Max Gekk created SPARK-57559:
--------------------------------
Summary: Add a TimeType case to PhysicalDataType
Key: SPARK-57559
URL: https://issues.apache.org/jira/browse/SPARK-57559
Project: Spark
Issue Type: Sub-task
Components: SQL
Affects Versions: 4.3.0
Reporter: Max Gekk
h2. What
Add an explicit {{TimeType}} branch to {{PhysicalDataType.apply}} so TIME maps
to
{{PhysicalLongType}} instead of falling through to
{{UninitializedPhysicalType}}.
h2. Why
{{PhysicalDataType.apply}} has no {{TimeType}} case. Ordering/comparison
currently work via
{{AtomicType}} + {{OrderUtils}} and the {{Long}} storage in column vectors, but
the missing
case is a latent inconsistency relative to TIMESTAMP/DATE and may surface in
physical-type
driven code paths.
h2. Scope
* Add {{case _: TimeType => PhysicalLongType}} (or equivalent) in
{{PhysicalDataType}}.
* Audit call sites that switch on {{PhysicalDataType}} to confirm TIME is
handled.
h2. Acceptance criteria
* {{PhysicalDataType(TimeType(p))}} returns the long physical type for all p.
* No behavior regressions; covered by a unit test.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
