@exodus/bytes 1.0.0-rc.9 → 1.1.0

package/fallback/_utils.js

@@ -1,22 +1,30 @@
 const { Buffer, TextEncoder, TextDecoder } = globalThis
 const haveNativeBuffer = Buffer && !Buffer.TYPED_ARRAY_SUPPORT
-let isNative = (x) => x && (haveNativeBuffer || `${x}`.includes('[native code]')) // we consider Node.js TextDecoder/TextEncoder native
+export const nativeBuffer = haveNativeBuffer ? Buffer : null
+export const isHermes = Boolean(globalThis.HermesInternal)
+export const isDeno = Boolean(globalThis.Deno)
+export const isLE = new Uint8Array(Uint16Array.of(258).buffer)[0] === 2
+
+let isNative = (x) => {
+  if (!x) return false
+  if (haveNativeBuffer) return true // we consider Node.js TextDecoder/TextEncoder native
+  const s = `${x}`
+  // See https://github.com/facebook/hermes/pull/1855#issuecomment-3659386410
+  return s.includes('[native code]') || s.includes(`[bytecode]`) // Static Hermes has [bytecode] for contrib, which includes TextEncoder/TextDecoder
+}
+
 if (!haveNativeBuffer && isNative(() => {})) isNative = () => false // e.g. XS, we don't want false positives
 
 export const nativeEncoder = isNative(TextEncoder) ? new TextEncoder() : null
 export const nativeDecoder = isNative(TextDecoder)
   ? new TextDecoder('utf-8', { ignoreBOM: true })
   : null
-export const nativeBuffer = haveNativeBuffer ? Buffer : null
-export const isHermes = Boolean(globalThis.HermesInternal)
-export const isDeno = Boolean(globalThis.Deno)
-export const isLE = new Uint8Array(Uint16Array.of(258).buffer)[0] === 2
 
 // Actually windows-1252, compatible with ascii and latin1 decoding
 // Beware that on non-latin1, i.e. on windows-1252, this is broken in ~all Node.js versions released
 // in 2025 due to a regression, so we call it Latin1 as it's usable only for that
 let nativeDecoderLatin1impl = null
-if (isNative(TextDecoder)) {
+if (nativeDecoder) {
   // Not all barebone engines with TextDecoder support something except utf-8, detect
   try {
     nativeDecoderLatin1impl = new TextDecoder('latin1', { ignoreBOM: true })
@@ -116,3 +124,7 @@ export function decode2string(arr, start, end, m) {
 
   return decodePart(arr, start, end, m)
 }
+
+export function assert(condition, msg) {
+  if (!condition) throw new Error(msg)
+}

