Skip to content

UTF-16 ​

Signatureutf16.encode(text: string): u16[]
Signatureutf16.encodeLE(text: string): u8[]
Signatureutf16.encodeBE(text: string): u8[]
Signatureutf16.decode(units: u16[]): Result<string, EncodingError>
Signatureutf16.decodeLossy(units: u16[]): string
Signatureutf16.decodeLE(bytes: u8[]): Result<string, EncodingError>
Signatureutf16.decodeLELossy(bytes: u8[]): string
Signatureutf16.decodeBE(bytes: u8[]): Result<string, EncodingError>
Signatureutf16.decodeBELossy(bytes: u8[]): string

Import utf16 from tyrus:encoding. encode(text) returns host-independent u16 code units. encodeLE(text) and encodeBE(text) return byte arrays in the explicitly selected order. decode, decodeLE, and decodeBE are strict; the matching Lossy methods replace malformed surrogate sequences with U+FFFD.

tyrus
import { EncodingError, utf16 } from "tyrus:encoding";

const units: u16[] = utf16.encode("A😀");
const littleEndian: u8[] = utf16.encodeLE("A😀");
const bigEndian: u8[] = utf16.encodeBE("A😀");
const malformed: u16[] = [55296];

try {
    print(utf16.decode(units)); // A😀
    print(utf16.decodeLE(littleEndian)); // A😀
    print(utf16.decodeBE(bigEndian)); // A😀
} catch (error: EncodingError) {
    print(error.code);
}
print(utf16.decodeLossy(malformed)); // �

Byte-oriented decoders require an even number of bytes. The strict variants return Err(EncodingError) for an odd byte length or malformed surrogate sequence; the lossy variants repair malformed surrogate sequences only.