browser: use TextDecoder and TextEncoder for buffer

Doing a profiling in chrome dev tools shows that the `Buffer.toString()` and `Buffer.from(string)` is using unexpected long cpu time. With the native TextDecoder and TextEncoder it can get much faster in browsers supporting it.
On browsers not supporting TextDecoder, like Internet Explorer, `Buffer.toString()` and `Buffer.from(string)` would not be changed.

References:
feross/buffer#268
feross/buffer#60
https://developer.mozilla.org/en-US/docs/Web/API/TextDecoder
https://developer.mozilla.org/en-US/docs/Web/API/TextEncoder

lib/exceljs.bare.js

@@ -1,4 +1,6 @@
 // this bundle is built without polyfill leaving apps the freedom to add their own
+require('./utils/browser-patch-buffer');
+
 const ExcelJS = {
   Workbook: require('./doc/workbook'),
 };

lib/exceljs.browser.js

@@ -5,6 +5,7 @@ require('core-js/modules/es.object.keys');
 require('core-js/modules/es.symbol');
 require('core-js/modules/es.symbol.async-iterator');
 require('regenerator-runtime/runtime');
+require('./utils/browser-patch-buffer');
 
 const ExcelJS = {
   Workbook: require('./doc/workbook'),

lib/utils/browser-patch-buffer.js

@@ -0,0 +1,34 @@
+/* eslint-disable node/no-unsupported-features/node-builtins */
+// Note: this is included only in browser
+if (typeof TextDecoder !== 'undefined' || typeof TextEncoder !== 'undefined') {
+  const {Buffer} = require('buffer');
+  if (typeof TextEncoder !== 'undefined') {
+    const textEncoder = new TextEncoder('utf-8');
+    const {from} = Buffer;
+    Buffer.from = function(str, encoding) {
+      if (typeof str === 'string') {
+        if (encoding) {
+          encoding = encoding.toLowerCase();
+        }
+        if (!encoding || encoding === 'utf8' || encoding === 'utf-8') {
+          return from.call(this, textEncoder.encode(str).buffer);
+        }
+      }
+      return from.apply(this, arguments);
+    };
+  }
+  if (typeof TextDecoder !== 'undefined') {
+    const textDecoder =  new TextDecoder('utf-8');
+    const {toString} = Buffer.prototype;
+    Buffer.prototype.toString = function(encoding, start, end) {
+      if (encoding) {
+        encoding = encoding.toLowerCase();
+      }
+      if (!start && end === undefined &&
+        (!encoding || encoding === 'utf8' || encoding === 'utf-8')) {
+        return textDecoder.decode(this);
+      }
+      return toString.apply(this, arguments);
+    };
+  }
+}