Changelog

All notable changes to this project are documented here. The format follows Keep a Changelog, and the project adheres to Semantic Versioning.

[Unreleased]

0.11.0 - 2026-06-24

Two major capability additions plus broader format coverage and two new VAT types (92 -> 94). The two real gaps versus python-stdnum - standalone check-digit algorithms and online VAT validation - are now closed.

Added

• More canonical format: :hk-id (A123456(3)), :kr-brn (124-81-00998), :au-abn (51 824 753 556), :au-tfn (123 456 782), :no-org (974 760 673).
• New namespace stdnum.checkdigit exposing the underlying algorithms as standalone primitives (the python-stdnum stdnum.luhn/verhoeff/iso7064 parallel): luhn-valid? / luhn-check-digit, verhoeff-valid? / verhoeff-check-digit, iso7064-mod11-2-valid? / iso7064-mod11-2-check, and iso7064-mod97-10-valid?.
• Two more VAT types (90 -> ... 94 total): :es-vat (Spain - covers the DNI, NIE and CIF forms; CIF accepts the digit- or letter-control convention) and :ie-vat (Ireland, mod-23 check letter). Verified against Telefónica's CIF and Google Ireland's VAT.
• New namespace stdnum.vies for online EU VAT validation against the official VIES service (vies/check), confirming a VAT number actually exists in the member-state registry and returning the trader name/address where disclosed. Member-state errors (e.g. MS_UNAVAILABLE) surface as {:error ...} so a transient outage is never mistaken for an invalid number. Requires JDK 11+ (uses java.net.http); adds an org.clojure/data.json dependency. stdnum.core remains pure and network-free.

0.10.0 - 2026-06-24

Depth pass, part 2: broader format coverage and parse decomposition of structural identifiers (BIC, CLABE, IMEI, ISIN) and non-date IDs. Still 92 types; this release deepens the surface so most identifiers now report their parts, not just {:valid? true}. 22 types have parse extraction and 22 have canonical format.

Added

• Canonical format for more existing types with standard display forms: :nhs (943 476 5919), :gb-nino (AB 12 34 56 C), :in-aadhaar (2341 2341 2346), :ch-ahv (756.9217.0769.85), :ca-sin (046 454 286), :fr-nir (2 55 08 14 168 025 38).
• More non-date parse extraction: :in-pan -> :holder-type (from the 4th character), :ec-ced -> :province-code :province, :pe-ruc -> :entity-type (:company/:natural-person), and :credit-card now also returns :iin (first 6) and :last4.
• Structural parse decomposition: :bic -> :bank-code :country :location-code :branch-code (via iban4j); :mx-clabe -> :bank-code :branch-code :account; :imei -> :tac :serial; :isin -> :country :nsin.

0.9.0 - 2026-06-24

Depth pass continued: more parse field-extraction and canonical format. Extraction now covers every shipped identifier that embeds a birth date. No new types (still 92); this release deepens the existing surface.

Added

• parse :cn-ric (China resident ID) now extracts :birth-date (full YYYY-MM-DD) and :gender.
• More parse extraction: :pl-pesel -> :birth-date (century resolved from the PESEL month offset, so unambiguous) + :gender; :fr-nir -> :gender :birth-year :birth-month :department; :se-pnr -> :birth-date :gender; :be-nn -> :birth-date (century from the validating mod-97 base) + :gender; :bg-egn -> :birth-date (century from month offset) + :gender.
• Canonical format for :se-pnr (811218-9876) and :be-nn (00.01.25-111.48).
• Canonical format for more types: :orcid (0000-0002-1825-0097), :isni (0000 0001 2103 2683), :cas (7732-18-5), :ar-cuit (30-70308853-4), :cl-rut (97.004.000-5), :co-nit (890.903.938-8).

0.8.0 - 2026-06-24

Two new types (90 -> 92) plus a depth pass: parse now extracts the structured data embedded in several identifiers, matching the most-used part of python-stdnum's per-type richness. Everything verified against published example numbers; stays EPL.

Added

• :gln (GS1 Global Location Number) and :mx-curp (Mexico CURP). Clean-room / engine-backed, each verified against a published example.
• parse now extracts the structured data embedded in several IDs (not just {:valid? true}):
  • :mx-curp -> :birth-date :gender :state :state-name
  • :ee-ik (Estonia), :jmbg (ex-Yugoslav) -> :birth-date :gender (+ :region for JMBG)
  • :za-id (South Africa) -> :gender :citizen :birth-date
  • :vin -> :wmi :vds :vis :model-year :plant :serial
  • :it-cf (Italy codice fiscale) -> :gender :birth-day :birth-month :birth-year :comune-code (the century is not encoded in a CF, so the 2-digit year is returned as-is)
  • :iban -> now also :bank-code :branch-code :account-number (where the country's BBAN defines them), via iban4j.

0.7.0 - 2026-06-24

Coverage expansion: 71 -> 90 identifier types. Adds LatAm tax IDs (Nubank/MercadoLibre audience), more EU national/person IDs, research IDs (ORCID/ISNI), and the longer GS1 keys. Every new checksum is clean-room from the public standard (or engine-backed) and verified against a real published number.

