For the first rows, we know exactly the row number. But for the "random access" ones, in the middle of a big CSV file, we only guess the row number, and the estimate is most probably slightly incorrect.
We know when such a row number is an estimate, but HighTable has no concept of an estimated row number.
For the first rows, we know exactly the row number. But for the "random access" ones, in the middle of a big CSV file, we only guess the row number, and the estimate is most probably slightly incorrect.
We know when such a row number is an estimate, but HighTable has no concept of an estimated row number.