package/fallback/encoding.js

@@ -0,0 +1,290 @@
+// We can't return native TextDecoder if it's present, as Node.js one is broken on windows-1252 and we fix that
+// We are also faster than Node.js built-in on both TextEncoder and TextDecoder
+
+import { utf16toString, utf16toStringLoose } from '@exodus/bytes/utf16.js'
+import { utf8fromStringLoose, utf8toString, utf8toStringLoose } from '@exodus/bytes/utf8.js'
+import { createSinglebyteDecoder } from '@exodus/bytes/single-byte.js'
+import labels from './encoding.labels.js'
+import { unfinishedBytes } from './encoding.util.js'
+
+const E_OPTIONS = 'The "options" argument must be of type object'
+const E_ENCODING = 'Unknown encoding'
+const replacementChar = '\uFFFD'
+
+const E_MULTI =
+  'Legacy multi-byte encodings are disabled in /encoding-lite.js, use /encoding.js for full encodings range support'
+const multibyteSet = new Set(['big5', 'euc-kr', 'euc-jp', 'iso-2022-jp', 'shift_jis', 'gbk', 'gb18030']) // prettier-ignore
+let createMultibyteDecoder
+
+export function setMultibyteDecoder(createDecoder) {
+  createMultibyteDecoder = createDecoder
+}
+
+let labelsMap
+
+// Warning: unlike whatwg-encoding, returns lowercased labels
+// Those are case-insensitive and that's how TextDecoder encoding getter normalizes them
+// https://encoding.spec.whatwg.org/#names-and-labels
+export function normalizeEncoding(label) {
+  // fast path
+  if (label === 'utf-8' || label === 'utf8' || label === 'UTF-8' || label === 'UTF8') return 'utf-8'
+  if (label === 'windows-1252' || label === 'ascii' || label === 'latin1') return 'windows-1252'
+  // full map
+  if (/[^\w\t\n\f\r .:-]/i.test(label)) return null // must be ASCII (with ASCII whitespace)
+  const low = `${label}`.trim().toLowerCase()
+  if (Object.hasOwn(labels, low)) return low
+  if (!labelsMap) {
+    labelsMap = new Map()
+    for (const [label, aliases] of Object.entries(labels)) {
+      for (const alias of aliases) labelsMap.set(alias, label)
+    }
+  }
+
+  const mapped = labelsMap.get(low)
+  if (mapped) return mapped
+  return null
+}
+
+const define = (obj, key, value) => Object.defineProperty(obj, key, { value, writable: false })
+
+const fromSource = (x) => {
+  if (x instanceof Uint8Array) return x
+  if (x instanceof ArrayBuffer) return new Uint8Array(x)
+  if (ArrayBuffer.isView(x)) return new Uint8Array(x.buffer, x.byteOffset, x.byteLength)
+  if (globalThis.SharedArrayBuffer && x instanceof SharedArrayBuffer) return new Uint8Array(x)
+  throw new TypeError('Argument must be a SharedArrayBuffer, ArrayBuffer or ArrayBufferView')
+}
+
+function unicodeDecoder(encoding, loose) {
+  if (encoding === 'utf-8') return loose ? utf8toStringLoose : utf8toString // likely
+  const form = encoding === 'utf-16le' ? 'uint8-le' : 'uint8-be'
+  return loose ? (u) => utf16toStringLoose(u, form) : (u) => utf16toString(u, form)
+}
+
+export class TextDecoder {
+  #decode
+  #unicode
+  #multibyte
+  #chunk
+  #canBOM
+
+  constructor(encoding = 'utf-8', options = {}) {
+    if (typeof options !== 'object') throw new TypeError(E_OPTIONS)
+    const enc = normalizeEncoding(encoding)
+    if (!enc || enc === 'replacement') throw new RangeError(E_ENCODING)
+    define(this, 'encoding', enc)
+    define(this, 'fatal', Boolean(options.fatal))
+    define(this, 'ignoreBOM', Boolean(options.ignoreBOM))
+    this.#unicode = enc === 'utf-8' || enc === 'utf-16le' || enc === 'utf-16be'
+    this.#multibyte = !this.#unicode && multibyteSet.has(enc)
+    this.#canBOM = this.#unicode && !this.ignoreBOM
+  }
+
+  get [Symbol.toStringTag]() {
+    return 'TextDecoder'
+  }
+
+  decode(input, options = {}) {
+    if (typeof options !== 'object') throw new TypeError(E_OPTIONS)
+    const stream = Boolean(options.stream)
+    let u = input === undefined ? new Uint8Array() : fromSource(input)
+
+    if (this.#unicode) {
+      let prefix
+      if (this.#chunk) {
+        if (u.length === 0) {
+          if (stream) return '' // no change
+          u = this.#chunk // process as final chunk to handle errors and state changes
+        } else if (u.length < 3) {
+          // No reason to bruteforce offsets, also it's possible this doesn't yet end the sequence
+          const a = new Uint8Array(u.length + this.#chunk.length)
+          a.set(this.#chunk)
+          a.set(u, this.#chunk.length)
+          u = a
+        } else {
+          // Slice off a small portion of u into prefix chunk so we can decode them separately without extending array size
+          const t = new Uint8Array(this.#chunk.length + 3) // We have 1-3 bytes and need 1-3 more bytes
+          t.set(this.#chunk)
+          t.set(u.subarray(0, 3), this.#chunk.length)
+
+          // Stop at the first offset where unfinished bytes reaches 0 or fits into u
+          // If that doesn't happen (u too short), just concat chunk and u completely
+          for (let i = 1; i <= 3; i++) {
+            const unfinished = unfinishedBytes(t, this.#chunk.length + i, this.encoding) // 0-3
+            if (unfinished <= i) {
+              // Always reachable at 3, but we still need 'unfinished' value for it
+              const add = i - unfinished // 0-3
+              prefix = add > 0 ? t.subarray(0, this.#chunk.length + add) : this.#chunk
+              if (add > 0) u = u.subarray(add)
+              break
+            }
+          }
+        }
+
+        this.#chunk = null
+      } else if (u.byteLength === 0) {
+        if (!stream) this.#canBOM = !this.ignoreBOM
+        return ''
+      }
+
+      // For non-stream utf-8 we don't have to do this as it matches utf8toStringLoose already
+      // For non-stream loose utf-16 we still have to do this as this API supports uneven byteLength unlike utf16toStringLoose
+      let suffix = ''
+      if (stream || (!this.fatal && this.encoding !== 'utf-8')) {
+        const trail = unfinishedBytes(u, u.byteLength, this.encoding)
+        if (trail > 0) {
+          if (stream) {
+            this.#chunk = Uint8Array.from(u.subarray(-trail)) // copy
+          } else {
+            // non-fatal mode as already checked
+            suffix = replacementChar
+          }
+
+          u = u.subarray(0, -trail)
+        }
+      }
+
+      if (this.#canBOM) {
+        const bom = this.#findBom(prefix ?? u)
+        if (bom) {
+          if (stream) this.#canBOM = false
+          if (prefix) {
+            prefix = prefix.subarray(bom)
+          } else {
+            u = u.subarray(bom)
+          }
+        }
+      }
+
+      if (!this.#decode) this.#decode = unicodeDecoder(this.encoding, !this.fatal)
+      try {
+        const res = (prefix ? this.#decode(prefix) : '') + this.#decode(u) + suffix
+        if (res.length > 0 && stream) this.#canBOM = false
+
+        if (!stream) this.#canBOM = !this.ignoreBOM
+        return res
+      } catch (err) {
+        this.#chunk = null // reset unfinished chunk on errors
+        throw err
+      }
+
+      // eslint-disable-next-line no-else-return
+    } else if (this.#multibyte) {
+      if (!createMultibyteDecoder) throw new Error(E_MULTI)
+      if (!this.#decode) this.#decode = createMultibyteDecoder(this.encoding, !this.fatal) // can contain state!
+      return this.#decode(u, stream)
+    } else {
+      if (!this.#decode) this.#decode = createSinglebyteDecoder(this.encoding, !this.fatal)
+      return this.#decode(u)
+    }
+  }
+
+  #findBom(u) {
+    switch (this.encoding) {
+      case 'utf-8':
+        return u.byteLength >= 3 && u[0] === 0xef && u[1] === 0xbb && u[2] === 0xbf ? 3 : 0
+      case 'utf-16le':
+        return u.byteLength >= 2 && u[0] === 0xff && u[1] === 0xfe ? 2 : 0
+      case 'utf-16be':
+        return u.byteLength >= 2 && u[0] === 0xfe && u[1] === 0xff ? 2 : 0
+    }
+
+    throw new Error('Unreachable')
+  }
+}
+
+export class TextEncoder {
+  constructor() {
+    define(this, 'encoding', 'utf-8')
+  }
+
+  get [Symbol.toStringTag]() {
+    return 'TextEncoder'
+  }
+
+  encode(str = '') {
+    if (typeof str !== 'string') str = `${str}`
+    const res = utf8fromStringLoose(str)
+    return res.byteOffset === 0 ? res : res.slice(0) // Ensure 0-offset, to match new Uint8Array (per spec), which is non-pooled
+  }
+
+  encodeInto(str, target) {
+    if (typeof str !== 'string') str = `${str}`
+    if (!(target instanceof Uint8Array)) throw new TypeError('Target must be an Uint8Array')
+    if (target.buffer.detached) return { read: 0, written: 0 } // Until https://github.com/whatwg/encoding/issues/324 is resolved
+
+    const tlen = target.length
+    if (tlen < str.length) str = str.slice(0, tlen)
+    let u8 = utf8fromStringLoose(str)
+    let read
+    if (tlen >= u8.length) {
+      read = str.length
+    } else if (u8.length === str.length) {
+      if (u8.length > tlen) u8 = u8.subarray(0, tlen) // ascii can be truncated
+      read = u8.length
+    } else {
+      u8 = u8.subarray(0, tlen)
+      const unfinished = unfinishedBytes(u8, u8.length, 'utf-8')
+      if (unfinished > 0) u8 = u8.subarray(0, u8.length - unfinished)
+
+      // We can do this because loose str -> u8 -> str preserves length, unlike loose u8 -> str -> u8
+      // Each unpaired surrogate (1 charcode) is replaced with a single charcode
+      read = utf8toStringLoose(u8).length // FIXME: Converting back is very inefficient
+    }
+
+    try {
+      target.set(u8)
+    } catch {
+      return { read: 0, written: 0 } // see above, likely detached but no .detached property support
+    }
+
+    return { read, written: u8.length }
+  }
+}
+
+// Warning: unlike whatwg-encoding, returns lowercased labels
+// Those are case-insensitive and that's how TextDecoder encoding getter normalizes them
+export function getBOMEncoding(input) {
+  const u8 = fromSource(input) // asserts
+  if (u8.length >= 3 && u8[0] === 0xef && u8[1] === 0xbb && u8[2] === 0xbf) return 'utf-8'
+  if (u8.length < 2) return null
+  if (u8[0] === 0xff && u8[1] === 0xfe) return 'utf-16le'
+  if (u8[0] === 0xfe && u8[1] === 0xff) return 'utf-16be'
+  return null
+}
+
+// https://encoding.spec.whatwg.org/#decode
+// Warning: encoding sniffed from BOM takes preference over the supplied one
+// Warning: lossy, performs replacement, no option of throwing
+// Completely ignores encoding and even skips validation when BOM is found
+// Unlike TextDecoder public API, additionally supports 'replacement' encoding
+export function legacyHookDecode(input, fallbackEncoding = 'utf-8') {
+  let u8 = fromSource(input)
+  const bomEncoding = getBOMEncoding(u8)
+  if (bomEncoding) u8 = u8.subarray(bomEncoding === 'utf-8' ? 3 : 2)
+  const enc = bomEncoding ?? normalizeEncoding(fallbackEncoding) // "the byte order mark is more authoritative than anything else"
+
+  if (enc === 'utf-8') return utf8toStringLoose(u8)
+  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)) {
+    if (!createMultibyteDecoder) throw new Error(E_MULTI)
+    return createMultibyteDecoder(enc, true)(u8)
+  }
+
+  // https://encoding.spec.whatwg.org/#replacement-decoder
+  // On non-streaming non-fatal case, it just replaces any non-empty input with a single replacement char
+  if (enc === 'replacement') return input.byteLength > 0 ? replacementChar : ''
+
+  return createSinglebyteDecoder(enc, true)(u8)
+}

