tricu

io-utf8.test.mjs (2189B)

---

 import test from 'node:test';
 import assert from 'node:assert/strict';
 import {encodeUTF8,decodeUTF8} from '../dist/io-utf8.js';
 test('all Unicode scalars match independent platform encoding and strict decoding',()=>{
  const encoder=new TextEncoder();
  for(let start=0;start<=0x10ffff;start+=1024){
   const scalars=[];for(let n=start;n<Math.min(start+1024,0x110000);n++)if(n<0xd800||n>0xdfff)scalars.push(n);
   const text=String.fromCodePoint(...scalars),expected=[...encoder.encode(text)];
   assert.deepEqual(encodeUTF8(text),expected);assert.deepEqual(decodeUTF8(expected,true),{text,carry:[]});
  }
 });
 test('every valid multibyte scalar prefix yields only copied bounded carry, and fails when final',()=>{
  const encoder=new TextEncoder();
  for(let n=128;n<=0x10ffff;n++){
   if(n>=0xd800&&n<=0xdfff)continue;
   const bytes=[...encoder.encode(String.fromCodePoint(n))];
   for(let split=1;split<bytes.length;split++){
    const prefix=bytes.slice(0,split),out=decodeUTF8(prefix,false);
    assert.equal(out.text,'');assert.deepEqual(out.carry,prefix);assert.notEqual(out.carry,prefix);
   }
  }
  // Exhaustive prefixes above; explicit final/invalid-prefix controls below keep
  // exception creation out of the million-scalar loop.
  for(const bytes of [[0xc2],[0xe0,0xa0],[0xf4,0x8f,0xbf]])assert.throws(()=>decodeUTF8(bytes,true),/incomplete/);
 });
 test('invalid lead/continuation/overlong/surrogate/range prefixes fail before finality',()=>{
  for(const bytes of [[0x80],[0xc0],[0xc1],[0xf5],[0xff],[0xe0,0x9f],[0xed,0xa0],[0xf0,0x8f],[0xf4,0x90],[0xc2,0x41],[0xe2,0x82,0x41]])
   for(const final of [false,true])assert.throws(()=>decodeUTF8(bytes,final),/invalid UTF-8/);
  assert.throws(()=>decodeUTF8([65,0xff],false),/at byte 1$/);
 });
 test('BOM, NUL, combining scalars and line endings are unchanged; scalar/byte bounds are separate',()=>{
  const text='\ufeffa\0\r\nλ😀e\u0301';assert.deepEqual(decodeUTF8(encodeUTF8(text),true),{text,carry:[]});
  assert.equal(encodeUTF8('😀'.repeat(1024)).length,4096);
  assert.throws(()=>encodeUTF8('a'.repeat(1025)),/scalar/);assert.throws(()=>encodeUTF8('\ud800'),/scalar/);
  assert.throws(()=>decodeUTF8(Array(4097).fill(65),false),/byte/);
 });