def test_charset_in_header(self) -> None:
html = """<html>
<head>
<meta property="og:title" content="中文" />
</head>
</html>""".encode("big5")
parser = OpenGraphParser(html, "text/html; charset=Big5")
result = parser.extract_data()
self.assertEqual(result["title"], "中文")
def test_page_with_og(self) -> None:
html = b"""<html>
<head>
<meta property="og:title" content="The Rock" />
<meta property="og:type" content="video.movie" />
<meta property="og:url" content="http://www.imdb.com/title/tt0117500/" />
<meta property="og:image" content="http://ia.media-imdb.com/images/rock.jpg" />
<meta property="og:description" content="The Rock film" />
</head>
</html>"""
parser = OpenGraphParser(html, "text/html; charset=UTF-8")
result = parser.extract_data()
self.assertEqual(result.title, "The Rock")
self.assertEqual(result.description, "The Rock film")
def test_page_with_og(self) -> None:
html = """<html>
<head>
<meta property="og:title" content="The Rock" />
<meta property="og:type" content="video.movie" />
<meta property="og:url" content="http://www.imdb.com/title/tt0117500/" />
<meta property="og:image" content="http://ia.media-imdb.com/images/rock.jpg" />
<meta property="og:description" content="The Rock film" />
</head>
</html>"""
parser = OpenGraphParser(html)
result = parser.extract_data()
self.assertIn('title', result)
self.assertEqual(result['title'], 'The Rock')
self.assertEqual(result.get('description'), 'The Rock film')
def test_page_with_og(self) -> None:
html = """<html>
<head>
<meta property="og:title" content="The Rock" />
<meta property="og:type" content="video.movie" />
<meta property="og:url" content="http://www.imdb.com/title/tt0117500/" />
<meta property="og:image" content="http://ia.media-imdb.com/images/rock.jpg" />
<meta property="og:description" content="The Rock film" />
</head>
</html>"""
parser = OpenGraphParser(html)
result = parser.extract_data()
self.assertIn('title', result)
self.assertEqual(result['title'], 'The Rock')
self.assertEqual(result.get('description'), 'The Rock film')
def test_page_with_evil_og_tags(self) -> None:
html = b"""<html>
<head>
<meta property="og:title" content="The Rock" />
<meta property="og:type" content="video.movie" />
<meta property="og:url" content="http://www.imdb.com/title/tt0117500/" />
<meta property="og:image" content="http://ia.media-imdb.com/images/rock.jpg" />
<meta property="og:description" content="The Rock film" />
<meta property="og:html" content="<script>alert(window.location)</script>" />
<meta property="og:oembed" content="True" />
</head>
</html>"""
parser = OpenGraphParser(html, "text/html; charset=UTF-8")
result = parser.extract_data()
self.assertIn("title", result)
self.assertEqual(result["title"], "The Rock")
self.assertEqual(result.get("description"), "The Rock film")
self.assertEqual(result.get("oembed"), None)
self.assertEqual(result.get("html"), None)