package/fallback/encoding.labels.js

@@ -2,7 +2,7 @@
 
 /* eslint-disable @exodus/export-default/named */
 // prettier-ignore
-export default {
+const labels = {
   'utf-8': ['unicode-1-1-utf-8', 'unicode11utf8', 'unicode20utf8', 'utf8', 'x-unicode20utf8'],
   ibm866: ['866', 'cp866', 'csibm866'],
   'iso-8859-2': ['csisolatin2', 'iso-ir-101', 'iso8859-2', 'iso88592', 'iso_8859-2', 'iso_8859-2:1987', 'l2', 'latin2'],
@@ -22,15 +22,6 @@ export default {
   'koi8-u': ['koi8-ru'],
   macintosh: ['csmacintosh', 'mac', 'x-mac-roman'],
   'windows-874': ['dos-874', 'iso-8859-11', 'iso8859-11', 'iso885911', 'tis-620'],
-  'windows-1250': ['cp1250', 'x-cp1250'],
-  'windows-1251': ['cp1251', 'x-cp1251'],
-  'windows-1252': ['ansi_x3.4-1968', 'ascii', 'cp1252', 'cp819', 'csisolatin1', 'ibm819', 'iso-8859-1', 'iso-ir-100', 'iso8859-1', 'iso88591', 'iso_8859-1', 'iso_8859-1:1987', 'l1', 'latin1', 'us-ascii', 'x-cp1252'],
-  'windows-1253': ['cp1253', 'x-cp1253'],
-  'windows-1254': ['cp1254', 'csisolatin5', 'iso-8859-9', 'iso-ir-148', 'iso8859-9', 'iso88599', 'iso_8859-9', 'iso_8859-9:1989', 'l5', 'latin5', 'x-cp1254'],
-  'windows-1255': ['cp1255', 'x-cp1255'],
-  'windows-1256': ['cp1256', 'x-cp1256'],
-  'windows-1257': ['cp1257', 'x-cp1257'],
-  'windows-1258': ['cp1258', 'x-cp1258'],
   'x-mac-cyrillic': ['x-mac-ukrainian'],
   gbk: ['chinese', 'csgb2312', 'csiso58gb231280', 'gb2312', 'gb_2312', 'gb_2312-80', 'iso-ir-58', 'x-gbk'],
   gb18030: [],
@@ -44,3 +35,12 @@ export default {
   'utf-16le': ['csunicode', 'iso-10646-ucs-2', 'ucs-2', 'unicode', 'unicodefeff', 'utf-16'],
   'x-user-defined': [],
 }
+
+for (let i = 0; i < 9; i++) labels[`windows-125${i}`] = [`cp125${i}`, `x-cp125${i}`]
+
+// prettier-ignore
+labels['windows-1252'].push('ansi_x3.4-1968', 'ascii', 'cp819', 'csisolatin1', 'ibm819', 'iso-8859-1', 'iso-ir-100', 'iso8859-1', 'iso88591', 'iso_8859-1', 'iso_8859-1:1987', 'l1', 'latin1', 'us-ascii')
+// prettier-ignore
+labels['windows-1254'].push('csisolatin5', 'iso-8859-9', 'iso-ir-148', 'iso8859-9', 'iso88599', 'iso_8859-9', 'iso_8859-9:1989', 'l5', 'latin5')
+
+export default labels

