]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/vk.py
Prepare to upload
[youtubedl] / youtube_dl / extractor / vk.py
index 3ee66e23e22b1d5350148d5cb21c2f55e4885f75..8b6dc0e244982f46cbc0d65bf5133b3505779fb0 100644 (file)
@@ -1,37 +1,36 @@
-# encoding: utf-8
+# coding: utf-8
 from __future__ import unicode_literals
 
+import collections
 import re
-import json
-import sys
 
 from .common import InfoExtractor
-from ..compat import (
-    compat_str,
-    compat_urlparse,
-)
+from ..compat import compat_urlparse
 from ..utils import (
     clean_html,
     ExtractorError,
     get_element_by_class,
     int_or_none,
     orderedSet,
-    parse_duration,
     remove_start,
+    str_or_none,
     str_to_int,
     unescapeHTML,
-    unified_strdate,
+    unified_timestamp,
+    url_or_none,
     urlencode_postdata,
 )
-from .vimeo import VimeoIE
+from .dailymotion import DailymotionIE
 from .pladform import PladformIE
+from .vimeo import VimeoIE
+from .youtube import YoutubeIE
 
 
 class VKBaseIE(InfoExtractor):
     _NETRC_MACHINE = 'vk'
 
     def _login(self):
-        (username, password) = self._get_login_info()
+        username, password = self._get_login_info()
         if username is None:
             return
 
@@ -45,26 +44,13 @@ class VKBaseIE(InfoExtractor):
             'pass': password.encode('cp1251'),
         })
 
-        # https://new.vk.com/ serves two same remixlhk cookies in Set-Cookie header
-        # and expects the first one to be set rather than second (see
-        # https://github.com/rg3/youtube-dl/issues/9841#issuecomment-227871201).
-        # As of RFC6265 the newer one cookie should be set into cookie store
-        # what actually happens.
-        # We will workaround this VK issue by resetting the remixlhk cookie to
-        # the first one manually.
-        cookies = url_handle.headers.get('Set-Cookie')
-        if cookies:
-            if sys.version_info[0] >= 3:
-                cookies = cookies.encode('iso-8859-1')
-            cookies = cookies.decode('utf-8')
-            remixlhk = re.search(r'remixlhk=(.+?);.*?\bdomain=(.+?)(?:[,;]|$)', cookies)
-            if remixlhk:
-                value, domain = remixlhk.groups()
-                self._set_cookie(domain, 'remixlhk', value)
+        # vk serves two same remixlhk cookies in Set-Cookie header and expects
+        # first one to be actually set
+        self._apply_first_set_cookie_header(url_handle, 'remixlhk')
 
         login_page = self._download_webpage(
             'https://login.vk.com/?act=login', None,
-            note='Logging in as %s' % username,
+            note='Logging in',
             data=urlencode_postdata(login_form))
 
