[twitter] Fix several failed tests

[youtube-dl] / youtube_dl / extractor / twitter.py

diff --git a/youtube_dl/extractor/twitter.py b/youtube_dl/extractor/twitter.py
index a161f046b2532805d864a26e083de06f68cf7a1f..64c429f02c61b1a1fea49c7900e93757740bc28d 100644 (file)
--- a/youtube_dl/extractor/twitter.py
+++ b/youtube_dl/extractor/twitter.py
@@ -14,13 +14,19 @@ from ..utils import (
 )
 
 
-class TwitterCardIE(InfoExtractor):
+class TwitterBaseIE(InfoExtractor):
+    def _get_vmap_video_url(self, vmap_url, video_id):
+        vmap_data = self._download_xml(vmap_url, video_id)
+        return xpath_text(vmap_data, './/MediaFile').strip()
+
+
+class TwitterCardIE(TwitterBaseIE):
     IE_NAME = 'twitter:card'
     _VALID_URL = r'https?://(?:www\.)?twitter\.com/i/cards/tfw/v1/(?P<id>\d+)'
     _TESTS = [
         {
             'url': 'https://twitter.com/i/cards/tfw/v1/560070183650213889',
-            'md5': '4fa26a35f9d1bf4b646590ba8e84be19',
+            # MD5 checksums are different in different places
             'info_dict': {
                 'id': '560070183650213889',
                 'ext': 'mp4',
@@ -42,7 +48,7 @@ class TwitterCardIE(InfoExtractor):
         },
         {
             'url': 'https://twitter.com/i/cards/tfw/v1/654001591733886977',
-            'md5': 'b6f35e8b08a0bec6c8af77a2f4b3a814',
+            'md5': 'd4724ffe6d2437886d004fa5de1043b3',
             'info_dict': {
                 'id': 'dq4Oj5quskI',
                 'ext': 'mp4',
@@ -62,8 +68,8 @@ class TwitterCardIE(InfoExtractor):
                 'ext': 'mp4',
                 'upload_date': '20151113',
                 'uploader_id': '1189339351084113920',
-                'uploader': '@ArsenalTerje',
-                'title': 'Vine by @ArsenalTerje',
+                'uploader': 'ArsenalTerje',
+                'title': 'Vine by ArsenalTerje',
             },
             'add_ie': ['Vine'],
         }
@@ -96,10 +102,8 @@ class TwitterCardIE(InfoExtractor):
                 video_id)
             if 'playlist' not in config:
                 if 'vmapUrl' in config:
-                    vmap_data = self._download_xml(config['vmapUrl'], video_id)
-                    video_url = xpath_text(vmap_data, './/MediaFile').strip()
                     formats.append({
-                        'url': video_url,
+                        'url': self._get_vmap_video_url(config['vmapUrl'], video_id),
                     })
                     break   # same video regardless of UA
                 continue
@@ -138,7 +142,7 @@ class TwitterIE(InfoExtractor):
 
     _TESTS = [{
         'url': 'https://twitter.com/freethenipple/status/643211948184596480',
-        'md5': 'db6612ec5d03355953c3ca9250c97e5e',
+        # MD5 checksums are different in different places
         'info_dict': {
             'id': '643211948184596480',
             'ext': 'mp4',
@@ -226,3 +230,32 @@ class TwitterIE(InfoExtractor):
             return info
 
         raise ExtractorError('There\'s not video in this tweet.')
+
+
+class TwitterAmplifyIE(TwitterBaseIE):
+    IE_NAME = 'twitter:amplify'
+    _VALID_URL = 'https?://amp\.twimg\.com/v/(?P<id>[0-9a-f\-]{36})'
+
+    _TEST = {
+        'url': 'https://amp.twimg.com/v/0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951',
+        'md5': '7df102d0b9fd7066b86f3159f8e81bf6',
+        'info_dict': {
+            'id': '0ba0c3c7-0af3-4c0a-bed5-7efd1ffa2951',
+            'ext': 'mp4',
+            'title': 'Twitter Video',
+        },
+    }
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        vmap_url = self._html_search_meta(
+            'twitter:amplify:vmap', webpage, 'vmap url')
+        video_url = self._get_vmap_video_url(vmap_url, video_id)
+
+        return {
+            'id': video_id,
+            'title': 'Twitter Video',
+            'url': video_url,
+        }