SEZ9 commented on PR #11222: URL: https://github.com/apache/seatunnel/pull/11222#issuecomment-6818356387
@MyeoungDev thanks for the update and for walking through the changes. - **Vector copy (full capacity, position 0):** Agreed on the approach. Since `VectorUtils.toFloatArray` reads by capacity, preserving the source position would just shift the underflow problem; a fresh heap buffer with the whole capacity and position 0 is what downstream consumers expect. Glad to hear the moved-position, direct-buffer and byte-order cases now have unit tests and that the vector E2E checks the copy via `VECTOR_DIMS` and `L2_DISTANCE`. - **`multiple_row` removal:** No need to bring it back. If it can't be declared as a column type and multi-table Copy already works without it, dropping the branch is cleaner than keeping dead code. - **Docs, `instanceof` guards, Assert comment, indentation:** Noted, thanks. I'll go through the latest changes and verify each of the earlier points against the diff, in particular that the Copy doc body describes the deep-copy semantics for bytes/vectors and that the `instanceof` fallback raises the structured `CommonError` with field context rather than a bare `ClassCastException`. Will follow up once I've done that pass. <!-- streview-comment:2296 --> -- 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]
