Hyukjin Kwon created SPARK-60723:
------------------------------------
Summary: Settle version 1 of the native data source interface
before 4.4.0
Key: SPARK-60723
URL: https://issues.apache.org/jira/browse/SPARK-60723
Project: Spark
Issue Type: Improvement
Components: SQL
Affects Versions: 5.0.0
Reporter: Hyukjin Kwon
A follow-up to SPARK-60101 (https://github.com/apache/spark/pull/59424).
Version 1 of spark_datasource.h first ships in 4.4.0, so the changes from the
review in https://github.com/apache/spark/pull/59424#issuecomment-6480975990
that change existing functions or the meaning of a flag have to land before
then:
- read and create_data_writer get size-prefixed SparkDsReadInfo and
SparkDsDataWriterInfo structs with the schema, and read also gets whether the
read is batch or streaming.
- SPARK_DS_TRUNCATE is split into SPARK_DS_BATCH_TRUNCATE and
SPARK_DS_STREAMING_TRUNCATE.
- partitions, stream_partitions and create_writer return the state of the
reader or the writer, replacing serialize_reader, serialize_stream_reader and
serialize_writer.
- create_stream_reader and latest_offset take size-prefixed structs, with the
checkpoint location and the start offset, so that admission control can be
added later.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
