summary refs log tree commit diff
path: root/tests
diff options
context:
space:
mode:
authorPatrick Cloke <clokep@users.noreply.github.com>2021-12-13 12:55:07 -0500
committerGitHub <noreply@github.com>2021-12-13 17:55:07 +0000
commiteb39da6782b57c939450839097f32a14cba3ebfc (patch)
tree8dbd3921d86cb87b7d97ebe7f0b61e374b5868ca /tests
parentType hint the constructors of the data store classes (#11555) (diff)
downloadsynapse-eb39da6782b57c939450839097f32a14cba3ebfc.tar.xz
Move HTML parsing to a separate file for URL previews. (#11566)
* Splits the logic for parsing HTML from the resource handling code.
* Fix a circular import in the oEmbed code (which uses the HTML parsing code).
* Renames some of the HTML parsing methods to:
  * Make it clear which methods are "internal" to the module.
  * Clarify what the methods do.
Diffstat (limited to 'tests')
-rw-r--r--tests/rest/media/v1/test_url_preview.py1
-rw-r--r--tests/test_preview.py46
2 files changed, 24 insertions, 23 deletions
diff --git a/tests/rest/media/v1/test_url_preview.py b/tests/rest/media/v1/test_url_preview.py
index 8698135a76..16e904f15b 100644
--- a/tests/rest/media/v1/test_url_preview.py
+++ b/tests/rest/media/v1/test_url_preview.py
@@ -1,4 +1,5 @@
 # Copyright 2018 New Vector Ltd
+# Copyright 2021 The Matrix.org Foundation C.I.C.
 #
 # Licensed under the Apache License, Version 2.0 (the "License");
 # you may not use this file except in compliance with the License.
diff --git a/tests/test_preview.py b/tests/test_preview.py
index 40b89fb2ef..46e02f483f 100644
--- a/tests/test_preview.py
+++ b/tests/test_preview.py
@@ -12,10 +12,10 @@
 # See the License for the specific language governing permissions and
 # limitations under the License.
 
-from synapse.rest.media.v1.preview_url_resource import (
-    _calc_og,
+from synapse.rest.media.v1.preview_html import (
+    _get_html_media_encodings,
     decode_body,
-    get_html_media_encodings,
+    parse_html_to_open_graph,
     summarize_paragraphs,
 )
 
@@ -160,7 +160,7 @@ class CalcOgTestCase(unittest.TestCase):
         """
 
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
 
         self.assertEqual(og, {"og:title": "Foo", "og:description": "Some text."})
 
@@ -176,7 +176,7 @@ class CalcOgTestCase(unittest.TestCase):
         """
 
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
 
         self.assertEqual(og, {"og:title": "Foo", "og:description": "Some text."})
 
@@ -195,7 +195,7 @@ class CalcOgTestCase(unittest.TestCase):
         """
 
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
 
         self.assertEqual(
             og,
@@ -217,7 +217,7 @@ class CalcOgTestCase(unittest.TestCase):
         """
 
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
 
         self.assertEqual(og, {"og:title": "Foo", "og:description": "Some text."})
 
@@ -231,7 +231,7 @@ class CalcOgTestCase(unittest.TestCase):
         """
 
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
 
         self.assertEqual(og, {"og:title": None, "og:description": "Some text."})
 
@@ -246,7 +246,7 @@ class CalcOgTestCase(unittest.TestCase):
         """
 
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
 
         self.assertEqual(og, {"og:title": "Title", "og:description": "Some text."})
 
@@ -261,7 +261,7 @@ class CalcOgTestCase(unittest.TestCase):
         """
 
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
 
         self.assertEqual(og, {"og:title": None, "og:description": "Some text."})
 
