fix: improve the performance of replacing returns with enclosing whit…

…espaces
josdejong · Nov 9, 2021 · de05b10 · de05b10
1 parent cbfde09
commit de05b10
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 5 deletions.
diff --git a/src/lib/utils/domUtils.js b/src/lib/utils/domUtils.js
@@ -115,7 +115,7 @@ export function traverseInnerText(element, buffer) {
   // text node
   if (element.nodeValue) {
     // remove return characters and the whitespaces surrounding those return characters
-    const trimmedValue = element.nodeValue.replace(regexReturnAndSurroundingWhitespace, '')
+    const trimmedValue = removeReturnsAndSurroundingWhitespace(element.nodeValue)
     if (trimmedValue !== '') {
       return buffer.flush() + trimmedValue
     } else {
@@ -160,7 +160,11 @@ export function traverseInnerText(element, buffer) {
 
 // regular expression matching one or multiple return characters with all their
 // enclosing white spaces
-export const regexReturnAndSurroundingWhitespace = /(\b|^)\s*\n\s*(\b|$)/g
+export function removeReturnsAndSurroundingWhitespace(text) {
+  return text.replace(/(\b|^)\s*(\b|$)/g, (match) => {
+    return /\n/.exec(match) ? '' : match
+  })
+}
 
 export function isChildOfNodeName(element, nodeName) {
   return isChildOf(element, (e) => e.nodeName.toUpperCase() === nodeName.toUpperCase())

diff --git a/src/lib/utils/domUtils.test.js b/src/lib/utils/domUtils.test.js
@@ -1,5 +1,5 @@
 import assert from 'assert'
-import { escapeHTML, regexReturnAndSurroundingWhitespace, unescapeHTML } from './domUtils.js'
+import { escapeHTML, removeReturnsAndSurroundingWhitespace, unescapeHTML } from './domUtils.js'
 
 describe('domUtils', () => {
   it('escapeHTML', () => {
@@ -22,8 +22,8 @@ describe('domUtils', () => {
 
   it('regex should match whitespace and surrounding whitespace', () => {
     assert.strictEqual(
-      'A\nB  \nC  \n  D \n\n E F'.replace(regexReturnAndSurroundingWhitespace, '*'),
-      'A*B*C*D*E F'
+      removeReturnsAndSurroundingWhitespace(' \n A\nB  \nC  \n  D \n\n E F\n '),
+      'ABCDE F'
     )
   })
 })