Data Quality

CSV Cleaner

The safe, explainable half of cleaning: whitespace, invisible characters, null tokens, empty rows and columns.

Loading the tool…

Processing happens locally in your browser. What you paste or load is processed by this page and is not uploaded to a server. Nothing is stored unless you use a control that says it stores something, and you can clear anything this site has kept from the privacy page.

How to use this tool

  1. Paste the messy file.
  2. Turn off anything you do not want touched — every option here is a change someone may later ask you to justify.
  3. Select Clean.
  4. Read the summary before downloading; it says exactly how many cells each change affected.

What csv cleaner does

This does the cleaning you can explain to whoever receives the file afterwards, and stops there. Leading and trailing spaces go, runs of spaces inside a value collapse, zero-width characters and non-breaking spaces that survived a copy and paste are removed, the null tokens you nominate become genuinely empty, and rows and columns with nothing in them at all are dropped.

What it deliberately does not do is impute a value, round a number, reformat a date or rename a column it was not asked about. That is the line between cleaning data and quietly altering someone else’s, and a tool that crosses it produces a file nobody can reconcile with the source. The summary lists every category of change and how many cells it touched, so the result is auditable rather than merely tidier.

Frequently asked questions

Only in the ways you leave switched on, and every one of them is listed in the summary with a count. It does not impute missing values, round numbers, reformat dates or rename columns, because those are changes the person receiving the file cannot reconcile with the source.

Zero-width spaces and non-breaking spaces survive a copy and paste from a web page or a PDF and are impossible to see. They then break every equality test downstream — two values that look identical simply are not — so removing them is one of the few cleaning steps with no downside.

The header cleaner works on the first row: names, casing, characters that need quoting. This works on the data underneath it and leaves the header alone. They are complementary, and running the header cleaner afterwards is a reasonable habit.

The first occurrence is kept and later identical rows are dropped, so the original order of what remains is unchanged. Rows are compared on every column after the other cleaning steps have run.