package/fallback/multi-byte.js

@@ -319,7 +319,6 @@ const mappers = {
 }
 
 export const isAsciiSuperset = (enc) => enc !== 'iso-2022-jp' // all others are ASCII supersets and can use fast path
-export const multibyteSupported = (enc) => Object.hasOwn(mappers, enc) || enc === 'big5'
 
 export function multibyteDecoder(enc, loose = false) {
   if (enc === 'big5') return big5decoder(loose)
@@ -337,7 +336,7 @@ export function multibyteDecoder(enc, loose = false) {
           // Decoders big5, euc-jp, euc-kr, shift_jis, gb18030 / gbk - all clear state before throwing unless EOF, so not affected
           // iso-2022-jp is the only tricky one one where this !stream check matters in non-stream mode
           if (!stream) mapper = null // destroy state, effectively the same as 'do not flush' = false, but early
-          throw new Error(E_STRICT)
+          throw new TypeError(E_STRICT)
         }
 
     let res = ''
@@ -400,7 +399,7 @@ function big5decoder(loose) {
       : () => {
           pushback.length = 0 // the queue is cleared on returning an error
           // Lead is always already cleared before throwing
-          throw new Error(E_STRICT)
+          throw new TypeError(E_STRICT)
         }
 
     let res = ''

package/fallback/multi-byte.table.js

@@ -1,5 +1,5 @@
-import { fromBase64url } from '@exodus/bytes/base64.js' // eslint-disable-line @exodus/import/no-unresolved
-import { utf16toString } from '@exodus/bytes/utf16.js' // eslint-disable-line @exodus/import/no-unresolved
+import { fromBase64url } from '@exodus/bytes/base64.js'
+import { utf16toString } from '@exodus/bytes/utf16.js'
 import loadEncodings from './multi-byte.encodings.cjs'
 import { to16input } from './utf16.js'
 

package/fallback/single-byte.encodings.js

@@ -1,45 +1,61 @@
-// See tests/fixtures/encodings/single-byte/dump.js for generator
+// See tests/encoding/fixtures/single-byte/dump.js for generator
 
-const c =
-  '\x80\x81\x82\x83\x84\x85\x86\x87\x88\x89\x8A\x8B\x8C\x8D\x8E\x8F\x90\x91\x92\x93\x94\x95\x96\x97\x98\x99\x9A\x9B\x9C\x9D\x9E\x9F\xA0'
-const k8a = '─│┌┐└┘├┤┬┴┼▀▄█▌▐░▒▓⌠■∙√≈≤≥\xA0⌡°²·÷═║╒ё'
-const k8b = '©юабцдефгхийклмнопярстужвьызшэщчъЮАБЦДЕФГХИЙКЛМНОПЯРСТУЖВЬЫЗШЭЩЧЪ'
-const i2 = 'żŔÁÂĂÄĹĆÇČÉĘËĚÍÎĎĐŃŇÓÔŐÖ×ŘŮÚŰÜÝŢßŕáâăäĺćçčéęëěíîďđńňóôőö÷řůúűüýţ˙'
-const ch = 'АБВГДЕЖЗИЙКЛМНОПРСТУФХЦЧШЩЪЫЬЭЮЯ'
-const cl1 = 'абвгдежзийклмнопрстуфхцчшщъыьэю'
-const i8 = 'אבגדהוזחטיךכלםמןנסעףפץצקרשת'
-const p = '€\x81‚ƒ„…†‡'
-const s = 'µ¶·ø¹ŗ»¼½¾æĄĮĀĆÄÅĘĒČÉŹĖĢĶĪĻŠŃŅÓŌÕÖ×ŲŁŚŪÜŻŽßąįāćäåęēčéźėģķīļšńņóōõö÷ųłśūüżž'
-const f = (n) => '\uFFFD'.repeat(n)
+const r = 0xff_fd
+const e = (x) => new Array(x).fill(1)
+const h = (x) => new Array(x).fill(r)
+
+/* eslint-disable unicorn/numeric-separators-style, @exodus/export-default/named */
+
+// Common ranges
+
+// prettier-ignore
+const k8a = [9345,2,10,4,4,4,4,8,8,8,8,68,4,4,4,4,1,1,1,-627,640,-903,1,46,28,1,-8645,8833,-8817,2,5,64,9305,1,1,-8449]
+// prettier-ignore
+const k8b = [-30,1,21,-18,1,15,-17,18,-13,...e(7),16,-15,1,1,1,-13,-4,26,-1,-20,17,5,-4,-2,3]
+const p1 = [8237, -8235, 8089, -7816, 7820, 8, -6, 1]
+const p2 = [-99, 12, 20, -12, 17, 37, -29, 2]
+// prettier-ignore
+const p3 = [1,1,65,-63,158,-156,1,1,1,40,30,42,-46,6,-66,1,83,-6,-6,-67,176,...p2,-114,121,-119,1,1,155,-49,25,16,-142,159,2,-158,38,42,-46,6,-35,1,52,-6,-6,-36,145,...p2,-83,90,-88,1,1,124,-49,25,16,-111,128,2]
+const i0 = e(33)
+// prettier-ignore
+const i2 = [-40,-147,1,64,-62,117,-51,-63,69,-67,79,-77,79,-77,1,64,2,51,4,-116,1,124,-122,1,129,22,-148,150,-148,1,133,-131,118,-116,1,33,-31,86,-51,-32,38,-36,48,-46,48,-46,1,33,2,51,4,-85,1,93,-91,1,98,22,-117,119,-117,1,102,374]
+const i4a = [-75, -63, ...e(5), 104, -34, -67, 79, -77, 75, -73, 1]
+const i4b = [34, -32, ...e(5), 73, -34, -36, 48, -46, 44, -42, 1]
+const i7 = [721, 1, 1, -719, 721, -719, 721, ...e(19), r, 2, ...e(43), r]
+const i8 = [...e(26), r, r, 6692, 1, r]
+const w0 = [8237, -8235, 8089, -8087, 8091, 8, -6, 1, -8089, 8104]
+const w8 = [8072, 1, 3, 1, 5, -15, 1]
+const w1 = [...w8, -7480, 7750, -8129, 7897, -7911, -182]
+const w3 = [...w8, -8060, 8330, -8328, 8096, -8094]
+const m0 = [8558, -8328, 8374, -66, -8539, 16, 8043, -8070]
 
