perf: speedup decodeLatin1 with atob

Author: ChALkeR

fallback/latin1.js

@@ -6,12 +6,19 @@ import {
   isHermes,
   isDeno,
   isLE,
+  skipWeb,
 } from './_utils.js'
 
+const { atob } = globalThis
+const { toBase64: web64 } = Uint8Array.prototype
+
 // See http://stackoverflow.com/a/22747272/680742, which says that lowest limit is in Chrome, with 0xffff args
 // On Hermes, actual max is 0x20_000 minus current stack depth, 1/16 of that should be safe
 const maxFunctionArgs = 0x20_00
 
+// toBase64+atob path is faster on everything where fromBase64 is fast
+const useLatin1atob = web64 && atob && !skipWeb
+
 export function asciiPrefix(arr) {
   let p = 0 // verified ascii bytes
   const length = arr.length
@@ -46,6 +53,18 @@ export function decodeLatin1(arr, start = 0, stop = arr.length) {
   stop |= 0
   const total = stop - start
   if (total === 0) return ''
+
+  if (
+    useLatin1atob &&
+    total >= 256 &&
+    total < 1e8 &&
+    arr.toBase64 === web64 &&
+    arr.BYTES_PER_ELEMENT === 1
+  ) {
+    const sliced = start === 0 && stop === arr.length ? arr : arr.subarray(start, stop)
+    return atob(sliced.toBase64())
+  }
+
   if (total > maxFunctionArgs) {
     let prefix = ''
     for (let i = start; i < stop; ) {

single-byte.js

@@ -105,7 +105,7 @@ export function createSinglebyteEncoder(encoding, { mode = 'fatal' } = {}) {
     if (typeof s !== 'string') throw new TypeError(E_STRING)
     if (isLatin1) {
       // max limit is to not produce base64 strings that are too long
-      if (s.length >= 1024 && s.length < 1e8 && useLatin1btoa) {
+      if (useLatin1btoa && s.length >= 1024 && s.length < 1e8) {
         try {
           return Uint8Array.fromBase64(btoa(s)) // fails on non-latin1
         } catch {