Dejan Krakovic created SPARK-61301:
--------------------------------------
Summary: Improve the FILE type error messages and their builders
Key: SPARK-61301
URL: https://issues.apache.org/jira/browse/SPARK-61301
Project: Spark
Issue Type: Sub-task
Components: SQL
Affects Versions: 5.0.0
Reporter: Dejan Krakovic
Follow-up of SPARK-60386 (https://github.com/apache/spark/pull/59640).
Improve the FILE type errors:
- FILE_TYPE_MISSING_QUALIFIER has no column name or query context, so on a wide
table the user cannot tell which column to fix. Add the column path.
- FILE_TYPE_UNKNOWN_WRITE reports a nested path only for schema evolution, and
the top-level column for CTAS and CREATE TABLE LIKE.
- Review the wording and SQLSTATEs of the FILE error conditions, and where
their builders live.
- Three public TypeUtils helpers throw FILE_TYPE_UNKNOWN_WRITE and differ only
by input type; fold them into one.
- A FILE bucket sort column gets _LEGACY_ERROR_TEMP_3072 on V1 tables but
FILE_TYPE_UNSUPPORTED_COLUMN_USAGE.PARTITION_COLUMN on V2.
- Simplifications: the usage string is spliced into the
FILE_TYPE_UNSUPPORTED_COLUMN_USAGE error class name, so a typo at a call site
is not caught at compile time.
Review threads:
https://github.com/apache/spark/pull/59640#discussion_r4353281828,
https://github.com/apache/spark/pull/59640#discussion_r4353281850,
https://github.com/apache/spark/pull/59640#discussion_r4382011871
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