-/* eslint-disable @exodus/export-default/named */
 // prettier-ignore
 export default {
-  ibm866: ch + "абвгдежзийклмноп░▒▓│┤╡╢╖╕╣║╗╝╜╛┐└┴┬├─┼╞╟╚╔╩╦╠═╬╧╨╤╥╙╘╒╓╫╪┘┌█▄▌▐▀рстуфхцчшщъыьэюяЁёЄєЇїЎў°∙·√№¤■\xA0",
-  "iso-8859-10": c + "ĄĒĢĪĨĶ§ĻĐŠŦŽ\xADŪŊ°ąēģīĩķ·ļđšŧž―ūŋĀÁÂÃÄÅÆĮČÉĘËĖÍÎÏÐŅŌÓÔÕÖŨØŲÚÛÜÝÞßāáâãäåæįčéęëėíîïðņōóôõöũøųúûüýþĸ",
-  "iso-8859-13": c + "”¢£¤„¦§Ø©Ŗ«¬\xAD®Æ°±²³“" + s + "’",
-  "iso-8859-14": c + "Ḃḃ£ĊċḊ§Ẁ©ẂḋỲ\xAD®ŸḞḟĠġṀṁ¶ṖẁṗẃṠỳẄẅṡÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏŴÑÒÓÔÕÖṪØÙÚÛÜÝŶßàáâãäåæçèéêëìíîïŵñòóôõöṫøùúûüýŷ",
-  "iso-8859-15": c + "¡¢£€¥Š§š©ª«¬\xAD®¯°±²³Žµ¶·ž¹º»ŒœŸ",
-  "iso-8859-16": c + "ĄąŁ€„Š§š©Ș«Ź\xADźŻ°±ČłŽ”¶·žčș»ŒœŸżÀÁÂĂÄĆÆÇÈÉÊËÌÍÎÏĐŃÒÓÔŐÖŚŰÙÚÛÜĘȚßàáâăäćæçèéêëìíîïđńòóôőöśűùúûüęț",
-  "iso-8859-2": c + "Ą˘Ł¤ĽŚ§¨ŠŞŤŹ\xADŽŻ°ą˛ł´ľśˇ¸šşťź˝ž" + i2,
-  "iso-8859-3": c + "Ħ˘£¤\uFFFDĤ§¨İŞĞĴ\xAD\uFFFDŻ°ħ²³´µĥ·¸ışğĵ½\uFFFDżÀÁÂ\uFFFDÄĊĈÇÈÉÊËÌÍÎÏ\uFFFDÑÒÓÔĠÖ×ĜÙÚÛÜŬŜßàáâ\uFFFDäċĉçèéêëìíîï\uFFFDñòóôġö÷ĝùúûüŭŝ˙",
-  "iso-8859-4": c + "ĄĸŖ¤ĨĻ§¨ŠĒĢŦ\xADŽ¯°ą˛ŗ´ĩļˇ¸šēģŧŊžŋĀÁÂÃÄÅÆĮČÉĘËĖÍÎĪĐŅŌĶÔÕÖ×ØŲÚÛÜŨŪßāáâãäåæįčéęëėíîīđņōķôõö÷øųúûüũū˙",
-  "iso-8859-5": c + "ЁЂЃЄЅІЇЈЉЊЋЌ\xADЎЏ" + ch + cl1 + "я№ёђѓєѕіїјљњћќ§ўџ",
-  "iso-8859-6": c + f(3) + "¤" + f(7) + "،\xAD" + f(13) + "؛" + f(3) + "؟\uFFFDءآأؤإئابةتثجحخدذرزسشصضطظعغ" + f(5) + "ـفقكلمنهوىي\u064B\u064C\u064D\u064E\u064F\u0650\u0651\u0652" + f(13),
-  "iso-8859-7": c + "‘’£€₯¦§¨©ͺ«¬\xAD\uFFFD―°±²³΄΅Ά·ΈΉΊ»Ό½ΎΏΐΑΒΓΔΕΖΗΘΙΚΛΜΝΞΟΠΡ\uFFFDΣΤΥΦΧΨΩΪΫάέήίΰαβγδεζηθικλμνξοπρςστυφχψωϊϋόύώ\uFFFD",
-  "iso-8859-8": c + "\uFFFD¢£¤¥¦§¨©×«¬\xAD®¯°±²³´µ¶·¸¹÷»¼½¾" + f(32) + "‗" + i8 + f(2) + "\u200E\u200F\uFFFD",
-  "koi8-r": k8a + "╓╔╕╖╗╘╙╚╛╜╝╞╟╠╡Ё╢╣╤╥╦╧╨╩╪╫╬" + k8b,
-  "koi8-u": k8a + "є╔ії╗╘╙╚╛ґў╞╟╠╡ЁЄ╣ІЇ╦╧╨╩╪ҐЎ" + k8b,
-  macintosh: "ÄÅÇÉÑÖÜáàâäãåçéèêëíìîïñóòôöõúùûü†°¢£§•¶ß®©™´¨≠ÆØ∞±≤≥¥µ∂∑∏π∫ªºΩæø¿¡¬√ƒ≈∆«»…\xA0ÀÃÕŒœ–—“”‘’÷◊ÿŸ⁄€‹›fifl‡·‚„‰ÂÊÁËÈÍÎÏÌÓÔ\uF8FFÒÚÛÙıˆ˜¯˘˙˚¸˝˛ˇ",
-  "windows-1250": "€\x81‚\x83„…†‡\x88‰Š‹ŚŤŽŹ\x90‘’“”•–—\x98™š›śťžź\xA0ˇ˘Ł¤Ą¦§¨©Ş«¬\xAD®Ż°±˛ł´µ¶·¸ąş»Ľ˝ľ" + i2,
-  "windows-1251": "ЂЃ‚ѓ„…†‡€‰Љ‹ЊЌЋЏђ‘’“”•–—\x98™љ›њќћџ\xA0ЎўЈ¤Ґ¦§Ё©Є«¬\xAD®Ї°±Ііґµ¶·ё№є»јЅѕї" + ch + cl1 + 'я',
-  "windows-1252": p + "ˆ‰Š‹Œ\x8DŽ\x8F\x90‘’“”•–—˜™š›œ\x9DžŸ",
-  "windows-1253": p + "\x88‰\x8A‹\x8C\x8D\x8E\x8F\x90‘’“”•–—\x98™\x9A›\x9C\x9D\x9E\x9F\xA0΅Ά£¤¥¦§¨©\uFFFD«¬\xAD®―°±²³΄µ¶·ΈΉΊ»Ό½ΎΏΐΑΒΓΔΕΖΗΘΙΚΛΜΝΞΟΠΡ\uFFFDΣΤΥΦΧΨΩΪΫάέήίΰαβγδεζηθικλμνξοπρςστυφχψωϊϋόύώ\uFFFD",
-  "windows-1254": p + "ˆ‰Š‹Œ\x8D\x8E\x8F\x90‘’“”•–—˜™š›œ\x9D\x9EŸ\xA0¡¢£¤¥¦§¨©ª«¬\xAD®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏĞÑÒÓÔÕÖ×ØÙÚÛÜİŞßàáâãäåæçèéêëìíîïğñòóôõö÷øùúûüış",
-  "windows-1255": p + "ˆ‰\x8A‹\x8C\x8D\x8E\x8F\x90‘’“”•–—˜™\x9A›\x9C\x9D\x9E\x9F\xA0¡¢£₪¥¦§¨©×«¬\xAD®¯°±²³´µ¶·¸¹÷»¼½¾¿\u05B0\u05B1\u05B2\u05B3\u05B4\u05B5\u05B6\u05B7\u05B8\u05B9\u05BA\u05BB\u05BC\u05BD־\u05BF׀\u05C1\u05C2׃װױײ׳״" + f(7) + i8 + f(2) + "\u200E\u200F\uFFFD",
-  "windows-1256": "€پ‚ƒ„…†‡ˆ‰ٹ‹Œچژڈگ‘’“”•–—ک™ڑ›œ\u200C\u200Dں\xA0،¢£¤¥¦§¨©ھ«¬\xAD®¯°±²³´µ¶·¸¹؛»¼½¾؟ہءآأؤإئابةتثجحخدذرزسشصض×طظعغـفقكàلâمنهوçèéêëىيîï\u064B\u064C\u064D\u064Eô\u064F\u0650÷\u0651ù\u0652ûü\u200E\u200Fے",
