pub fn row_id_column_index(chunk: &DataChunk) -> Option<usize>Expand description
Locate the physical row index column in a scan-produced chunk.
Node scans append an internal node id column (<alias>._id = row offset)
as the last column of each chunk (see resolve_scan_arrow_data and
resolve_scan_data). Write operators (DELETE/SET) must read row indices
from that column; reading column 0 would use the first property value
as a row index. Falls back to None when the chunk carries no _id
column (e.g. rel table scans), letting callers keep the legacy behaviour.