From 01df5bacac3aa0e8356fed889ea0b69c4c044535 Mon Sep 17 00:00:00 2001 From: Patrick Cloke Date: Fri, 3 Jun 2022 12:09:12 -0400 Subject: Improve URL previews for some pages (#12951) * Skip `og` and `meta` tags where the value is empty. * Fallback to the favicon if there are no other images. * Ignore tags meant for navigation. --- tests/rest/media/v1/test_html_preview.py | 37 +++++++++++++++++++++++++++++++- 1 file changed, 36 insertions(+), 1 deletion(-) (limited to 'tests/rest') diff --git a/tests/rest/media/v1/test_html_preview.py b/tests/rest/media/v1/test_html_preview.py index 62e308814d..ea9e5889bf 100644 --- a/tests/rest/media/v1/test_html_preview.py +++ b/tests/rest/media/v1/test_html_preview.py @@ -145,7 +145,7 @@ class SummarizeTestCase(unittest.TestCase): ) -class CalcOgTestCase(unittest.TestCase): +class OpenGraphFromHtmlTestCase(unittest.TestCase): if not lxml: skip = "url preview feature requires lxml" @@ -235,6 +235,21 @@ class CalcOgTestCase(unittest.TestCase): self.assertEqual(og, {"og:title": None, "og:description": "Some text."}) + # Another variant is a title with no content. + html = b""" + + + +

Title

+ + + """ + + tree = decode_body(html, "http://example.com/test.html") + og = parse_html_to_open_graph(tree) + + self.assertEqual(og, {"og:title": "Title", "og:description": "Title"}) + def test_h1_as_title(self) -> None: html = b""" @@ -250,6 +265,26 @@ class CalcOgTestCase(unittest.TestCase): self.assertEqual(og, {"og:title": "Title", "og:description": "Some text."}) + def test_empty_description(self) -> None: + """Description tags with empty content should be ignored.""" + html = b""" + + + + + + + +