-  "windows-1257": "€\x81‚\x83„…†‡\x88‰\x8A‹\x8C¨ˇ¸\x90‘’“”•–—\x98™\x9A›\x9C¯˛\x9F\xA0\uFFFD¢£¤\uFFFD¦§Ø©Ŗ«¬\xAD®Æ°±²³´" + s + "˙",
-  "windows-1258": p + "ˆ‰\x8A‹Œ\x8D\x8E\x8F\x90‘’“”•–—˜™\x9A›œ\x9D\x9EŸ\xA0¡¢£¤¥¦§¨©ª«¬\xAD®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂĂÄÅÆÇÈÉÊË\u0300ÍÎÏĐÑ\u0309ÓÔƠÖ×ØÙÚÛÜƯ\u0303ßàáâăäåæçèéêë\u0301íîïđñ\u0323óôơö÷øùúûüư₫",
-  "windows-874": "€\x81\x82\x83\x84…\x86\x87\x88\x89\x8A\x8B\x8C\x8D\x8E\x8F\x90‘’“”•–—\x98\x99\x9A\x9B\x9C\x9D\x9E\x9F\xA0กขฃคฅฆงจฉชซฌญฎฏฐฑฒณดตถทธนบปผฝพฟภมยรฤลฦวศษสหฬอฮฯะ\u0E31าำ\u0E34\u0E35\u0E36\u0E37\u0E38\u0E39\u0E3A" + f(4) + "฿เแโใไๅๆ\u0E47\u0E48\u0E49\u0E4A\u0E4B\u0E4C\u0E4D\u0E4E๏๐๑๒๓๔๕๖๗๘๙๚๛" + f(4),
-  "x-mac-cyrillic": ch + "†°Ґ£§•¶І®©™Ђђ≠Ѓѓ∞±≤≥іµґЈЄєЇїЉљЊњјЅ¬√ƒ≈∆«»…\xA0ЋћЌќѕ–—“”‘’÷„ЎўЏџ№Ёёя" + cl1 + "€"
+  ibm866: [913,...e(47),8530,1,1,-145,34,61,1,-12,-1,14,-18,6,6,-1,-1,-75,4,32,-8,-16,-28,60,34,1,-5,-6,21,-3,-6,-16,28,-5,1,-4,1,-12,-1,-6,1,24,-1,-82,-12,124,-4,8,4,-16,-8512,...e(15),-78,80,-77,80,-77,80,-73,80,-942,8553,-8546,8547,-260,-8306,9468,-9472],
+  'iso-8859-10': [...i0,100,14,16,8,-2,14,-143,148,-43,80,6,23,-208,189,-32,-154,85,14,16,8,-2,14,-128,133,-43,80,6,23,7831,-7850,-32,...i4a,1,1,117,7,-121,1,1,1,146,-144,154,-152,...e(5),...i4b,1,1,86,7,-90,1,1,1,115,-113,123,-121,1,1,1,1,58],
+  'iso-8859-13': [...i0,8061,-8059,1,1,8058,-8056,1,49,-47,173,-171,1,1,1,24,-22,1,1,1,8041,-8039,...p3,7835],
+  'iso-8859-14': [...i0,7522,1,-7520,103,1,7423,-7523,7641,-7639,7641,-119,231,-7749,1,202,7334,1,-7423,1,7455,1,-7563,7584,43,-42,44,-35,147,-111,1,-36,-7585,...e(15),165,-163,...e(5),7572,-7570,...e(5),153,-151,...e(16),134,-132,...e(5),7541,-7539,...e(5),122],
+  'iso-8859-15': [...i0,1,1,1,8201,-8199,187,-185,186,-184,...e(10),202,-200,1,1,199,-197,1,1,151,1,37],
+  'iso-8859-16': [...i0,100,1,60,8043,-142,-7870,-185,186,-184,367,-365,206,-204,205,1,-203,1,91,54,59,7840,-8039,1,199,-113,268,-350,151,1,37,4,-188,1,1,64,-62,66,-64,...e(9),65,51,-113,1,1,124,-122,132,22,-151,1,1,1,60,258,-315,1,1,1,33,-31,35,-33,...e(9),34,51,-82,1,1,93,-91,101,22,-120,1,1,1,29,258],
+  'iso-8859-2': [...i0,100,468,-407,-157,153,29,-179,1,184,-2,6,21,-204,208,-2,-203,85,470,-409,-142,138,29,364,-527,169,-2,6,21,355,-351,-2,...i2],
+  'iso-8859-3': [...i0,134,434,-565,1,r,128,-125,1,136,46,-64,22,-135,r,206,-203,119,-117,1,1,1,112,-110,1,121,46,-64,22,-120,r,191,-188,1,1,r,2,70,-2,-65,...e(8),r,2,1,1,1,76,-74,1,69,-67,1,1,1,144,-16,-125,1,1,1,r,2,39,-2,-34,...e(8),r,2,1,1,1,45,-43,1,38,-36,1,1,1,113,-16,380],
+  'iso-8859-4': [...i0,100,52,30,-178,132,19,-148,1,184,-78,16,68,-185,208,-206,1,85,470,-388,-163,117,19,395,-527,169,-78,16,68,-29,52,-51,...i4a,92,-26,53,7,-22,-98,1,1,1,1,154,-152,1,1,140,2,-139,...i4b,61,-26,53,7,-22,-67,1,1,1,1,123,-121,1,1,109,2,366],
+  'iso-8859-5': [...i0,865,...e(11),-863,865,...e(65),7367,-7365,...e(11),-949,951,1],
+  'iso-8859-6': [...i0,r,r,r,4,...h(7),1384,-1375,...h(13),1390,r,r,r,4,r,2,...e(25),r,r,r,r,r,6,...e(18),...h(13)],
+  'iso-8859-7': [...i0,8056,1,-8054,8201,3,-8201,1,1,1,721,-719,1,1,r,8040,-8037,1,1,1,721,1,1,-719,...i7],
