Improve decoding performance

This yields a performance improvement over indexing into `source` over and over. `source` is asserted to be a string, `psz` is monotonically increasing, and `psz >= source.length` is a good indication that you've run out of characters.

Originally proposed by @oscxc in #74.

Co-authored-by: Rand <[email protected]>

Author: steveluscher

src/cjs/index.cjs

@@ -81,7 +81,7 @@ function base (ALPHABET) {
     const size = (((source.length - psz) * FACTOR) + 1) >>> 0 // log(58) / log(256), rounded up.
     const b256 = new Uint8Array(size)
     // Process the characters.
-    while (source[psz]) {
+    while (psz < source.length) {
       // Decode character
       let carry = BASE_MAP[source.charCodeAt(psz)]
       // Invalid character

src/esm/index.js

@@ -79,7 +79,7 @@ function base (ALPHABET) {
     const size = (((source.length - psz) * FACTOR) + 1) >>> 0 // log(58) / log(256), rounded up.
     const b256 = new Uint8Array(size)
     // Process the characters.
-    while (source[psz]) {
+    while (psz < source.length) {
       // Decode character
       let carry = BASE_MAP[source.charCodeAt(psz)]
       // Invalid character

ts_src/index.ts

@@ -100,7 +100,7 @@ function base (ALPHABET: string): base.BaseConverter {
     const b256 = new Uint8Array(size)
 
     // Process the characters.
-    while (source[psz]) {
+    while (psz < source.length) {
       // Decode character
       let carry = BASE_MAP[source.charCodeAt(psz)]