HippoBaro opened a new issue, #11265: URL: https://github.com/apache/arrow-rs/issues/11265
### Describe the bug The Parquet writer accepts some encoding and physical-type combinations that its encoders cannot represent correctly. Invalid configurations can therefore fail only when values are written, when dictionary encoding falls back, or when a reader encounters the output, rather than when the physical column encoder is constructed. Dictionary encoding can also hide the problem entirely: a configured fallback encoding may be invalid for the column's physical type, but writing succeeds if all values remain dictionary-encoded. Whether an unsupported configuration is accepted should not depend on the input values or whether dictionary fallback happens to occur. ### To Reproduce For example, `DELTA_LENGTH_BYTE_ARRAY` is not supported for `INT32`. Nevertheless, the writer allows this column to be constructed. With dictionary encoding enabled, a small write completes successfully without exercising the invalid fallback; with dictionary encoding disabled, the same values cause a panic inside `DeltaLengthByteArrayEncoder`. ### Expected behavior Unsupported encoding/type combinations should be rejected when the physical column encoder is constructed, before any values reach the affected encoder. This must include dictionary fallback encodings even if the column would remain dictionary-encoded for the entire write. The same validation should apply to the Arrow byte-array writer's fallback path and across both Parquet writer versions. Valid encoding/type combinations and their wire representations should remain unchanged. ### Additional context _No response_ -- 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]