+  'iso-8859-8': [...i0,r,2,...e(7),46,-44,...e(14),62,-60,1,1,1,...h(32),8025,-6727,...i8],
+  'koi8-r': [...k8a,8450,...e(14),-8544,8545,...e(10),-9411,933,...k8b,-28,...k8b],
+  'koi8-u': [...k8a,3,8448,-8446,1,8448,1,1,1,1,-8394,-51,8448,1,1,1,-8544,3,8543,-8541,1,8543,1,1,1,1,-8410,-130,-869,933,...k8b,-28,...k8b],
+  macintosh: [69,1,2,2,8,5,6,5,-1,2,2,-1,2,2,2,-1,2,1,2,-1,2,1,2,2,-1,2,2,-1,5,-1,2,1,7972,-8048,-14,1,4,8059,-8044,41,-49,-5,8313,-8302,-12,8632,-8602,18,8518,-8557,8627,1,-8640,16,8525,15,-2,-7759,7787,-8577,16,751,-707,18,-57,-30,11,...m0,32,3,18,125,1,7872,1,8,1,-5,1,-7970,9427,-9419,121,7884,104,-115,1,56007,1,-56033,-8042,8035,4,18,-8046,8,-9,10,-3,5,1,1,-3,7,1,63531,-63533,8,1,-2,88,405,22,-557,553,1,1,-546,549,-2,-20],
+  'windows-1250': [...w0,-7888,7897,-7903,10,25,-4,-233,...w8,-8060,8330,-8129,7897,-7903,10,25,-4,-218,551,17,-407,-157,96,-94,1,1,1,181,-179,1,1,1,205,-203,1,554,-409,-142,1,1,1,1,77,90,-164,130,416,-415,62,...i2],
+  'windows-1251': [899,1,7191,-7111,7115,8,-6,1,139,-124,-7207,7216,-7215,2,-1,4,67,7110,1,3,1,5,-15,1,-8060,8330,-7369,7137,-7136,2,-1,4,-959,878,80,-86,-868,1004,-1002,1,858,-856,859,-857,1,1,1,857,-855,1,853,80,59,-988,1,1,922,7365,-7362,-921,925,-83,80,2,-71,...e(63)],
+  'windows-1252': [...p1,-7515,7530,-7888,7897,-7911,-197,240,-238,1,...w1,225,-6],
+  'windows-1253': [...p1,-8089,8104,-8102,8111,-8109,1,1,1,1,...w3,1,1,1,1,741,1,-739,1,1,1,1,1,1,r,2,1,1,1,8039,-8037,1,1,1,721,-719,1,1,...i7],
+  'windows-1254': [...p1,-7515,7530,-7888,7897,-7911,-197,1,1,1,...w1,1,218,-216,...e(47),79,-77,...e(11),84,46,-127,...e(16),48,-46,...e(11),53,46],
+  'windows-1255': [...p1,-7515,7530,-8102,8111,-8109,1,1,1,1,...w8,-7480,7750,-8328,8096,-8094,...e(7),8199,-8197,1,1,1,1,46,-44,...e(14),62,-60,1,1,1,1,1265,...e(19),45,1,1,1,1,...h(7),-36,...i8],
+  'windows-1256': [8237,-6702,6556,-7816,7820,8,-6,1,-7515,7530,-6583,6592,-7911,1332,18,-16,39,6505,1,3,1,5,-15,1,-6507,6777,-6801,6569,-7911,7865,1,-6483,-1562,1388,-1386,...e(7),1557,-1555,...e(14),1378,-1376,1,1,1,1377,162,-160,...e(21),-1375,1376,1,1,1,6,1,1,1,-1379,1380,-1378,1379,1,1,1,-1377,1,1,1,1,1374,1,-1372,1,1372,1,1,1,-1370,1371,1,-1369,1370,-1368,1369,-1367,1,7954,1,-6461],
+  'windows-1257': [...w0,-8102,8111,-8109,28,543,-527,-40,...w3,19,556,-572,1,r,2,1,1,r,2,1,49,-47,173,-171,1,1,1,24,-22,...e(5),...p3,347],
+  'windows-1258': [...p1,-7515,7530,-8102,8111,-7911,-197,1,1,1,...w8,-7480,7750,-8328,8096,-7911,-182,1,218,-216,...e(34),64,-62,...e(7),565,-563,1,1,65,-63,568,-566,1,204,-202,1,1,1,1,1,1,211,340,-548,1,1,1,33,-31,...e(7),534,-532,1,1,34,-32,562,-560,1,173,-171,1,1,1,1,1,1,180,7931],
+  'windows-874': [8237,-8235,1,1,1,8098,-8096,...e(10),...w8,-8060,...e(8),3425,...e(57),r,r,r,r,5,...e(28),r,r,r,r],
+  'x-mac-cyrillic': [913,...e(31),7153,-8048,992,-1005,4,8059,-8044,848,-856,-5,8313,-7456,80,7694,-7773,80,7627,-8557,8627,1,-7695,-929,988,-137,-4,80,-77,80,-78,80,-79,80,-2,-83,-857,...m0,875,80,-79,80,-7,7102,1,8,1,-5,1,-7970,7975,-7184,80,-79,80,7351,-7445,80,-2,-31,...e(30),7262]
 }

