Supported formats

The usual file is comma-separated UTF-8, with or without a byte-order mark, up to 32 MB. Quoted commas, doubled quotes, and line breaks inside quotes are part of the cell. A header-only file is valid. Duplicate header names stay as separate columns.

If parsing fails, or a comma parse produces one column that still contains semicolons, you can reparse that same file with semicolons or tabs. The file is still read on this device.

A .xlsx file opens in the browser. The first row of the chosen sheet becomes the header, and every cell becomes text. When the workbook has more than one sheet, use the Sheet menu. The CSV cleaner downloads a CSV, not a workbook.

Stored formula results are shown. Formulas are not calculated, and merged cells are not filled. Blank rows after the last value are omitted. Older .xls files, and password-protected workbooks, need to be saved as an unprotected .xlsx first. Characters that are not valid UTF-8 in a CSV may appear as a replacement mark, and the page says so.

A download can optionally prefix cells that start with =, +, -, or @. That option changes the downloaded text only. The table on screen stays as it is until you apply a cleaning rule.