mirror of
https://github.com/hanzoai/tsgo.git
synced 2026-10-30 19:37:15 +00:00
048da4e459
Co-authored-by: copilot-swe-agent[bot] <[email protected]> Co-authored-by: jakebailey <[email protected]>
40 lines
1.5 KiB
TypeScript
40 lines
1.5 KiB
TypeScript
// @strict: true
|
|
// @noEmit: true
|
|
|
|
// Verifies that template literal type inference handles multi-byte characters correctly,
|
|
// slicing by Unicode code point (rune) instead of by byte.
|
|
|
|
type GetFirst<S extends string> = S extends `${infer Head}${string}` ? Head : never;
|
|
type GetRest<S extends string> = S extends `${string}${infer Rest}` ? Rest : never;
|
|
type Split<S extends string> = S extends `${infer Head}${infer Rest}` ? [Head, Rest] : never;
|
|
|
|
// Single-byte (ASCII) characters
|
|
type A = GetFirst<"ABC">; // "A"
|
|
type BC = GetRest<"ABC">; // "BC"
|
|
type ABC_Split = Split<"ABC">; // ["A", "BC"]
|
|
|
|
// Multi-byte characters (Japanese, 3 bytes in UTF-8, 1 code unit in UTF-16)
|
|
type Ah = GetFirst<"あいう">; // "あ"
|
|
type IU = GetRest<"あいう">; // "いう"
|
|
type AIU_Split = Split<"あいう">; // ["あ", "いう"]
|
|
|
|
// Mixed single-byte and multi-byte
|
|
type M1 = GetFirst<"aあ">; // "a"
|
|
type M2 = GetFirst<"あa">; // "あ"
|
|
|
|
// 2-byte UTF-8 characters (Latin Extended, 1 code unit in UTF-16)
|
|
type E = GetFirst<"éàü">; // "é"
|
|
|
|
// Emoji (4 bytes in UTF-8, 2 code units in UTF-16 as surrogate pair)
|
|
type Emoji = GetFirst<"😀abc">;
|
|
|
|
// Verification
|
|
type Expect<T, Expected> = T extends Expected ? true : false;
|
|
const check1: Expect<A, "A"> = true;
|
|
const check2: Expect<Ah, "あ"> = true;
|
|
const check3: Expect<BC, "BC"> = true;
|
|
const check4: Expect<IU, "いう"> = true;
|
|
const check5: Expect<M1, "a"> = true;
|
|
const check6: Expect<M2, "あ"> = true;
|
|
const check7: Expect<E, "é"> = true;
|