hasattr(right, "dtype") and right.dtype == "bool")
Review comment:
How do you think about
```py
def _is_valid_for_logical_operator(right: Any) -> bool:
from pyspark.pandas.base import IndexOpsMixin
return (
isinstance(right, IndexOpsMixin)
and (
isinstance(right.spark.data_type, BooleanType)
or isinstance(right.spark.data_type, IntegralType)
)
or isinstance(right, (int, bool))
)
```
IndexOpsMixin provides a unified interface for Series and Index.
DataTypeOps aims to make operations of Series and Index data-type-based.
We may access the DataTypeOps from the property `_dtype_op` of IndexOpsMixin.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
