[
https://issues.apache.org/jira/browse/SPARK-60880?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Uroš Bojanić updated SPARK-60880:
---------------------------------
Affects Version/s: 4.4.0
(was: 5.0.0)
> Extract shared CREATE TABLE analysis helpers
> --------------------------------------------
>
> Key: SPARK-60880
> URL: https://issues.apache.org/jira/browse/SPARK-60880
> Project: Spark
> Issue Type: Improvement
> Components: SQL
> Affects Versions: 4.4.0
> Reporter: Vladan Vasić
> Assignee: Vladan Vasić
> Priority: Major
> Labels: pull-request-available
>
> Separate per-table operations from analyzer rule traversal so analyzer paths
> can reuse the existing CREATE TABLE and REPLACE TABLE logic.
> Extract V1 lowering, storage construction, default-value folding,
> preprocessing, and CHAR/VARCHAR replacement into CreateTableHelper. Move V2
> CREATE/REPLACE checks into Catalyst CreateTableValidation. Share
> column-definition output construction through CatalogIdentifierResolution and
> expose per-specification resolution in ResolveTableSpec.
> Keep existing catalog/provider selection, storage and option precedence,
> constraint and generated-column validation, write distribution and ordering,
> collation behavior, and table metadata unchanged.
> Add focused tests in the existing CatalogSuite, AnalysisSuite, and
> PlanResolutionSuite. Run the existing V2 CREATE/REPLACE, create-time
> write-ordering, and CHAR/VARCHAR DDL regression cases and Scala style checks.
> Validation with JDK 21: 285 tests passed across the Catalyst
> catalog/analysis, table-plan, V2 CREATE/REPLACE, create-time write-ordering,
> and CHAR/VARCHAR DDL cases. One existing V2 case is ignored. Catalyst and SQL
> main/test scalastyle passed with zero errors or warnings; whitespace checks
> passed.
--
This message was sent by Atlassian Jira
(v10.3.26#10030026)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