         if re.search(r'onLoginFailed', login_page):
@@ -96,28 +82,30 @@ class VKIE(VKBaseIE):
     _TESTS = [
         {
             'url': 'http://vk.com/videos-77521?z=video-77521_162222515%2Fclub77521',
-            'md5': '0deae91935c54e00003c2a00646315f0',
+            'md5': '7babad3b85ea2e91948005b1b8b0cb84',
             'info_dict': {
-                'id': '162222515',
-                'ext': 'flv',
+                'id': '-77521_162222515',
+                'ext': 'mp4',
                 'title': 'ProtivoGunz - Хуёвая песня',
                 'uploader': 're:(?:Noize MC|Alexander Ilyashenko).*',
+                'uploader_id': '-77521',
                 'duration': 195,
+                'timestamp': 1329049880,
                 'upload_date': '20120212',
-                'view_count': int,
             },
         },
         {
             'url': 'http://vk.com/video205387401_165548505',
             'md5': '6c0aeb2e90396ba97035b9cbde548700',
             'info_dict': {
-                'id': '165548505',
+                'id': '205387401_165548505',
                 'ext': 'mp4',
-                'uploader': 'Tom Cruise',
                 'title': 'No name',
+                'uploader': 'Tom Cruise',
+                'uploader_id': '205387401',
                 'duration': 9,
-                'upload_date': '20130721',
-                'view_count': int,
+                'timestamp': 1374364108,
+                'upload_date': '20130720',
             }
         },
         {
@@ -125,7 +113,7 @@ class VKIE(VKBaseIE):
             'url': 'http://vk.com/video_ext.php?oid=32194266&id=162925554&hash=7d8c2e0d5e05aeaa&hd=1',
             'md5': 'c7ce8f1f87bec05b3de07fdeafe21a0a',
             'info_dict': {
-                'id': '162925554',
+                'id': '32194266_162925554',
                 'ext': 'mp4',
                 'uploader': 'Vladimir Gavrin',
                 'title': 'Lin Dan',
@@ -142,7 +130,7 @@ class VKIE(VKBaseIE):
             'md5': 'a590bcaf3d543576c9bd162812387666',
             'note': 'Only available for registered users',
             'info_dict': {
-                'id': '164049491',
+                'id': '-8871596_164049491',
                 'ext': 'mp4',
                 'uploader': 'Триллеры',
                 'title': '► Бойцовский клуб / Fight Club 1999 [HD 720]',
@@ -156,7 +144,7 @@ class VKIE(VKBaseIE):
             'url': 'http://vk.com/hd_kino_mania?z=video-43215063_168067957%2F15c66b9b533119788d',
             'md5': '4d7a5ef8cf114dfa09577e57b2993202',
             'info_dict': {
-                'id': '168067957',
+                'id': '-43215063_168067957',
                 'ext': 'mp4',
                 'uploader': 'Киномания - лучшее из мира кино',
                 'title': ' ',
@@ -170,7 +158,7 @@ class VKIE(VKBaseIE):
             'md5': '0c45586baa71b7cb1d0784ee3f4e00a6',
             'note': 'ivi.ru embed',
             'info_dict': {
-                'id': '60690',
+                'id': '-43215063_169084319',
                 'ext': 'mp4',
                 'title': 'Книга Илая',
                 'duration': 6771,
@@ -184,7 +172,7 @@ class VKIE(VKBaseIE):
             'url': 'https://vk.com/video30481095_171201961?list=8764ae2d21f14088d4',
             'md5': '091287af5402239a1051c37ec7b92913',
             'info_dict': {
-                'id': '171201961',
+                'id': '30481095_171201961',
                 'ext': 'mp4',
                 'title': 'ТюменцевВВ_09.07.2015',
                 'uploader': 'Anton Ivanov',
@@ -192,35 +180,80 @@ class VKIE(VKBaseIE):
                 'upload_date': '20150709',
                 'view_count': int,
             },
+            'skip': 'Removed',
         },
         {
             # youtube embed
             'url': 'https://vk.com/video276849682_170681728',
             'info_dict': {
                 'id': 'V3K4mi0SYkc',
-                'ext': 'webm',
+                'ext': 'mp4',
                 'title': "DSWD Awards 'Children's Joy Foundation, Inc.' Certificate of Registration and License to Operate",
-                'description': 'md5:d9903938abdc74c738af77f527ca0596',
+                'description': 'md5:bf9c26cfa4acdfb146362682edd3827a',
                 'duration': 178,
                 'upload_date': '20130116',
-                'uploader': "Children's Joy Foundation",
+                'uploader': "Children's Joy Foundation Inc.",
                 'uploader_id': 'thecjf',
                 'view_count': int,
             },
         },
+        {
+            # dailymotion embed
+            'url': 'https://vk.com/video-37468416_456239855',
+            'info_dict': {
+                'id': 'k3lz2cmXyRuJQSjGHUv',
+                'ext': 'mp4',
+                'title': 'md5:d52606645c20b0ddbb21655adaa4f56f',
+                # TODO: fix test by fixing dailymotion description extraction
+                'description': 'md5:c651358f03c56f1150b555c26d90a0fd',
+                'uploader': 'AniLibria.Tv',
+                'upload_date': '20160914',
+                'uploader_id': 'x1p5vl5',
+                'timestamp': 1473877246,
+            },
+            'params': {
+                'skip_download': True,
+            },
+        },
         {
             # video key is extra_data not url\d+
             'url': 'http://vk.com/video-110305615_171782105',
             'md5': 'e13fcda136f99764872e739d13fac1d1',
             'info_dict': {
-                'id': '171782105',
+                'id': '-110305615_171782105',
                 'ext': 'mp4',
                 'title': 'S-Dance, репетиции к The way show',
                 'uploader': 'THE WAY SHOW | 17 апреля',
+                'uploader_id': '-110305615',
+                'timestamp': 1454859345,
                 'upload_date': '20160207',
-                'view_count': int,
+            },
+            'params': {
+                'skip_download': True,
             },
         },
+        {
+            # finished live stream, postlive_mp4
+            'url': 'https://vk.com/videos-387766?z=video-387766_456242764%2Fpl_-387766_-2',
+            'info_dict': {
+                'id': '-387766_456242764',
+                'ext': 'mp4',
+                'title': 'ИгроМир 2016 День 1 — Игромания Утром',
+                'uploader': 'Игромания',
+                'duration': 5239,
+                # TODO: use act=show to extract view_count
+                # 'view_count': int,
+                'upload_date': '20160929',
+                'uploader_id': '-387766',
+                'timestamp': 1475137527,
+            },
+        },
+        {
+            # live stream, hls and rtmp links, most likely already finished live
+            # stream by the time you are reading this comment
+            'url': 'https://vk.com/video-140332_456239111',
+            'only_matching': True,
+        },
         {
             # removed video, just testing that we match the pattern
             'url': 'http://vk.com/feed?z=video-43215063_166094326%2Fbb50cacd3177146d7a',
@@ -239,15 +272,24 @@ class VKIE(VKBaseIE):
         {
             'url': 'http://new.vk.com/video205387401_165548505',
             'only_matching': True,
-        }
-    ]
+        },
+        {
+            # This video is no longer available, because its author has been blocked.
+            'url': 'https://vk.com/video-10639516_456240611',
+            'only_matching': True,
+        },
+        {
+            # The video is not available in your region.
+            'url': 'https://vk.com/video-51812607_171445436',
+            'only_matching': True,
+        }]
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         video_id = mobj.group('videoid')
 
