zhengruifeng opened a new pull request, #55604:
URL: https://github.com/apache/spark/pull/55604
### What changes were proposed in this pull request?
Pin `spark.sql.analyzer.strictDataFrameColumnResolution=true` around the
body of the `lazy column validation` test in `DataFrameSuite`. The config is
set via `spark.conf.set/unset` rather than `withSQLConf` because the lazy
SQLConf entry trips `withSQLConf`'s `isModifiable` check on the Connect server.
### Why are the changes needed?
The test asserts that `df4.schema` throws `AnalysisException` for `df1("x")`
when `df1` does not contain `x`. This holds only under strict plan-id-based
resolution; if the name-based fallback path is enabled, `df1("x")` resolves to
`df2.x` from the join output and `df4.schema` succeeds. Today this works
because `STRICT_DATAFRAME_COLUMN_RESOLUTION` defaults to `true`, but the test
should not silently rely on that default; pinning it makes the assumption
explicit and keeps the test robust against future default changes or
environments where the default is overridden.
### Does this PR introduce _any_ user-facing change?
No. Test-only change.
### How was this patch tested?
Existing `DataFrameSuite."lazy column validation"`.
### Was this patch authored or co-authored using generative AI tooling?
Generated-by: Claude Code (Anthropic), claude-opus-4-7
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
