Zero-dependency UPC / EAN / ISBN / ITF-14 barcode check-digit math, validation, UPC-E expansion, and random / sequential code generation for JavaScript and TypeScript. Calculate and verify the GS1 mod-10 check digit for UPC-A, UPC-E, EAN-13, EAN-8, ISBN-13, and ITF-14 — no image rendering, no dependencies, just the numbers.
This is the production barcode engine behind upcgen.com — the rendering layer stays in the app; the math is open.
npm install @x1ee7/barcode-corepnpm add @x1ee7/barcode-core
# or: yarn add @x1ee7/barcode-coreESM + CommonJS + TypeScript types. No dependencies.
import {
computeCheckDigit,
validateBarcodeText,
upcEExpand,
generateRandom,
buildSequential,
} from "@x1ee7/barcode-core";
// Append the GS1 mod-10 check digit
computeCheckDigit("upca", "03600029145"); // → "036000291452"
computeCheckDigit("ean13", "978013110362"); // → "9780131103627"
// Validate digit count + check digit
validateBarcodeText("upca", "036000291452"); // → { ok: true }
validateBarcodeText("upca", "036000291450"); // → { ok: false, error: "invalid check digit — expected 2, got 0" }
// UPC-E (6-digit body + number system) → 11-digit UPC-A payload
upcEExpand("0", "123455"); // → "01234500005"
// One random, structurally valid code
generateRandom("ean13"); // → e.g. "4006381333931"
// A contiguous run sharing a prefix, each with a valid check digit
buildSequential("ean13", "978", 1, 3);
// → ["9780000000012", "9780000000029", "9780000000036"]UPC-A, EAN-13, EAN-8, ISBN-13, and ITF-14 all terminate in a single GS1 mod-10 check digit. From the right of the payload, digits are weighted alternately by 3 and 1, summed, and the check digit is whatever brings that sum up to the next multiple of 10:
checkDigit = (10 − (weightedSum mod 10)) mod 10
computeCheckDigit(format, payload) applies the correct positional weights
for each format (exposed as SPEC) and returns the full code with the check
digit appended. validateBarcodeText recomputes it and also enforces the
expected digit count for fixed-length formats.
Fixed-length (have a computable check digit):
upca, upce, ean13, ean8, isbn, itf14
Variable-length (no fixed check digit — validation always passes):
code128, datamatrix, fnsku
| Export | Description |
|---|---|
computeCheckDigit(format, payload) |
Returns payload + GS1 mod-10 check digit. Throws for variable-length formats. |
validateBarcodeText(format, text) |
{ ok: true } or { ok: false, error }. Variable-length formats always pass. |
normalizeText(format, text) |
Upper-cases except Data Matrix (case-sensitive). |
upcEExpand(numSystem, comp) |
Expands a UPC-E body to its 11-digit UPC-A payload. |
generateRandom(format) |
One random, structurally valid code. |
buildSequential(format, prefix, start, count) |
Contiguous run of check-digit-terminated codes. |
isVariableLength(format) / isCaseSensitive(format) |
Format predicates. |
SPEC |
Payload length + positional weights per fixed-length format. |
computeCheckDigit("upca", payload) or computeCheckDigit("ean13", payload).
Pass the payload without the final digit and it returns the full,
check-digit-terminated code.
validateBarcodeText(format, text) returns { ok: true } or
{ ok: false, error }, checking both the digit count and the GS1 mod-10
check digit.
upcEExpand(numberSystem, sixDigitBody) returns the equivalent 11-digit
UPC-A payload; run it through computeCheckDigit("upca", …) for the full
12-digit code.
No — by design. barcode-core is the numeric layer (check digits,
validation, code generation). Pair it with any SVG/canvas renderer for
visuals; the math here is what's hard to get right.
Yes, as the isbn format (13-digit ISBN with the GS1 mod-10 check digit, the
modern Bookland EAN form).
This library is extracted from a commercial product and maintained in the open. If it saves you the afternoon of re-deriving GS1 check-digit weights, please consider sponsoring on GitHub — it directly funds maintenance, more symbologies, and keeping the math correct.
MIT © upcgen.com