fix: TextNode text property is not decoded (fixes #134)

taoqf · Jun 30, 2021 · 7dea8d7 · 7dea8d7
1 parent e35e4ff
commit 7dea8d7
Show file tree

Hide file tree

Showing 3 changed files with 63 additions and 32 deletions.
diff --git a/src/nodes/html.ts b/src/nodes/html.ts
@@ -458,7 +458,7 @@ export default class HTMLElement extends Node {
 				if ((node as TextNode).isWhitespace) {
 					return;
 				}
-				node.rawText = (<TextNode>node).trimmedText;
+				node.rawText = (<TextNode>node).trimmedRawText;
 			} else if (node.nodeType === NodeType.ELEMENT_NODE) {
 				(node as HTMLElement).removeWhitespace();
 			}

diff --git a/src/nodes/text.ts b/src/nodes/text.ts
@@ -1,14 +1,18 @@
 import NodeType from './type';
 import Node from './node';
 import HTMLElement from './html';
+import { trimText } from '../utils';
+import { decode } from 'he';
+
 
 /**
  * TextNode to contain a text element in DOM tree.
  * @param {string} value [description]
  */
 export default class TextNode extends Node {
-	public constructor(public rawText: string, parentNode: HTMLElement) {
+	public constructor(rawText: string, parentNode: HTMLElement) {
 		super(parentNode);
+		this._rawText = rawText;
 	}
 
 	/**
@@ -17,42 +21,38 @@ export default class TextNode extends Node {
 	 */
 	public nodeType = NodeType.TEXT_NODE;
 
+	private _rawText: string;
+	private _trimmedRawText?: string;
 	private _trimmedText?: string;
 
+	public get rawText() {
+	  return this._rawText;
+  }
+
+  /**
+   * Set rawText and invalidate trimmed caches
+   */
+	public set rawText(text: string) {
+	  this._rawText = text;
+	  this._trimmedRawText = void 0;
+	  this._trimmedText = void 0;
+  }
+
+	/**
+	 * Returns raw text with all whitespace trimmed except single leading/trailing non-breaking space
+	 */
+	public get trimmedRawText() {
+		if (this._trimmedRawText !== undefined) return this._trimmedRawText;
+		this._trimmedRawText = trimText(this.rawText);
+		return this._trimmedRawText;
+	}
+
 	/**
 	 * Returns text with all whitespace trimmed except single leading/trailing non-breaking space
 	 */
 	public get trimmedText() {
 		if (this._trimmedText !== undefined) return this._trimmedText;
-
-		const text = this.rawText;
-		let i = 0;
-		let startPos;
-		let endPos;
-
-		while (i >= 0 && i < text.length) {
-			if (/\S/.test(text[i])) {
-				if (startPos === undefined) {
-					startPos = i;
-					i = text.length;
-				} else {
-					endPos = i;
-					i = void 0;
-				}
-			}
-
-			if (startPos === undefined) i++;
-			else i--;
-		}
-
-		if (startPos === undefined) startPos = 0;
-		if (endPos === undefined) endPos = text.length - 1;
-
-		const hasLeadingSpace = startPos > 0 && /[^\S\r\n]/.test(text[startPos-1]);
-		const hasTrailingSpace = endPos < (text.length - 1) && /[^\S\r\n]/.test(text[endPos+1]);
-
-		this._trimmedText = (hasLeadingSpace ? ' ' : '') + text.slice(startPos, endPos + 1) + (hasTrailingSpace ? ' ' : '');
-
+		this._trimmedText = trimText(this.text);
 		return this._trimmedText;
 	}
 
@@ -61,7 +61,7 @@ export default class TextNode extends Node {
 	 * @return {string} text content
 	 */
 	public get text() {
-		return this.rawText;
+		return decode(this.rawText);
 	}
 
 	/**

diff --git a/src/utils.ts b/src/utils.ts
@@ -0,0 +1,31 @@
+/**
+ * Trim whitespace except single leading/trailing non-breaking space
+ */
+export function trimText(text: string): string {
+  let i = 0;
+  let startPos;
+  let endPos;
+
+  while (i >= 0 && i < text.length) {
+    if (/\S/.test(text[i])) {
+      if (startPos === undefined) {
+        startPos = i;
+        i = text.length;
+      } else {
+        endPos = i;
+        i = void 0;
+      }
+    }
+
+    if (startPos === undefined) i++;
+    else i--;
+  }
+
+  if (startPos === undefined) startPos = 0;
+  if (endPos === undefined) endPos = text.length - 1;
+
+  const hasLeadingSpace = startPos > 0 && /[^\S\r\n]/.test(text[startPos-1]);
+  const hasTrailingSpace = endPos < (text.length - 1) && /[^\S\r\n]/.test(text[endPos+1]);
+
+  return (hasLeadingSpace ? ' ' : '') + text.slice(startPos, endPos + 1) + (hasTrailingSpace ? ' ' : '');
+}