tartarus created FLINK-23715:
--------------------------------
Summary: Support for reading fields that do not exist in Parquet
files
Key: FLINK-23715
URL: https://issues.apache.org/jira/browse/FLINK-23715
Project: Flink
Issue Type: Improvement
Components: Formats (JSON, Avro, Parquet, ORC, SequenceFile)
Reporter: tartarus
In a production environment, it is often encountered that users add fields to
the hive table, but do not refresh the data of the historical partition.
Therefore, if the new field is not in the historical partition file, an error
will be reported when reading the historical partition.
General users would expect that if there is no such field, then fill in null
and return.
The current flink Parquet format does not support this function.
--
This message was sent by Atlassian Jira
(v8.3.4#803005)
