Pull just the columns you need out of a CSV

The job

Vendor exports dump everything. You need column 1, 4, and 7 for a report — and naïve splitting on commas corrupts the row whose address field contains 'Apt 4, Suite B'.

Extract them

Paste the delimited text, type a selector like 1,4-6 in keep mode (or drop mode to remove instead), and run. The shared CSV parser respects quoted fields, so 'Apt 4, Suite B' stays one cell.

• 1-based positions: 1 is the first column, 2-4 is a range.
• keep = output only the selected; drop = remove them, keep the rest.
• Output delimiter is separate — read a CSV, emit a TSV.

Worked example

A contacts export with columns name,phone,email,notes,company — selector '1,3,5' in keep mode emits just name, email, company rows, quoting preserved: "Smith, John" stays one field.

Limits

Up to 300,000 characters / 50,000 rows. Ragged rows (short lines) keep what they have — no padding. Column names are positional only; there is no select-by-header-name yet.