         if video_id:
-            info_url = 'https://vk.com/al_video.php?act=show&al=1&module=video&video=%s' % video_id
+            info_url = 'https://vk.com/al_video.php?act=show_inline&al=1&video=' + video_id
             # Some videos (removed?) can only be downloaded with list id specified
             list_id = mobj.group('list_id')
             if list_id:
@@ -270,9 +312,14 @@ class VKIE(VKBaseIE):
                 'You are trying to log in from an unusual location. You should confirm ownership at vk.com to log in with this IP.',
                 expected=True)
 
+        ERROR_COPYRIGHT = 'Video %s has been removed from public access due to rightholder complaint.'
+
         ERRORS = {
             r'>Видеозапись .*? была изъята из публичного доступа в связи с обращением правообладателя.<':
-            'Video %s has been removed from public access due to rightholder complaint.',
+            ERROR_COPYRIGHT,
+
+            r'>The video .*? was removed from public access by request of the copyright holder.<':
+            ERROR_COPYRIGHT,
 
             r'<!>Please log in or <':
             'Video %s is only available for registered users, '
@@ -286,19 +333,29 @@ class VKIE(VKBaseIE):
 
             r'<!>Access denied':
             'Access denied to video %s.',
+
+            r'<!>Видеозапись недоступна, так как её автор был заблокирован.':
+            'Video %s is no longer available, because its author has been blocked.',
+
+            r'<!>This video is no longer available, because its author has been blocked.':
+            'Video %s is no longer available, because its author has been blocked.',
+
+            r'<!>This video is no longer available, because it has been deleted.':
+            'Video %s is no longer available, because it has been deleted.',
+
+            r'<!>The video .+? is not available in your region.':
+            'Video %s is not available in your region.',
         }
 
         for error_re, error_msg in ERRORS.items():
             if re.search(error_re, info_page):
                 raise ExtractorError(error_msg % video_id, expected=True)
 
-        youtube_url = self._search_regex(
-            r'<iframe[^>]+src="((?:https?:)?//www.youtube.com/embed/[^"]+)"',
-            info_page, 'youtube iframe', default=None)
+        youtube_url = YoutubeIE._extract_url(info_page)
         if youtube_url:
-            return self.url_result(youtube_url, 'Youtube')
+            return self.url_result(youtube_url, ie=YoutubeIE.ie_key())
 
-        vimeo_url = VimeoIE._extract_vimeo_url(url, info_page)
+        vimeo_url = VimeoIE._extract_url(url, info_page)
         if vimeo_url is not None:
             return self.url_result(vimeo_url)
 
@@ -313,6 +370,10 @@ class VKIE(VKBaseIE):
                 m_rutube.group(1).replace('\\', ''))
             return self.url_result(rutube_url)
 
