CSV Column Extractor

Keep the columns you name by number, starting at 1.

Use this tool

Input CSV format
Output CSV format

Parsed table

Result

The result appears in this panel after you run the tool.

About this tool

Enter numbers such as 1, 3. Those columns are copied, in that order, into a new table and written with the output delimiter. Missing cells become empty.

How to use it

  1. Paste a table or open a file.
  2. List the column numbers to keep, starting at 1.
  3. Use Extract columns.

Limitations

Listed columns kept, in the order listed

Numbers start at one, and unlisted cells are discarded

After the input dialect splits the file, you name columns by number, starting at one. A number outside one through the widest row is an error. The surviving row contains only those columns, in the order you typed the numbers, not in their old order. Typing the same number twice copies that column twice. A short row contributes an empty cell when the requested index is past its end. Every other column is omitted. The omitted text is not written beside the result and cannot be recovered from it.

The output dialect serializes what remains. You can change separator, quote, and escape while you extract. The header flag from the input tells the preview whether to treat the first row as names. Extraction does not rewrite those names. If column 1 was a header cell and you keep column 1, that cell is still the first cell of the first row. If you drop it, the header name is gone along with the data under it.

Why a list of numbers is not a schema

There is no check that a column holds dates or that two kept columns form a key. Empty cells stay empty. Quoted commas stay inside the cell they belonged to, because quoting was handled at split time. A later JSON conversion will turn the remaining header into keys if you enable a header then, and it will still store strings. This page does not do that conversion.

Plan the loss

Write down the numbers you kept. The result looks like a complete table, and nothing in the file says which source columns were abandoned. If you needed every column with a new order, the reorder page moves the listed columns to the front and keeps the rest after them. Extraction is the page that throws the rest away.

Privacy

Your input is processed in your browser and is not uploaded to UtilChain. File names, text, and results are not sent to analytics.