[Bug] NestedRow.getVector still throws "Unsupported type: VectorType" for nested VECTOR columns

[thswlsqls]

Search before asking

• I searched in the issues and found nothing similar.

Paimon version
master @ bbae07d

Compute Engine
Any engine; the Java API reproduces it.

Minimal reproduce step

1. Create a primary-key table with a column r ROW<id INT, v VECTOR(3, FLOAT)>.
2. Write one row with a non-null r.v and call prepareCommit.

The flush fails with:

java.lang.IllegalArgumentException: Unsupported type: VectorType
    at org.apache.paimon.data.NestedRow.getVector(NestedRow.java:311)
    at org.apache.paimon.format.parquet.writer.ParquetRowDataWriter$ArrayWriter.write(ParquetRowDataWriter.java:523)
    at org.apache.paimon.format.parquet.writer.ParquetRowDataWriter$RowWriter.write(ParquetRowDataWriter.java:609)

The Avro writer fails the same way. The write buffer holds the row as BinaryRow, so the format writer reads the nested row through BinaryRow.getRow() → NestedRow, whose getVector(int) is still the stub from #7204. BinaryWriter.writeVector already writes the nested slot; only the read accessor is missing.

What doesn't meet your expectations?
A nested VECTOR should be written, committed and read back like a top-level VECTOR (BinaryRow.getVector) and ARRAY<VECTOR> (BinaryArray.getVector). #7204 added the stub to both BinaryArray and NestedRow; #9773 fixed BinaryArray; NestedRow is the last one.

Anything else?
N/A

Are you willing to submit a PR?

• I'm willing to submit a PR!