If the user can provide a CSV, and if we expect certain columns to exist in that CSV, the code should do some validation on those expectations and fail in a reasonable way if the input in invalid. One example is missing the column name.
It might be nice if the code was more tolerant on the input structure (case-insensitive on headers), perhaps supporting aliases (like sequence_name -> name), but this is less critical.
If the user can provide a CSV, and if we expect certain columns to exist in that CSV, the code should do some validation on those expectations and fail in a reasonable way if the input in invalid. One example is missing the column name.
It might be nice if the code was more tolerant on the input structure (case-insensitive on headers), perhaps supporting aliases (like sequence_name -> name), but this is less critical.