Merge branch 'release/v0.17.3'

Author: holtwick

package.json

@@ -1,7 +1,7 @@
 {
   "name": "zeed-dom",
   "type": "module",
-  "version": "0.17.2",
+  "version": "0.17.3",
   "description": "🌱 Lightweight offline DOM",
   "author": {
     "name": "Dirk Holtwick",
@@ -77,6 +77,7 @@
     "@vitest/coverage-v8": "^3.2.4",
     "c8": "^10.1.3",
     "eslint": "^9.31.0",
+    "ts-node": "^10.9.2",
     "tsup": "^8.5.0",
     "typedoc": "^0.28.7",
     "typescript": "^5.8.3",

src/htmlparser.ts

@@ -41,45 +41,63 @@ export class HtmlParser {
   parse(html: string) {
     let treatAsChars = false
     let index, match, characters
+    // Precompile regex for script/style end tags
+    let scriptEndRe: RegExp | null = null
+    let styleEndRe: RegExp | null = null
     while (html.length) {
       // comment
-      if (html.substring(0, 4) === '<!--') {
+      if (html.startsWith('<!--')) {
         index = html.indexOf('-->')
         if (index !== -1) {
           this.scanner.comment(html.substring(4, index))
-          html = html.substring(index + 3)
+          html = html.slice(index + 3)
           treatAsChars = false
         }
         else {
           treatAsChars = true
         }
       }
-
       // end tag
-      else if (html.substring(0, 2) === '</') {
-        match = this.endTagRe.exec(html)
+      else if (html.startsWith('</')) {
+        match = html.match(this.endTagRe)
         if (match) {
-          html = RegExp.rightContext
+          const matchLen = match[0].length
+          html = html.slice(matchLen)
           treatAsChars = false
-          this.parseEndTag(RegExp.lastMatch, match[1])
+          this.parseEndTag(match[0], match[1])
         }
         else {
           treatAsChars = true
         }
       }
-
       // start tag
-      else if (html.charAt(0) === '<') {
-        match = this.startTagRe.exec(html)
+      else if (html[0] === '<') {
+        match = html.match(this.startTagRe)
         if (match) {
-          html = RegExp.rightContext
+          const matchLen = match[0].length
+          html = html.slice(matchLen)
           treatAsChars = false
-          const tagName = this.parseStartTag(RegExp.lastMatch, match[1], match)
-          if (tagName === 'script' || tagName === 'style') {
-            index = html.search(new RegExp(`<\/${tagName}`, 'i'))
+          const tagName = this.parseStartTag(match[0], match[1], match)
+          if (tagName === 'script') {
+            if (!scriptEndRe)
+              scriptEndRe = /<\/script/i
+            index = html.search(scriptEndRe)
             if (index !== -1) {
               this.scanner.characters(html.substring(0, index))
-              html = html.substring(index)
+              html = html.slice(index)
+              treatAsChars = false
+            }
+            else {
+              treatAsChars = true
+            }
+          }
+          else if (tagName === 'style') {
+            if (!styleEndRe)
+              styleEndRe = /<\/style/i
+            index = html.search(styleEndRe)
+            if (index !== -1) {
+              this.scanner.characters(html.substring(0, index))
+              html = html.slice(index)
               treatAsChars = false
             }
             else {
@@ -91,31 +109,25 @@ export class HtmlParser {
           treatAsChars = true
         }
       }
-
       if (treatAsChars) {
         index = html.indexOf('<')
         let offset = index
-
         if (index === 0) {
-          // First char is a < so find the next one
-          index = html.substring(1).indexOf('<')
-          // We're at substring(1) so add 1 to the index
-          offset = offset + 1
+          index = html.indexOf('<', 1)
+          offset = 1 + (index === -1 ? html.length : index - 1)
         }
-
         if (index === -1) {
           characters = html
           html = ''
         }
         else {
           characters = html.substring(0, offset)
-          html = html.substring(offset)
+          html = html.slice(offset)
         }
-
-        if (!this.options.ignoreWhitespaceText || !/^\s*$/.test(characters))
+        // Fast whitespace check
+        if (!this.options.ignoreWhitespaceText || (characters.length && /[^\s]/.test(characters)))
           this.scanner.characters(characters)
       }
-
       treatAsChars = true
       match = null
     }