Improve XML handling

.travis.yml

@@ -3,5 +3,7 @@ os:
   - osx
 language: node_js
 node_js:
+  - '12'
+  - '10'
+  - '8'
   - '6'
-  - '4'

fixture/youtube-with-complex-query.webloc

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
+<plist version="1.0">
+<dict>
+	<key>URL</key>
+	<string>https://www.youtube.com/watch?v=XXXXXXXXXXX&amp;complex=true</string>
+</dict>
+</plist>

fixture/youtube-with-query.webloc

@@ -0,0 +1,8 @@
+<?xml version="1.0" encoding="UTF-8"?>
+<!DOCTYPE plist PUBLIC "-//Apple//DTD PLIST 1.0//EN" "http://www.apple.com/DTDs/PropertyList-1.0.dtd">
+<plist version="1.0">
+<dict>
+	<key>URL</key>
+	<string>https://www.youtube.com/watch?v=XXXXXXXXXXX</string>
+</dict>
+</plist>

index.js

@@ -15,16 +15,39 @@ const getExt = () => {
 	}
 };
 
+const xmlTagRegexp = /<.*?>/g;
+const xmlEscapeRegexp = /&(#\d+|lt|gt|quot|apos|amp);/g;
+const xmlUnescape = str => str.replace(xmlEscapeRegexp, (_, escape) => {
+	switch (escape) {
+		case 'lt': return '<';
+		case 'gt': return '>';
+		case 'quot': return '"';
+		case 'apos': return '\'';
+		case 'amp': return '&';
+		default: return String.fromCharCode(parseInt(escape.substr(1), 10));
+	}
+});
+
 module.exports = filepath => {
 	filepath += path.extname(filepath) ? '' : getExt();
 
 	return pify(fs.readFile)(filepath, 'utf8')
-		.then(data => getUrls(data.replace(/<!doctype.*/i, ''))[0].trim())
-		.catch(err => {
-			if (err.code === 'ENOENT') {
-				err.message = `Couldn't find a web shortcut with the name \`${path.basename(`${filepath}\``)}`;
+		.then(text => {
+			let data = text.trim();
+
+			const isXml = data[0] === '<';
+			if (isXml) {
+				data = data.replace(xmlTagRegexp, ' ');
+				data = xmlUnescape(data);
+			}
+
+			return getUrls(data)[0].trim();
+		})
+		.catch(error => {
+			if (error.code === 'ENOENT') {
+				error.message = `Couldn't find a web shortcut with the name \`${path.basename(`${filepath}\``)}`;
 			}
 
-			throw err;
+			throw error;
 		});
 };

test.js

@@ -1,5 +1,5 @@
 import test from 'ava';
-import m from './';
+import m from '.';
 
 test('main', async t => {
 	t.is(await m('fixture/google'), 'https://google.com');
@@ -9,6 +9,14 @@ test('with extension', async t => {
 	t.is(await m('fixture/google.webloc'), 'https://google.com');
 });
 
-test('unknown shortcut', t => {
-	t.throws(m('fixture/unknown.webloc'), 'Couldn\'t find a web shortcut with the name `unknown.webloc`');
+test('webloc with query, #2', async t => {
+	t.is(await m('fixture/youtube-with-query.webloc'), 'https://youtube.com/watch?v=XXXXXXXXXXX');
+});
+
+test('webloc complex query, #2', async t => {
+	t.is(await m('fixture/youtube-with-complex-query.webloc'), 'https://youtube.com/watch?complex=true&v=XXXXXXXXXXX');
+});
+
+test('unknown shortcut', async t => {
+	await t.throwsAsync(() => m('fixture/unknown.webloc'), 'Couldn\'t find a web shortcut with the name `unknown.webloc`');
 });