Check a CSV column against an allowed set of values
Paste CSV, pick a column, list the allowed categories, and flag every cell whose value falls outside the set — with a count of each distinct unexpected value so typos jump out.
About this tool
A domain check (also called a controlled-vocabulary, allowed-values, or categorical
membership check) confirms that every value in a column is one of a known, agreed-upon set of
categories. It is one of the most common data-quality checks: a status column should only ever
hold active, inactive, or pending; a country column should only hold your supported ISO
codes. A single typo — activ instead of active — silently breaks downstream filters, joins,
and reports.
This tool takes CSV text, a target column, and the allowed set of categories, then flags every
cell whose value is not in the set. It reports how many cells were checked, how many were
valid, how many were invalid, the exact offending rows (with their line numbers), and — most
usefully — a summary of the distinct unexpected values with a count of each. That summary
turns "37 bad rows" into "you have one typo, activ, appearing 37 times", which is what you
actually need to fix the data.
Worked example
Paste this CSV, set the column to status, and add active, inactive, and pending as the
allowed values:
name,status
Ada,active
Bo,activ
Cy,pending
De,active
The report flags row 2 — activ is not in the allowed set — and lists it under Unexpected
values as "activ" ×1, while confirming the other three rows are valid.
Options
- Ignore case — off by default (an exact, case-sensitive match). Turn on to treat
ACTIVE,Active, andactiveas the same category. - Trim surrounding whitespace — on by default, so
" active "matchesactive. Turn off to require an exact, unpadded value. - First row is a header — on by default; the column is then chosen by header name. Turn off to select the column by its 0-based index.
- Allow blank cells — on by default (empty cells are skipped). Turn off to flag blank cells as invalid.
- CSV delimiter — auto-detects comma, tab, semicolon, or pipe from the first line; override it if detection guesses wrong.
- Max issues to list — caps how many offending rows and distinct unexpected values are listed (the total invalid count is always reported in full).
- Output format — a readable text report, or a structured JSON object for scripting.
Everything runs locally in your browser — your CSV is never uploaded.
FAQ
How do I select the column — by name or by number?
Either. With First row is a header on (the default), type the header name (for example
status). With it off, type the 0-based index of the column instead (0 is the first column,
1 the second). A purely numeric value is always treated as an index, even when a header row is
present, so you can index into a file whose headers happen to be numbers.
What is the difference between this and a regex or type validator?
This tool checks membership in a fixed set of categories — the value must be exactly one of the allowed strings. A regex validator checks that values match a pattern (like a ZIP code or email shape), and a type validator checks that values are the right data type (integer, date, boolean). Use a domain check when the valid values are a short, known list you can enumerate.
Why does it summarize distinct unexpected values with counts?
Because a plain row-by-row list buries the real problem. If one misspelling appears in hundreds of
rows, the per-row list is a wall of duplicates. The Unexpected values summary collapses that to
one line — "activ" ×342 — showing at a glance whether you have one systematic typo or many
scattered bad values. It is the single most useful output for cleaning categorical data.
Can it check several columns or apply a percentage tolerance in one run?
No — this is the focused single-column tool: it validates one column against one allowed set and reports exact valid/invalid counts. There is no built-in "pass if ≥ 95% are valid" tolerance, but the report gives you the precise counts, so you can apply any threshold you like yourself. For multi-column or multi-rule schema validation, use a general CSV validator instead.
Developer & Automation Access
Run it from the terminal
Same engine as this page, headless — via the gizza CLI:
gizza tool enum-domain-check "name,status
Ada,active
Bo,activ
Cy,pending
De,active" 'column=status' 'allowed=active,inactive,pending'New to the CLI? Get gizza →
Open it by URL
Pre-fill and auto-run this tool with query parameters — the names match the API/CLI:
https://gizza.ai/tools/enum-domain-check/?data=name%2Cstatus%0AAda%2Cactive%0ABo%2Cactiv%0ACy%2Cpending%0ADe%2Cactive&column=status&allowed=active%2Cinactive%2Cpending&ignore_case=true&trim=true&has_header=true&allow_blank=true&delimiter=auto&max_issues=50&output=textMachine-readable descriptor: tool.json — title + parameters JSON Schema for agents.