feat: implement block level elements spec

taoqf · Jun 23, 2021 · 044a1bf · 044a1bf
1 parent bcc21a6
commit 044a1bf
Show file tree

Hide file tree

Showing 2 changed files with 51 additions and 18 deletions.
diff --git a/src/nodes/html.ts b/src/nodes/html.ts
@@ -29,24 +29,52 @@ export interface RawAttributes {
 
 export type InsertPosition = 'beforebegin' | 'afterbegin' | 'beforeend' | 'afterend';
 
-const kBlockElements = new Map<string, true>();
+const kBlockElements = {
+	address: true,
+	article: true,
+	aside: true,
+	blockquote: true,
+	br: true,
+	details: true,
+	dialog: true,
+	dd: true,
+	div: true,
+	dl: true,
+	dt: true,
+	fieldset: true,
+	figcaption: true,
+	figure: true,
+	footer: true,
+	form: true,
+	h1: true,
+	h2: true,
+	h3: true,
+	h4: true,
+	h5: true,
+	h6: true,
+	header: true,
+	hgroup: true,
+	hr: true,
+	li: true,
+	main: true,
+	nav: true,
+	ol: true,
+	p: true,
+	pre: true,
+	section: true,
+	table: true,
+	td: true,
+	tr: true,
+	ul: true
+};
+
+function isBlockElement(node: HTMLElement) {
+	if (node && node.tagName) {
+		return Boolean(kBlockElements[node.tagName.toLowerCase()]);
+	}
+	return false;
+}
 
-kBlockElements.set('DIV', true);
-kBlockElements.set('div', true);
-kBlockElements.set('P', true);
-kBlockElements.set('p', true);
-// ul: true,
-// ol: true,
-kBlockElements.set('LI', true);
-kBlockElements.set('li', true);
-// table: true,
-// tr: true,
-kBlockElements.set('TD', true);
-kBlockElements.set('td', true);
-kBlockElements.set('SECTION', true);
-kBlockElements.set('section', true);
-kBlockElements.set('BR', true);
-kBlockElements.set('br', true);
 
 class DOMTokenList {
 	private _set: Set<string>;
@@ -245,7 +273,7 @@ export default class HTMLElement extends Node {
 		const blocks = [currentBlock];
 		function dfs(node: Node) {
 			if (node.nodeType === NodeType.ELEMENT_NODE) {
-				if (kBlockElements.get((node as HTMLElement).rawTagName)) {
+				if (isBlockElement(node as HTMLElement)) {
 					if (currentBlock.length > 0) {
 						blocks.push(currentBlock = []);
 					}

diff --git a/test/html.js b/test/html.js
@@ -371,6 +371,11 @@ describe('HTML Parser', function () {
 				const root = parseHTML('<span>o<p>a</p><!-- my comment --></span>', { comment: true });
 				root.structuredText.should.eql('o\na');
 			});
+
+			it('should return correct structured text (block level elements)', function () {
+				const root = parseHTML('<p>content</p><span><u><h1>inside</h1><i>htm<u>l</u></i></u></span>');
+				root.structuredText.should.eql('content\ninside\nhtml');
+			});
 		});
 		describe('#set_content', function () {
 			it('set content string', function () {