]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/tumblr.py
Merge tag 'upstream/2016.06.25'
[youtubedl] / youtube_dl / extractor / tumblr.py
index 2882c1809e0bd55c1e6c8b441c19293aeb64d301..4d8b57111897f3c936e11f55fcea60d6a6bd30d6 100644 (file)
@@ -4,13 +4,11 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..utils import (
-    ExtractorError,
-)
+from ..utils import int_or_none
 
 
 class TumblrIE(InfoExtractor):
-    _VALID_URL = r'http://(?P<blog_name>.*?)\.tumblr\.com/((post)|(video))/(?P<id>\d*)($|/)'
+    _VALID_URL = r'https?://(?P<blog_name>[^/?#&]+)\.tumblr\.com/(?:post|video)/(?P<id>[0-9]+)(?:$|[/?#])'
     _TESTS = [{
         'url': 'http://tatianamaslanydaily.tumblr.com/post/54196191430/orphan-black-dvd-extra-behind-the-scenes',
         'md5': '479bb068e5b16462f5176a6828829767',
@@ -18,7 +16,7 @@ class TumblrIE(InfoExtractor):
             'id': '54196191430',
             'ext': 'mp4',
             'title': 'tatiana maslany news, Orphan Black || DVD extra - behind the scenes ↳...',
-            'description': 'md5:dfac39636969fe6bf1caa2d50405f069',
+            'description': 'md5:37db8211e40b50c7c44e95da14f630b7',
             'thumbnail': 're:http://.*\.jpg',
         }
     }, {
@@ -27,10 +25,76 @@ class TumblrIE(InfoExtractor):
         'info_dict': {
             'id': '90208453769',
             'ext': 'mp4',
-            'title': '5SOS STRUM ;)',
+            'title': '5SOS STRUM ;]',
             'description': 'md5:dba62ac8639482759c8eb10ce474586a',
             'thumbnail': 're:http://.*\.jpg',
         }
+    }, {
+        'url': 'http://hdvideotest.tumblr.com/post/130323439814/test-description-for-my-hd-video',
+        'md5': '7ae503065ad150122dc3089f8cf1546c',
+        'info_dict': {
+            'id': '130323439814',
+            'ext': 'mp4',
+            'title': 'HD Video Testing \u2014 Test description for my HD video',
+            'description': 'md5:97cc3ab5fcd27ee4af6356701541319c',
+            'thumbnail': 're:http://.*\.jpg',
+        },
+        'params': {
+            'format': 'hd',
+        },
+    }, {
+        'url': 'http://naked-yogi.tumblr.com/post/118312946248/naked-smoking-stretching',
+        'md5': 'de07e5211d60d4f3a2c3df757ea9f6ab',
+        'info_dict': {
+            'id': 'Wmur',
+            'ext': 'mp4',
+            'title': 'naked smoking & stretching',
+            'upload_date': '20150506',
+            'timestamp': 1430931613,
+            'age_limit': 18,
+            'uploader_id': '1638622',
+            'uploader': 'naked-yogi',
+        },
+        'add_ie': ['Vidme'],
+    }, {
+        'url': 'http://camdamage.tumblr.com/post/98846056295/',
+        'md5': 'a9e0c8371ea1ca306d6554e3fecf50b6',
+        'info_dict': {
+            'id': '105463834',
+            'ext': 'mp4',
+            'title': 'Cam Damage-HD 720p',
+            'uploader': 'John Moyer',
+            'uploader_id': 'user32021558',
+        },
+        'add_ie': ['Vimeo'],
+    }, {
+        'url': 'http://sutiblr.tumblr.com/post/139638707273',
+        'md5': '2dd184b3669e049ba40563a7d423f95c',
+        'info_dict': {
+            'id': 'ir7qBEIKqvq',
+            'ext': 'mp4',
+            'title': 'Vine by sutiblr',
+            'alt_title': 'Vine by sutiblr',
+            'uploader': 'sutiblr',
+            'uploader_id': '1198993975374495744',
+            'upload_date': '20160220',
+            'like_count': int,
+            'comment_count': int,
+            'repost_count': int,
+        },
+        'add_ie': ['Vine'],
+    }, {
+        'url': 'http://vitasidorkina.tumblr.com/post/134652425014/joskriver-victoriassecret-invisibility-or',
+        'md5': '01c12ceb82cbf6b2fe0703aa56b3ad72',
+        'info_dict': {
+            'id': '-7LnUPGlSo',
+            'ext': 'mp4',
+            'title': 'Video by victoriassecret',
+            'description': 'Invisibility or flight…which superpower would YOU choose? #VSFashionShow #ThisOrThat',
+            'uploader_id': 'victoriassecret',
+            'thumbnail': 're:^https?://.*\.jpg'
+        },
+        'add_ie': ['Instagram'],
     }]
 
