Support non-alphanumeric characters in ToC

```js
// Note: character range in regex is roughly "word characters including accented" (eg: bublé)
const id = text
  .toLowerCase()
  .replace(/[\s-]+/g, '-')
  .replace(/[^a-z0-9\u00C0-\u024F-]/g, '');
  // korean
  .replace(/[^가-힣]/g, '');
  // Japanese hiragana, katakana, kanji
  .replace(/[^\u3041-\u3096\u30A0-\u30FF\u2E80-\u2FD5]/g, '');
```

If want to keep existing intent, the above code might be more suitable.

Reference issue comment: preactjs#754 (comment)

src/lib/content.js

@@ -25,24 +25,24 @@ export function getContent([lang, name]) {
 
 	let fallback = false;
 	const res = fetch(url)
-		.then(r => {
+		.then((r) => {
 			// fall back to english
 			if (!r.ok && lang != 'en') {
 				fallback = true;
 				return fetch(url.replace(/content\/[^/]+\//, 'content/en/'));
 			}
 			return r;
 		})
-		.then(r => {
+		.then((r) => {
 			if (r.ok) return r;
 			ext = 'md';
 			return fetch(`${path}/${r.status}.md`);
 		})
-		.then(r => r.text())
+		.then((r) => r.text())
 		.then(parseContent)
 		.then(applyEmojiToContent)
 		.then(parseMarkdownContent)
-		.then(data => {
+		.then((data) => {
 			data.fallback = fallback;
 			return data;
 		});
@@ -112,7 +112,7 @@ export function parseContent(text) {
 
 const markedWorker = !PRERENDER && new MarkedWorker();
 function parseMarkdownContent(data) {
-	return markedWorker.convert(data.content).then(html => {
+	return markedWorker.convert(data.content).then((html) => {
 		data.html = html;
 		return data;
 	});
@@ -121,7 +121,7 @@ function parseMarkdownContent(data) {
 let processEmojis, pendingEmojiProcessor;
 
 function applyEmojiToContent(data) {
-	return applyEmoji(data.content).then(content => {
+	return applyEmoji(data.content).then((content) => {
 		data.content = content;
 		return data;
 	});
@@ -144,7 +144,7 @@ function applyEmoji(content) {
 			/* webpackChunkName: "emoji" */ './gh-emoji'
 		).then(({ replace }) => (processEmojis = replace || String));
 	}
-	return pendingEmojiProcessor.then(processEmojis => processEmojis(content));
+	return pendingEmojiProcessor.then((processEmojis) => processEmojis(content));
 }
 
 /* generate ToC from markdown */
@@ -156,10 +156,7 @@ function generateToc(markdown) {
 		const level = token[1].length;
 		const text = sanitizeTitle(token[2]);
 		// Note: character range in regex is roughly "word characters including accented" (eg: bublé)
-		const id = text
-			.toLowerCase()
-			.replace(/[\s-]+/g, '-')
-			.replace(/[^a-z0-9\u00C0-\u024F-]/g, '');
+		const id = text.toLowerCase().replace(/[\s-]+/g, '-');
 		toc.push({ text, id, level });
 	}
 	return toc;