Data Quality

Data Validity Rule Checker

Checks each column against rules you write and lists every violation with its row number.

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. Put the data in the first box.
  2. Write the rules in the second box, one column per line, as in "age: integer, min(0), max(120)".
  3. Select Check rules.
  4. Check the summary for rules that named a column the file does not have — a typo there looks like a clean pass otherwise.

What validity rule checker does

Profiling can tell you what is in a file. It cannot tell you what should be, because only you know that status has three legal values, that a quantity is never negative and that the customer reference has to be unique. That knowledge is the part of data quality no tool can infer, so this one asks you to write it down in a few lines and then checks every row against it.

An empty cell passes every rule here except required, and that is deliberate: otherwise one missing value fails five rules at once and the report becomes unreadable. Missingness is required’s job, and the completeness tools do it better. Rules that name a column the file does not have are reported as skipped rather than silently ignored, because a typo in a rule name should not look like a clean pass.

Frequently asked questions

required, email, url, integer, number, date, boolean and unique take no argument. min, max, length, in and regex take one, written in brackets — min(0), length(3-40), in(active|inactive), regex(^[A-Z]{2}\d{4}$). Several rules for a column go on one line separated by commas.

Because otherwise one missing value fails every rule on the column at once and the report becomes unreadable. Emptiness is what the required rule is for; add it to any column where a gap is itself a violation.

The rule is skipped and the summary says so. Silently ignoring it would be worse, because a typo in a rule name would then look exactly like a clean pass.

The first five hundred are shown, and the download contains all of them. Beyond that number the useful information is the pattern rather than the individual rows, and the summary gives you the totals.