Validation Library
The 55 validators behind DataLens, in one place: what each one checks, how, and on what. Pick one to read and test it, then paste a CSV to see which of your columns it fits.
Validators
55 validators
- Format
- Australian phone number
- BIC / SWIFT code
- Boolean
- BSB number
- Canadian postal code
- Currency amount
- Date
- Digits only
- Email address
- Indian PAN
- International phone number (E.164)
- IPv4 address
- IPv6 address
- ISO 8601 date
- MAC address
- National Insurance number
- Number
- Phone number (any country)
- UK postcode
- UK sort code
- URL
- US Social Security number
- US ZIP code
- UUID
- Whole number
- Checksum
- Aadhaar number
- ABA routing number
- Australian Business Number (ABN)
- Australian Company Number (ACN)
- Australian Tax File Number (TFN)
- Canadian Social Insurance Number
- IBAN
- Indian GSTIN
- ISIN (securities identifier)
- Luhn check digit
- Medicare card number
- NHS number
- Payment card number
- Singapore NRIC / FIN
- Range
- Australian postcode
- Date of birth
- Latitude
- Longitude
- Numeric range
- Text length
- Pattern
- Regular expression
- Reference
- Allowed values
- Country code (ISO 3166-1 alpha-2)
- Uniqueness
- Unique key (rows)
- Unique values
- Completeness
- Completeness threshold
- Required value
- Schema
- Declared type
- Privacy
- Re-identification risk (composite PII)
- Security
- JSON Web Token
Details
Choose a validator from the list.
Your data
Paste a CSV with a header row. Each column is classified, the validators that fit it are listed, and the validator you picked is run over the columns it fits. Nothing leaves this page.
Powered by DataLens by BNS Data
How the library works
Each validator is a small, tested definition: what it recognises, the standard it follows, the parameters it takes and the checks it runs, including check digits where the standard defines them. The same package runs on these pages, in the DataLens app and on the DataLens server, so a value judged valid here is judged valid there.
What differs is what happens next. Here a failure is flagged and nothing more. In DataLens the same validators can be saved as a ruleset on a dataset, run on every refresh, stop a load at a quality gate, and keep a masked record of each failing row.
Structural validation does not establish that the identifier exists or belongs to a particular person.
Tools built on it
- Value Validator — check a list of values against one validator.
- Column Classifier — see what each column of a CSV holds.