Do not raise a 500 exception when previewing empty media. (#8883)

3 files changed, 22 insertions, 12 deletions

diff --git a/changelog.d/8883.bugfix b/changelog.d/8883.bugfix
new file mode 100644
index 0000000000..6137fc5b2b
--- /dev/null
+++ b/changelog.d/8883.bugfix
@@ -0,0 +1 @@
+Fix a 500 error when attempting to preview an empty HTML file.
diff --git a/synapse/rest/media/v1/preview_url_resource.py b/synapse/rest/media/v1/preview_url_resource.py
index dce6c4d168..1082389d9b 100644
--- a/synapse/rest/media/v1/preview_url_resource.py
+++ b/synapse/rest/media/v1/preview_url_resource.py
@@ -676,7 +676,11 @@ class PreviewUrlResource(DirectServeJsonResource):
             logger.debug("No media removed from url cache")
 
 
-def decode_and_calc_og(body, media_uri, request_encoding=None):
+def decode_and_calc_og(body, media_uri, request_encoding=None) -> Dict[str, str]:
+    # If there's no body, nothing useful is going to be found.
+    if not body:
+        return {}
+
     from lxml import etree
 
     try:
diff --git a/tests/test_preview.py b/tests/test_preview.py
index 7f67ee9e1f..a883d707df 100644
--- a/tests/test_preview.py
+++ b/tests/test_preview.py
@@ -56,7 +56,7 @@ class PreviewTestCase(unittest.TestCase):
 
         desc = summarize_paragraphs(example_paras, min_size=200, max_size=500)
 
-        self.assertEquals(
+        self.assertEqual(
             desc,
             "Tromsø (Norwegian pronunciation: [ˈtrʊmsœ] ( listen); Northern Sami:"
             " Romsa; Finnish: Tromssa[2] Kven: Tromssa) is a city and municipality in"
@@ -69,7 +69,7 @@ class PreviewTestCase(unittest.TestCase):
 
         desc = summarize_paragraphs(example_paras[1:], min_size=200, max_size=500)
 
-        self.assertEquals(
+        self.assertEqual(
             desc,
             "Tromsø lies in Northern Norway. The municipality has a population of"
             " (2015) 72,066, but with an annual influx of students it has over 75,000"
@@ -96,7 +96,7 @@ class PreviewTestCase(unittest.TestCase):
 
         desc = summarize_paragraphs(example_paras, min_size=200, max_size=500)
 
-        self.assertEquals(
+        self.assertEqual(
             desc,
             "Tromsø (Norwegian pronunciation: [ˈtrʊmsœ] ( listen); Northern Sami:"
             " Romsa; Finnish: Tromssa[2] Kven: Tromssa) is a city and municipality in"
@@ -122,7 +122,7 @@ class PreviewTestCase(unittest.TestCase):
         ]
 
         desc = summarize_paragraphs(example_paras, min_size=200, max_size=500)
-        self.assertEquals(
+        self.assertEqual(
             desc,
             "Tromsø (Norwegian pronunciation: [ˈtrʊmsœ] ( listen); Northern Sami:"
             " Romsa; Finnish: Tromssa[2] Kven: Tromssa) is a city and municipality in"
@@ -149,7 +149,7 @@ class PreviewUrlTestCase(unittest.TestCase):
 
         og = decode_and_calc_og(html, "http://example.com/test.html")
 
-        self.assertEquals(og, {"og:title": "Foo", "og:description": "Some text."})
+        self.assertEqual(og, {"og:title": "Foo", "og:description": "Some text."})
 
     def test_comment(self):
         html = """
@@ -164,7 +164,7 @@ class PreviewUrlTestCase(unittest.TestCase):
 
         og = decode_and_calc_og(html, "http://example.com/test.html")
 
-        self.assertEquals(og, {"og:title": "Foo", "og:description": "Some text."})
+        self.assertEqual(og, {"og:title": "Foo", "og:description": "Some text."})
 
     def test_comment2(self):
         html = """
@@ -182,7 +182,7 @@ class PreviewUrlTestCase(unittest.TestCase):
 
         og = decode_and_calc_og(html, "http://example.com/test.html")
 
-        self.assertEquals(
+        self.assertEqual(
             og,
             {
                 "og:title": "Foo",
@@ -203,7 +203,7 @@ class PreviewUrlTestCase(unittest.TestCase):
 
         og = decode_and_calc_og(html, "http://example.com/test.html")
 
-        self.assertEquals(og, {"og:title": "Foo", "og:description": "Some text."})
+        self.assertEqual(og, {"og:title": "Foo", "og:description": "Some text."})
 
     def test_missing_title(self):
         html = """
@@ -216,7 +216,7 @@ class PreviewUrlTestCase(unittest.TestCase):
 
         og = decode_and_calc_og(html, "http://example.com/test.html")
 
-        self.assertEquals(og, {"og:title": None, "og:description": "Some text."})
+        self.assertEqual(og, {"og:title": None, "og:description": "Some text."})
 
     def test_h1_as_title(self):
         html = """
@@ -230,7 +230,7 @@ class PreviewUrlTestCase(unittest.TestCase):
 
         og = decode_and_calc_og(html, "http://example.com/test.html")
 
-        self.assertEquals(og, {"og:title": "Title", "og:description": "Some text."})
+        self.assertEqual(og, {"og:title": "Title", "og:description": "Some text."})
 
     def test_missing_title_and_broken_h1(self):
         html = """
@@ -244,4 +244,9 @@ class PreviewUrlTestCase(unittest.TestCase):
 
         og = decode_and_calc_og(html, "http://example.com/test.html")
 
-        self.assertEquals(og, {"og:title": None, "og:description": "Some text."})
+        self.assertEqual(og, {"og:title": None, "og:description": "Some text."})
+
+    def test_empty(self):
+        html = ""
+        og = decode_and_calc_og(html, "http://example.com/test.html")
+        self.assertEqual(og, {})