Added

• More commerce / industry identifiers: :ean8 (EAN-8/GTIN-8), :ismn (sheet-music ISMN), :cas (CAS Registry Number for chemicals), :imo (IMO ship number). Clean-room / engine-backed, each verified against a real published number.
• National tax / person IDs: :fr-nir (France social security), :pl-pesel (Poland), :ar-cuit (Argentina), :cl-rut (Chile), :co-nit (Colombia). Clean-room, each verified against a real published number.
• More national IDs: :pe-ruc (Peru), :ie-pps (Ireland PPS), :ee-ik (Estonia isikukood), :jmbg (shared ex-Yugoslav number: RS/BA/ME/MK/SI/HR), :ec-ced (Ecuador cedula). Clean-room, each verified against a real published number.
• :bg-egn (Bulgaria), :orcid and :isni (researcher / name IDs, ISO 7064 MOD 11-2), and the longer GS1 keys :gtin14 and :sscc. Clean-room, each verified against a real published number.

0.6.0 - 2026-06-24

Coverage expansion: 54 -> 71 identifier types, adding a commerce/vehicle/health category and APAC business IDs. Every new checksum is clean-room from the public standard (or engine-backed) and verified against a real published number.

Added

• :nz-ird (New Zealand IRD), :be-nn (Belgium national number), :fi-hetu (Finland HETU). Clean-room, verified against published numbers.
• :figi (Financial Instrument Global Identifier, OMG check digit) and four more EU VAT numbers: :mt-vat (Malta), :sk-vat (Slovakia), :lt-vat (Lithuania), :cy-vat (Cyprus). Clean-room, each verified against a real published number.
• :ro-vat (Romania CUI), :sg-nric (Singapore NRIC/FIN, S/T/F/G series), :hk-id (Hong Kong HKID), :kr-brn (South Korea Business Registration Number). Clean-room, each verified against a real published number.
• Commerce / vehicle / healthcare: :ean13 (EAN-13/GTIN-13 barcode), :upc (UPC-A), :vin (ISO 3779 vehicle ID), :nhs (UK NHS number), :npi (US National Provider Identifier). Barcode checks are engine-backed (Commons Validator EAN-13); VIN/NHS/NPI are clean-room, each verified against a real published number.

0.5.0 - 2026-06-24

Coverage expansion: 40 -> 54 identifier types, spanning ~24 countries. Every checksum clean-room from the public standard and verified against published example numbers.

Added

• More VAT/GST: :gr-vat :lu-vat :si-vat :ee-vat :hu-vat :in-gstin (India GST, base-36 check char).
• More tax & national IDs: :pt-nif :cz-ico :jp-cn :au-tfn :hr-oib (Croatia OIB), :it-cf (Italy codice fiscale, mod-26 check letter), :ch-uid :ch-ahv (Switzerland).

0.4.0 - 2026-06-23

Big coverage expansion: 13 -> 40 identifier types. Every new checksum is clean-room from the public standard and verified against published example numbers (no third-party port; stays EPL).

Added

• VAT numbers, country prefix optional on input: :de-vat :fr-vat :it-vat :be-vat :pl-vat :gb-vat :at-vat :dk-vat :fi-vat :se-vat.
• National IDs: :gb-nino, :ca-sin, :au-abn, :in-pan, :in-aadhaar, :es-dni, :es-nie, :nl-bsn, :cn-ric, :se-pnr, :za-id, :no-org, :tr-tc.
• Banking: :mx-clabe (Mexico CLABE bank account).

Changed

• README: trimmed Usage to the core API and grouped the identifier types by category in a collapsible block, so the page stays scannable as the set grows.

0.3.0 - 2026-06-23

Added

• Securities identifiers (engine-backed check digits): :cusip (US/Canada), :sedol (UK/Ireland) - rounding out the :isin/:lei set.
• US national numbers, clean-room from the public structural rules: :us-ssn (area/group/serial rules + the SSA's reserved advertising numbers, with format) and :us-ein (IRS campus prefix validation, with format).

0.2.0 - 2026-06-23

Added

• Three new identifier types, implemented clean-room from their public standards (no third-party port; stays EPL):
  • :lei - Legal Entity Identifier (ISO 17442, ISO 7064 mod-97-10).
  • :br-cpf - Brazil individual taxpayer registry, with format.
  • :br-cnpj - Brazil company registry, with format.
• compact/format now also strip / (used by the CNPJ written form).

Country-specific identifiers are keyed by an ISO-3166 prefix (e.g. :br-cpf).

0.1.0 - 2026-06-23

Initial release.

Added

• stdnum.core - a unified facade for validating, parsing, and formatting standard identifier numbers, dispatched on a type keyword:
  • valid?, parse, format, compact, detect, card-network, and the types set.
  • Identifier types: :credit-card (with network detection), :iban, :bic, :isbn, :issn, :isin, :aba (US bank routing), :imei, :luhn.
• Idiomatic facade over Apache Commons Validator 1.10.1 and iban4j 3.2.11 - no algorithm reimplementation. Bad input data never throws; only an unknown identifier type does.