]> git.bitcoin.ninja Git - youtube-dl/commitdiff
[instagram] Extract metadata from JSON
authorSergey M․ <dstftw@gmail.com>
Sat, 11 Jun 2016 23:06:04 +0000 (06:06 +0700)
committerSergey M․ <dstftw@gmail.com>
Sat, 11 Jun 2016 23:06:04 +0000 (06:06 +0700)
youtube_dl/extractor/instagram.py

index 3cbe77ad80f2fc9a03c738745524d5dac98c9d37..fc0197ae19d2c3db4a504268256400d6ef81702d 100644 (file)
@@ -8,6 +8,7 @@ from ..utils import (
     int_or_none,
     limit_length,
     lowercase_escape,
+    try_get,
 )
 
 
@@ -19,10 +20,16 @@ class InstagramIE(InfoExtractor):
         'info_dict': {
             'id': 'aye83DjauH',
             'ext': 'mp4',
-            'uploader_id': 'naomipq',
             'title': 'Video by naomipq',
             'description': 'md5:1f17f0ab29bd6fe2bfad705f58de3cb8',
-        }
+            'thumbnail': 're:^https?://.*\.jpg',
+            'timestamp': 1371748545,
+            'upload_date': '20130620',
+            'uploader_id': 'naomipq',
+            'uploader': 'Naomi Leonor Phan-Quang',
+            'like_count': int,
+            'comment_count': int,
+        },
     }, {
         # missing description
         'url': 'https://www.instagram.com/p/BA-pQFBG8HZ/?taken-by=britneyspears',
@@ -31,6 +38,13 @@ class InstagramIE(InfoExtractor):
             'ext': 'mp4',
             'uploader_id': 'britneyspears',
             'title': 'Video by britneyspears',
+            'thumbnail': 're:^https?://.*\.jpg',
+            'timestamp': 1453760977,
+            'upload_date': '20160125',
+            'uploader_id': 'britneyspears',
+            'uploader': 'Britney Spears',
+            'like_count': int,
+            'comment_count': int,
         },
         'params': {
             'skip_download': True,
@@ -67,21 +81,57 @@ class InstagramIE(InfoExtractor):
         url = mobj.group('url')
 
         webpage = self._download_webpage(url, video_id)
-        uploader_id = self._search_regex(r'"owner":{"username":"(.+?)"',
-                                         webpage, 'uploader id', fatal=False)
-        desc = self._search_regex(
-            r'"caption":"(.+?)"', webpage, 'description', default=None)
-        if desc is not None:
-            desc = lowercase_escape(desc)
+
+        (video_url, description, thumbnail, timestamp, uploader,
+         uploader_id, like_count, comment_count) = [None] * 8
+
+        shared_data = self._parse_json(
+            self._search_regex(
+                r'window\._sharedData\s*=\s*({.+?});',
+                webpage, 'shared data', default='{}'),
+            video_id, fatal=False)
+        if shared_data:
+            media = try_get(
+                shared_data, lambda x: x['entry_data']['PostPage'][0]['media'], dict)
+            if media:
+                video_url = media.get('video_url')
+                description = media.get('caption')
+                thumbnail = media.get('display_src')
+                timestamp = int_or_none(media.get('date'))
+                uploader = media.get('owner', {}).get('full_name')
+                uploader_id = media.get('owner', {}).get('username')
+                like_count = int_or_none(media.get('likes', {}).get('count'))
+                comment_count = int_or_none(media.get('comments', {}).get('count'))
+
+        if not video_url:
+            video_url = self._og_search_video_url(webpage, secure=False)
+
+        if not uploader_id:
+            uploader_id = self._search_regex(
+                r'"owner"\s*:\s*{\s*"username"\s*:\s*"(.+?)"',
+                webpage, 'uploader id', fatal=False)
+
+        if not description:
+            description = self._search_regex(
+                r'"caption"\s*:\s*"(.+?)"', webpage, 'description', default=None)
+            if description is not None:
+                description = lowercase_escape(description)
+
+        if not thumbnail:
+            thumbnail = self._og_search_thumbnail(webpage)
 
         return {
             'id': video_id,
-            'url': self._og_search_video_url(webpage, secure=False),
+            'url': video_url,
             'ext': 'mp4',
             'title': 'Video by %s' % uploader_id,
-            'thumbnail': self._og_search_thumbnail(webpage),
+            'description': description,
+            'thumbnail': thumbnail,
+            'timestamp': timestamp,
             'uploader_id': uploader_id,
-            'description': desc,
+            'uploader': uploader,
+            'like_count': like_count,
+            'comment_count': comment_count,
         }