summary refs log tree commit diff
diff options
context:
space:
mode:
authorBarbara Miller <galgeek@me.com>2019-07-22 14:09:21 -0700
committerSergey M․ <dstftw@gmail.com>2019-09-01 01:25:48 +0700
commitd78657fd18ae6413239137298eee4c54f3efee32 (patch)
tree6104848dc8593c551397f60d3713c9959f58603c
parentcc73d5ad15aed96f6462b8079ccb6716c2ef9f85 (diff)
downloadyoutube-dl-d78657fd18ae6413239137298eee4c54f3efee32.tar.gz
youtube-dl-d78657fd18ae6413239137298eee4c54f3efee32.tar.xz
youtube-dl-d78657fd18ae6413239137298eee4c54f3efee32.zip
[extractor/generic] Add support for squarespace embeds (closes #21294)
-rw-r--r--youtube_dl/extractor/generic.py18
1 files changed, 18 insertions, 0 deletions
diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py
index d34fc4b15..7dd2e2d5f 100644
--- a/youtube_dl/extractor/generic.py
+++ b/youtube_dl/extractor/generic.py
@@ -2076,6 +2076,17 @@ class GenericIE(InfoExtractor):
             'playlist_count': 6,
         },
         {
+            # Squarespace video embed, 2019-08-28
+            'url': 'http://ootboxford.com',
+            'info_dict': {
+                'id': 'Tc7b_JGdZfw',
+                'title': 'Out of the Blue, at Childish Things 10',
+            },
+            'params': {
+                'skip_download': True,
+            },
+        },
+        {
             # Zype embed
             'url': 'https://www.cookscountry.com/episode/554-smoky-barbecue-favorites',
             'info_dict': {
@@ -2395,6 +2406,13 @@ class GenericIE(InfoExtractor):
         # Unescaping the whole page allows to handle those cases in a generic way
         webpage = compat_urllib_parse_unquote(webpage)
 
+        # unescape re.sub replacement
+        def unescape_resub(m):
+            return unescapeHTML(m.group(0))
+
+        # unescape squarespace video embeds
+        webpage = re.sub(r'<div[^>]+class=[^>]*?sqs-video-wrapper[^>]*>', unescape_resub, webpage)
+
         # it's tempting to parse this further, but you would
         # have to take into account all the variations like
         #   Video Title - Site Name