Hyukjin Kwon created SPARK-60986:
------------------------------------
Summary: Take datetime.timedelta values in lit() and
createDataFrame in Spark Connect Rust as PySpark does
Key: SPARK-60986
URL: https://issues.apache.org/jira/browse/SPARK-60986
Project: Spark
Issue Type: Bug
Components: Connect
Affects Versions: connect-rust-4.3.0
Reporter: Hyukjin Kwon
PySpark takes a {{datetime.timedelta}} as a value of {{DayTimeIntervalType}}
(INTERVAL DAY TO SECOND), of {{floor(total_seconds())}} seconds and its
microseconds: {{F.lit(datetime.timedelta(days=1, seconds=2))}} is such a
literal, and {{createDataFrame([(datetime.timedelta(seconds=1),)])}} infers
{{interval day to second}}. The Rust-backed Python client raises
{{UNSUPPORTED_DATA_TYPE}} for the literal and {{TypeError: unsupported value
type for createDataFrame: timedelta}} for the rows. Take a {{timedelta}} as
PySpark does.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