package/fallback/single-byte.js

@@ -4,19 +4,21 @@ import { decode2string } from './_utils.js'
 
 export const E_STRICT = 'Input is not well-formed for this encoding'
 const xUserDefined = 'x-user-defined'
+const iso8i = 'iso-8859-8-i'
 
 export const assertEncoding = (encoding) => {
-  if (Object.hasOwn(encodings, encoding) || encoding === xUserDefined) return
+  if (Object.hasOwn(encodings, encoding) || encoding === xUserDefined || encoding === iso8i) return
   throw new RangeError('Unsupported encoding')
 }
 
+const r = 0xff_fd
+
 function getEncoding(encoding) {
   assertEncoding(encoding)
-  if (encoding === xUserDefined) {
-    return Array.from({ length: 128 }, (_, i) => String.fromCharCode(0xf7_80 + i)).join('')
-  }
-
-  return encodings[encoding]
+  if (encoding === xUserDefined) return Array.from({ length: 128 }, (_, i) => 0xf7_80 + i)
+  if (encoding === iso8i) encoding = 'iso-8859-8'
+  let prev = 127
+  return encodings[encoding].map((x) => (x === r ? x : (prev += x))) // eslint-disable-line no-return-assign
 }
 
 const mappers = new Map()
@@ -28,13 +30,13 @@ export function encodingMapper(encoding) {
   const cached = mappers.get(encoding)
   if (cached) return cached
 
-  const incomplete = getEncoding(encoding).includes('\uFFFD')
+  const codes = getEncoding(encoding)
+  const incomplete = codes.includes(0xff_fd)
   let map
   const mapper = (arr, start = 0) => {
     if (!map) {
-      map = Uint16Array.from({ length: 256 }, (_, i) => i) // Unicode subset
-      const strings = getEncoding(encoding).split('')
-      map.set(Uint16Array.from(strings.map((x) => x.charCodeAt(0))), 128)
+      map = new Uint16Array(256).map((_, i) => i) // Unicode subset
+      map.set(Uint16Array.from(codes), 128)
     }
 
     const o = Uint16Array.from(start === 0 ? arr : arr.subarray(start)) // copy to modify in-place, also those are 16-bit now
@@ -63,12 +65,13 @@ export function encodingDecoder(encoding) {
   if (cached) return cached
 
   let strings
-  const incomplete = getEncoding(encoding).includes('\uFFFD')
+  const codes = getEncoding(encoding)
+  const incomplete = codes.includes(0xff_fd)
   const decoder = (arr, loose = false) => {
     if (!strings) {
-      const part = getEncoding(encoding).split('')
-      strings = Array.from({ length: 128 }, (_, i) => String.fromCharCode(i)).concat(part)
-      while (strings.length < 256) strings.push(String.fromCharCode(strings.length))
+      const allCodes = Array.from({ length: 128 }, (_, i) => i).concat(codes)
+      while (allCodes.length < 256) allCodes.push(allCodes.length)
+      strings = allCodes.map((c) => String.fromCharCode(c))
     }
 
     const prefix = decodeLatin1(arr, 0, asciiPrefix(arr))

package/hex.d.ts

@@ -0,0 +1,22 @@
+/// <reference types="node" />
+
+import type { OutputFormat, Uint8ArrayBuffer } from './array.js';
+
+/**
+ * Encodes a Uint8Array to a lowercase hex string
+ * @param arr - The input bytes
+ * @returns The hex encoded string
+ */
+export function toHex(arr: Uint8ArrayBuffer): string;
+
+/**
+ * Decodes a hex string to bytes
+ * Unlike Buffer.from(), throws on invalid input
+ * @param str - The hex encoded string (case-insensitive)
+ * @param format - Output format (default: 'uint8')
+ * @returns The decoded bytes
+ */
+export function fromHex(str: string, format?: 'uint8'): Uint8ArrayBuffer;
+export function fromHex(str: string, format: 'buffer'): Buffer;
+export function fromHex(str: string, format?: OutputFormat): Uint8ArrayBuffer | Buffer;
+