Fix attribute names changed (#371)

Closes #369
website-scraper · Aug 28, 2019 · 6988b86 · 6988b86
1 parent 6c31bf5
commit 6988b86
Show file tree

Hide file tree

Showing 2 changed files with 22 additions and 1 deletion.
diff --git a/lib/resource-handler/html/index.js b/lib/resource-handler/html/index.js
@@ -81,7 +81,8 @@ function prepareToLoad ($, resource) {
 
 function loadTextToCheerio (text) {
 	return cheerio.load(text, {
-		decodeEntities: false
+		decodeEntities: false,
+		lowerCaseAttributeNames: false,
 	});
 }
 

diff --git a/test/unit/resource-handler/html.test.js b/test/unit/resource-handler/html.test.js
@@ -144,6 +144,26 @@ describe('ResourceHandler: Html', () => {
 		});
 	});
 
+	it('should not update attribute names to lowercase', () => {
+		htmlHandler = new HtmlHandler({ sources: [] }, {downloadChildrenPaths});
+		const html = `
+			<html>
+			<body>
+				<svg width="50" height="50" viewBox="0 0 100 100">
+  					<ellipse cx="50" cy="50" rx="50" ry="50"></ellipse>
+				</svg>
+			</body>
+			</html>
+		`;
+
+		const resource = new Resource('http://example.com', 'index.html');
+		resource.setText(html);
+
+		return htmlHandler.handle(resource).then(() => {
+			resource.getText().should.containEql('viewBox="0 0 100 100"');
+		});
+	});
+
 	it('should call downloadChildrenResources for each source', () => {
 		const sources = [{ selector: 'img', attr: 'src' }];
 		htmlHandler = new HtmlHandler({sources}, {downloadChildrenPaths});