Skip to content

[SPARK-51920][SS][PYTHON] Fix type handling of namedTuple for transfromWithState #53314

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
zeruibao wants to merge 10 commits into
base: master
Choose a base branch
from
+59 −13
Open

[SPARK-51920][SS][PYTHON] Fix type handling of namedTuple for transfromWithState #53314

Changes from 1 commit
Commits
Show all changes
10 commits
Select commit Hold shift + click to select a range
1de810d
save
zeruibao Dec 3, 2025
34b6dcf
save
zeruibao Dec 3, 2025
d86956a
save
zeruibao Dec 4, 2025
d660137
save
zeruibao Dec 4, 2025
ce1191b
save
zeruibao Dec 4, 2025
0a1a21e
save
zeruibao Dec 4, 2025
22b956c
save
zeruibao Dec 4, 2025
8523efc
save
zeruibao Dec 4, 2025
9028fa8
Update helper_pandas_transform_with_state.py
zeruibao Dec 5, 2025
06fad44
save
zeruibao Dec 5, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
5 changes: 5 additions & 0 deletions python/pyspark/sql/streaming/stateful_processor_api_client.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -501,6 +501,11 @@ def normalize_value(v: Any) -> Any:
# Convert NumPy types to Python primitive types.
if isinstance(v, np.generic):
return v.tolist()
# Named tuples (collections.namedtuple or typing.NamedTuple) have a
# _fields attribute. Spark Row has __fields__. Both require positional
# arguments and cannot be instantiated with a generator expression.
if hasattr(v, '_fields') or hasattr(v, '__fields__'):
Copy link
Contributor

@gaogaotiantian gaogaotiantian Dec 3, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I think checking the type is much safer than checking an attribute, especially considering that _fields is a not a rare attribute name. If we know what type we are targeting, we should just check type.

Copy link
Contributor Author

@zeruibao zeruibao Dec 3, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

sounds good! You are so fast. Did not get a chance to add a UT yet 😛 just convert to draft.

Copy link
Contributor Author

@zeruibao zeruibao Dec 4, 2025

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Hey @gaogaotiantian, I use

                if (
                    isinstance(v, Row) or
                    (isinstance(v, tuple) and hasattr(v, "_fields"))
                ):

instead. isinstance(v, NamedTuple) won’t work because typing.NamedTuple is a class factory, not a runtime parent of instances. Checking isinstance(v, tuple) and _fields is the correct way. Please take another look. Thanks!

return type(v)(*[normalize_value(e) for e in v])
# List / tuple: recursively normalize each element
if isinstance(v, (list, tuple)):
return type(v)(normalize_value(e) for e in v)
Expand Down