+        dailymotion_urls = DailymotionIE._extract_urls(info_page)
+        if dailymotion_urls:
+            return self.url_result(dailymotion_urls[0], DailymotionIE.ie_key())
+
         m_opts = re.search(r'(?s)var\s+opts\s*=\s*({.+?});', info_page)
         if m_opts:
             m_opts_url = re.search(r"url\s*:\s*'((?!/\b)[^']+)", m_opts.group(1))
@@ -322,46 +383,93 @@ class VKIE(VKBaseIE):
                     opts_url = 'http:' + opts_url
                 return self.url_result(opts_url)
 
-        data_json = self._search_regex(r'var\s+vars\s*=\s*({.+?});', info_page, 'vars')
-        data = json.loads(data_json)
-
-        # Extract upload date
-        upload_date = None
-        mobj = re.search(r'id="mv_date(?:_views)?_wrap"[^>]*>([a-zA-Z]+ [0-9]+), ([0-9]+) at', info_page)
-        if mobj is not None:
-            mobj.group(1) + ' ' + mobj.group(2)
-            upload_date = unified_strdate(mobj.group(1) + ' ' + mobj.group(2))
-
-        view_count = None
-        views = self._html_search_regex(
-            r'"mv_views_count_number"[^>]*>(.+?\bviews?)<',
-            info_page, 'view count', default=None)
-        if views:
-            view_count = str_to_int(self._search_regex(
-                r'([\d,.]+)', views, 'view count', fatal=False))
+        # vars does not look to be served anymore since 24.10.2016
+        data = self._parse_json(
+            self._search_regex(
+                r'var\s+vars\s*=\s*({.+?});', info_page, 'vars', default='{}'),
+            video_id, fatal=False)
+
+        # <!json> is served instead
+        if not data:
+            data = self._parse_json(
+                self._search_regex(
+                    [r'<!json>\s*({.+?})\s*<!>', r'<!json>\s*({.+})'],
+                    info_page, 'json', default='{}'),
+                video_id)
+            if data:
+                data = data['player']['params'][0]
+
+        if not data:
+            data = self._parse_json(
+                self._search_regex(
+                    r'var\s+playerParams\s*=\s*({.+?})\s*;\s*\n', info_page,
+                    'player params', default='{}'),
+                video_id)
+            if data:
+                data = data['params'][0]
+
+        # <!--{...}
+        if not data:
+            data = self._parse_json(
+                self._search_regex(
+                    r'<!--\s*({.+})', info_page, 'payload'),
+                video_id)['payload'][-1][-1]['player']['params'][0]
+
+        title = unescapeHTML(data['md_title'])
+
+        # 2 = live
+        # 3 = post live (finished live)
+        is_live = data.get('live') == 2
+        if is_live:
+            title = self._live_title(title)
+
+        timestamp = unified_timestamp(self._html_search_regex(
+            r'class=["\']mv_info_date[^>]+>([^<]+)(?:<|from)', info_page,
+            'upload date', default=None)) or int_or_none(data.get('date'))
+
+        view_count = str_to_int(self._search_regex(
+            r'class=["\']mv_views_count[^>]+>\s*([\d,.]+)',
+            info_page, 'view count', default=None))
 
         formats = []
-        for k, v in data.items():
-            if not k.startswith('url') and not k.startswith('cache') and k != 'extra_data' or not v:
+        for format_id, format_url in data.items():
+            format_url = url_or_none(format_url)
+            if not format_url or not format_url.startswith(('http', '//', 'rtmp')):
                 continue
-            height = int_or_none(self._search_regex(
-                r'^(?:url|cache)(\d+)', k, 'height', default=None))
-            formats.append({
-                'format_id': k,
-                'url': v,
-                'height': height,
-            })
+            if (format_id.startswith(('url', 'cache'))
+                    or format_id in ('extra_data', 'live_mp4', 'postlive_mp4')):
+                height = int_or_none(self._search_regex(
+                    r'^(?:url|cache)(\d+)', format_id, 'height', default=None))
+                formats.append({
+                    'format_id': format_id,
+                    'url': format_url,
+                    'height': height,
+                })
+            elif format_id == 'hls':
+                formats.extend(self._extract_m3u8_formats(
+                    format_url, video_id, 'mp4', 'm3u8_native',
+                    m3u8_id=format_id, fatal=False, live=is_live))
+            elif format_id == 'rtmp':
+                formats.append({
+                    'format_id': format_id,
+                    'url': format_url,
+                    'ext': 'flv',
+                })
         self._sort_formats(formats)
 
