Charset utility

Encode and decode HL7v2 wire bytes as UTF-8 (fatal, no silent corruption).

What it does

HL7v2 travels as bytes, but the rest of an HL7v2 pipeline (parsing, querying, linting) works in strings — so something has to convert between the two. This package is that step, in both directions.

For now it handles UTF-8 only — the HL7 2.x baseline and a strict superset of 7-bit ASCII. Decoding is fatal: a non-UTF-8 feed throws here instead of being silently corrupted into U+FFFD replacement characters (the failure behind issue #659 (opens in a new tab)). The package is zero-dependency and runtime-agnostic — just the Web platform's TextEncoder / TextDecoder.

Honouring the character set a message declares (MSH-18, byte-order marks, and other encodings for legacy peers) is deferred and tracked in issue #662 (opens in a new tab); the package exists as the home for that future work.

Install

npm install @glion/util-charset

Use

import { decodeBytes, encodeBytes } from "@glion/util-charset";

// Inbound: de-framed payload bytes -> text.
const text = decodeBytes(payload);

// Outbound: text -> UTF-8 wire bytes.
const wire = encodeBytes(text);

API

decodeBytes(bytes)

Decodes UTF-8 payload bytes (after MLLP de-framing) to text. A leading UTF-8 BOM is stripped.

• bytes (Uint8Array) — The de-framed payload bytes.
• Returns string — The decoded HL7v2 text.
• Throws CharsetError (code: "INCOMPATIBLE_CHARSET") — When the bytes carry a non-UTF-8 byte-order mark or are otherwise not valid UTF-8.

encodeBytes(text)

Encodes HL7v2 text to UTF-8 wire bytes.

• text (string) — The HL7v2 message text.
• Returns Uint8Array — The UTF-8 wire bytes.