Uroš Bojanić created SPARK-58931:
------------------------------------
Summary: Reject negative randstr length during analysis
Key: SPARK-58931
URL: https://issues.apache.org/jira/browse/SPARK-58931
Project: Spark
Issue Type: Improvement
Components: SQL
Affects Versions: 5.0.0
Reporter: Uroš Bojanić
`randstr(length[, seed])` requires a non-negative `length`, but that guard
currently lives only on the execution paths (`RandStr.lengthInteger()`, called
from interpreted eval and codegen), so a negative constant `length` is not
rejected until the query executes.
This moves the check into `RandStr.checkInputDataTypes()` so it is enforced
during Catalyst analysis, reusing the existing `INVALID_PARAMETER_VALUE.LENGTH`
error. As a result, `randstr` with a negative constant `length` now fails
during analysis instead of at execution. The error itself is unchanged (same
`SparkRuntimeException` and error class); only the phase at which it is raised
changes. Queries with a valid (non-negative) `length` are unaffected.
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
