test: add legacyHookDecode tests

Author: ChALkeR

fallback/encoding.js

@@ -265,9 +265,18 @@ export function legacyHookDecode(input, fallbackEncoding) {
   const bomEncoding = getBOMEncoding(u8)
   if (bomEncoding) u8 = u8.subarray(bomEncoding === 'utf-8' ? 3 : 2)
   const enc = bomEncoding ?? fallbackEncoding ?? 'utf-8' // "the byte order mark is more authoritative than anything else"
+
   if (enc === 'utf-8') return utf8toStringLoose(u8)
-  if (enc === 'utf-16le') return utf16toStringLoose(u8, 'uint8-le')
-  if (enc === 'utf-16be') return utf16toStringLoose(u8, 'uint8-be')
+  if (enc === 'utf-16le' || enc === 'utf-16be') {
+    let suffix = ''
+    if (u8.byteLength % 2 !== 0) {
+      suffix = replacementChar
+      u8 = u8.subarray(0, -1)
+    }
+
+    return utf16toStringLoose(u8, enc === 'utf-16le' ? 'uint8-le' : 'uint8-be') + suffix
+  }
+
   if (!Object.hasOwn(labels, enc)) throw new RangeError(E_ENCODING)
 
   if (multibyteSet.has(enc)) {

tests/encoding/generic.test.js

@@ -1,4 +1,9 @@
-import { TextDecoder, TextEncoder, getBOMEncoding } from '@exodus/bytes/encoding.js'
+import {
+  TextDecoder,
+  TextEncoder,
+  getBOMEncoding,
+  legacyHookDecode,
+} from '@exodus/bytes/encoding.js'
 import { fromHex } from '@exodus/bytes/hex.js'
 import { test, describe } from 'node:test'
 import unfinishedBytesFixtures from '../fixtures/text-encoding.unfinishedBytes.js'
@@ -83,6 +88,40 @@ describe('encodings are ASCII supersets, except utf-16 and iso-2022-jp', () => {
   }
 })
 
+describe('legacyHookDecode', () => {
+  const fixtures = {
+    replacement: [
+      ['', ''],
+      ['00', '\uFFFD'],
+      ['ff', '\uFFFD'],
+      ['20', '\uFFFD'],
+      ['2020', '\uFFFD'],
+      // BOM takes preference
+      ['efbbbf', ''],
+      ['efbbbf2a', '*'],
+      ['efbbbf202a', ' *'],
+      ['fffe', ''],
+      ['fffe2a20', '\u202A'],
+      ['fffe2a', '\uFFFD'],
+      ['fffe00d72a', '\uD700\uFFFD'],
+      ['fffe00d82a', '\uFFFD\uFFFD'],
+      ['feff', ''],
+      ['feff202a', '\u202A'],
+      ['feff20', '\uFFFD'],
+      ['feffd70020', '\uD700\uFFFD'],
+      ['feffd80820', '\uFFFD\uFFFD'],
+    ],
+  }
+
+  for (const [encoding, data] of Object.entries(fixtures)) {
+    test(encoding, (t) => {
+      for (const [hex, string] of data) {
+        t.assert.strictEqual(legacyHookDecode(fromHex(hex), encoding), string, `${hex}`)
+      }
+    })
+  }
+})
+
 test('getBOMEncoding', (t) => {
   const fixtures = [
     [null, ''],