     def _real_extract(self, url):
@@ -39,30 +103,58 @@ class TumblrIE(InfoExtractor):
         blog = m_url.group('blog_name')
 
         url = 'http://%s.tumblr.com/post/%s/' % (blog, video_id)
-        webpage = self._download_webpage(url, video_id)
+        webpage, urlh = self._download_webpage_handle(url, video_id)
+
+        iframe_url = self._search_regex(
+            r'src=\'(https?://www\.tumblr\.com/video/[^\']+)\'',
+            webpage, 'iframe url', default=None)
+        if iframe_url is None:
+            return self.url_result(urlh.geturl(), 'Generic')
+
+        iframe = self._download_webpage(iframe_url, video_id, 'Downloading iframe page')
 
-        re_video = r'src=\\x22(?P<video_url>http://%s\.tumblr\.com/video_file/%s/(.*?))\\x22 type=\\x22video/(?P<ext>.*?)\\x22' % (blog, video_id)
-        video = re.search(re_video, webpage)
-        if video is None:
-            raise ExtractorError('Unable to extract video')
-        video_url = video.group('video_url')
-        ext = video.group('ext')
+        duration = None
+        sources = []
 
-        video_thumbnail = self._search_regex(
-            r'posters.*?\[\\x22(.*?)\\x22',
-            webpage, 'thumbnail', fatal=False)  # We pick the first poster
-        if video_thumbnail:
-            video_thumbnail = video_thumbnail.replace('\\\\/', '/')
+        sd_url = self._search_regex(
+            r'<source[^>]+src=(["\'])(?P<url>.+?)\1', iframe,
+            'sd video url', default=None, group='url')
+        if sd_url:
+            sources.append((sd_url, 'sd'))
+
+        options = self._parse_json(
+            self._search_regex(
+                r'data-crt-options=(["\'])(?P<options>.+?)\1', iframe,
+                'hd video url', default='', group='options'),
+            video_id, fatal=False)
+        if options:
+            duration = int_or_none(options.get('duration'))
+            hd_url = options.get('hdUrl')
+            if hd_url:
+                sources.append((hd_url, 'hd'))
+
+        formats = [{
+            'url': video_url,
+            'ext': 'mp4',
+            'format_id': format_id,
+            'height': int_or_none(self._search_regex(
+                r'/(\d{3,4})$', video_url, 'height', default=None)),
+            'quality': quality,
+        } for quality, (video_url, format_id) in enumerate(sources)]
+
+        self._sort_formats(formats)
 
         # The only place where you can get a title, it's not complete,
         # but searching in other places doesn't work for all videos
-        video_title = self._html_search_regex(r'<title>(?P<title>.*?)(?: \| Tumblr)?</title>',
-            webpage, 'title', flags=re.DOTALL)
+        video_title = self._html_search_regex(
+            r'(?s)<title>(?P<title>.*?)(?: \| Tumblr)?</title>',
+            webpage, 'title')
 
-        return [{'id': video_id,
-                 'url': video_url,
-                 'title': video_title,
-                 'description': self._html_search_meta('description', webpage),
-                 'thumbnail': video_thumbnail,
-                 'ext': ext
-                 }]
+        return {
+            'id': video_id,
+            'title': video_title,
+            'description': self._og_search_description(webpage, default=None),
+            'thumbnail': self._og_search_thumbnail(webpage, default=None),
+            'duration': duration,
+            'formats': formats,
+        }