@@ -289,7 +289,7 @@ class CalcOgTestCase(unittest.TestCase):
         <head><title>Foo</title></head><body>Some text.</body></html>
         """.strip()
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
         self.assertEqual(og, {"og:title": "Foo", "og:description": "Some text."})
 
     def test_invalid_encoding(self):
@@ -303,7 +303,7 @@ class CalcOgTestCase(unittest.TestCase):
         </html>
         """
         tree = decode_body(html, "http://example.com/test.html", "invalid-encoding")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
         self.assertEqual(og, {"og:title": "Foo", "og:description": "Some text."})
 
     def test_invalid_encoding2(self):
@@ -318,7 +318,7 @@ class CalcOgTestCase(unittest.TestCase):
         </html>
         """
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
         self.assertEqual(og, {"og:title": "ÿÿ Foo", "og:description": "Some text."})
 
     def test_windows_1252(self):
@@ -332,14 +332,14 @@ class CalcOgTestCase(unittest.TestCase):
         </html>
         """
         tree = decode_body(html, "http://example.com/test.html")
-        og = _calc_og(tree, "http://example.com/test.html")
+        og = parse_html_to_open_graph(tree, "http://example.com/test.html")
         self.assertEqual(og, {"og:title": "ó", "og:description": "Some text."})
 
 
 class MediaEncodingTestCase(unittest.TestCase):
     def test_meta_charset(self):
         """A character encoding is found via the meta tag."""
-        encodings = get_html_media_encodings(
+        encodings = _get_html_media_encodings(
             b"""
         <html>
         <head><meta charset="ascii">
@@ -351,7 +351,7 @@ class MediaEncodingTestCase(unittest.TestCase):
         self.assertEqual(list(encodings), ["ascii", "utf-8", "cp1252"])
 
         # A less well-formed version.
-        encodings = get_html_media_encodings(
+        encodings = _get_html_media_encodings(
             b"""
         <html>
         <head>< meta charset = ascii>
@@ -364,7 +364,7 @@ class MediaEncodingTestCase(unittest.TestCase):
 
     def test_meta_charset_underscores(self):
         """A character encoding contains underscore."""
-        encodings = get_html_media_encodings(
+        encodings = _get_html_media_encodings(
             b"""
         <html>
         <head><meta charset="Shift_JIS">
@@ -377,7 +377,7 @@ class MediaEncodingTestCase(unittest.TestCase):
 
     def test_xml_encoding(self):
         """A character encoding is found via the meta tag."""
-        encodings = get_html_media_encodings(
+        encodings = _get_html_media_encodings(
             b"""
         <?xml version="1.0" encoding="ascii"?>
         <html>
@@ -389,7 +389,7 @@ class MediaEncodingTestCase(unittest.TestCase):
 
     def test_meta_xml_encoding(self):
         """Meta tags take precedence over XML encoding."""
-        encodings = get_html_media_encodings(
+        encodings = _get_html_media_encodings(
             b"""
         <?xml version="1.0" encoding="ascii"?>
         <html>
@@ -413,17 +413,17 @@ class MediaEncodingTestCase(unittest.TestCase):
             'text/html; charset=ascii";',
         )
         for header in headers:
-            encodings = get_html_media_encodings(b"", header)
+            encodings = _get_html_media_encodings(b"", header)
             self.assertEqual(list(encodings), ["ascii", "utf-8", "cp1252"])
 
     def test_fallback(self):
         """A character encoding cannot be found in the body or header."""
-        encodings = get_html_media_encodings(b"", "text/html")
+        encodings = _get_html_media_encodings(b"", "text/html")
         self.assertEqual(list(encodings), ["utf-8", "cp1252"])
 
     def test_duplicates(self):
         """Ensure each encoding is only attempted once."""
-        encodings = get_html_media_encodings(
+        encodings = _get_html_media_encodings(
             b"""
         <?xml version="1.0" encoding="utf8"?>
         <html>
@@ -437,7 +437,7 @@ class MediaEncodingTestCase(unittest.TestCase):
 
     def test_unknown_invalid(self):
         """A character encoding should be ignored if it is unknown or invalid."""
-        encodings = get_html_media_encodings(
+        encodings = _get_html_media_encodings(
             b"""
         <html>
         <head><meta charset="invalid">
generated by cgit-pink 1.4.1 (git 2.36.1) at 2024-11-01 16:30:50 +0000