         return {
-            'id': compat_str(data['vid']),
+            'id': video_id,
             'formats': formats,
-            'title': unescapeHTML(data['md_title']),
+            'title': title,
             'thumbnail': data.get('jpg'),
             'uploader': data.get('md_author'),
+            'uploader_id': str_or_none(data.get('author_id')),
             'duration': data.get('duration'),
-            'upload_date': upload_date,
+            'timestamp': timestamp,
             'view_count': view_count,
+            'like_count': int_or_none(data.get('liked')),
+            'dislike_count': int_or_none(data.get('nolikes')),
+            'is_live': is_live,
         }
 
 
@@ -445,6 +553,9 @@ class VKWallPostIE(VKBaseIE):
                 'skip_download': True,
             },
         }],
+        'params': {
+            'usenetrc': True,
+        },
         'skip': 'Requires vk account credentials',
     }, {
         # single YouTube embed, no leading -
@@ -454,6 +565,9 @@ class VKWallPostIE(VKBaseIE):
             'title': 'Sergey Gorbunov - Wall post 85155021_6319',
         },
         'playlist_count': 1,
+        'params': {
+            'usenetrc': True,
+        },
         'skip': 'Requires vk account credentials',
     }, {
         # wall page URL
@@ -481,37 +595,41 @@ class VKWallPostIE(VKBaseIE):
             raise ExtractorError('VK said: %s' % error, expected=True)
 
         description = clean_html(get_element_by_class('wall_post_text', webpage))
-        uploader = clean_html(get_element_by_class(
-            'fw_post_author', webpage)) or self._og_search_description(webpage)
+        uploader = clean_html(get_element_by_class('author', webpage))
         thumbnail = self._og_search_thumbnail(webpage)
 
         entries = []
 
-        for audio in re.finditer(r'''(?sx)
-                            <input[^>]+
-                                id=(?P<q1>["\'])audio_info(?P<id>\d+_\d+).*?(?P=q1)[^>]+
-                                value=(?P<q2>["\'])(?P<url>http.+?)(?P=q2)
-                                .+?
-                            </table>''', webpage):
-            audio_html = audio.group(0)
-            audio_id = audio.group('id')
-            duration = parse_duration(get_element_by_class('duration', audio_html))
-            track = self._html_search_regex(
-                r'<span[^>]+id=["\']title%s[^>]*>([^<]+)' % audio_id,
-                audio_html, 'title', default=None)
-            artist = self._html_search_regex(
-                r'>([^<]+)</a></b>\s*&ndash', audio_html,
-                'artist', default=None)
-            entries.append({
-                'id': audio_id,
-                'url': audio.group('url'),
-                'title': '%s - %s' % (artist, track) if artist and track else audio_id,
-                'thumbnail': thumbnail,
-                'duration': duration,
-                'uploader': uploader,
-                'artist': artist,
-                'track': track,
-            })
+        audio_ids = re.findall(r'data-full-id=["\'](\d+_\d+)', webpage)
+        if audio_ids:
+            al_audio = self._download_webpage(
+                'https://vk.com/al_audio.php', post_id,
+                note='Downloading audio info', fatal=False,
+                data=urlencode_postdata({
+                    'act': 'reload_audio',
+                    'al': '1',
+                    'ids': ','.join(audio_ids)
+                }))
+            if al_audio:
+                Audio = collections.namedtuple(
+                    'Audio', ['id', 'user_id', 'url', 'track', 'artist', 'duration'])
+                audios = self._parse_json(
+                    self._search_regex(
+                        r'<!json>(.+?)<!>', al_audio, 'audios', default='[]'),
+                    post_id, fatal=False, transform_source=unescapeHTML)
+                if isinstance(audios, list):
+                    for audio in audios:
+                        a = Audio._make(audio[:6])
+                        entries.append({
+                            'id': '%s_%s' % (a.user_id, a.id),
+                            'url': a.url,
+                            'title': '%s - %s' % (a.artist, a.track) if a.artist and a.track else a.id,
+                            'thumbnail': thumbnail,
+                            'duration': a.duration,
+                            'uploader': uploader,
+                            'artist': a.artist,
+                            'track': a.track,
+                        })
 
         for video in re.finditer(
                 r'<a[^>]+href=(["\'])(?P<url>/video(?:-?[\d_]+).*?)\1', webpage):