Merge branch 'master' of https://github.com/rg3/youtube-dl into bilibili

1 files changed, 71 insertions, 68 deletions

diff --git a/youtube_dl/extractor/vimeo.py b/youtube_dl/extractor/vimeo.py
index fa1b22049..f392ccf1c 100644
--- a/youtube_dl/extractor/vimeo.py
+++ b/youtube_dl/extractor/vimeo.py
@@ -8,15 +8,15 @@ import itertools
 from .common import InfoExtractor
 from ..compat import (
     compat_HTTPError,
-    compat_urllib_parse,
-    compat_urllib_request,
     compat_urlparse,
 )
 from ..utils import (
+    encode_dict,
     ExtractorError,
     InAdvancePagedList,
     int_or_none,
     RegexNotFoundError,
+    sanitized_Request,
     smuggle_url,
     std_headers,
     unified_strdate,
@@ -40,17 +40,17 @@ class VimeoBaseInfoExtractor(InfoExtractor):
         self.report_login()
         webpage = self._download_webpage(self._LOGIN_URL, None, False)
         token, vuid = self._extract_xsrft_and_vuid(webpage)
-        data = urlencode_postdata({
+        data = urlencode_postdata(encode_dict({
             'action': 'login',
             'email': username,
             'password': password,
             'service': 'vimeo',
             'token': token,
-        })
-        login_request = compat_urllib_request.Request(self._LOGIN_URL, data)
+        }))
+        login_request = sanitized_Request(self._LOGIN_URL, data)
         login_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
-        login_request.add_header('Cookie', 'vuid=%s' % vuid)
         login_request.add_header('Referer', self._LOGIN_URL)
+        self._set_vimeo_cookie('vuid', vuid)
         self._download_webpage(login_request, None, False, 'Wrong login info')
 
     def _extract_xsrft_and_vuid(self, webpage):
@@ -62,6 +62,9 @@ class VimeoBaseInfoExtractor(InfoExtractor):
             webpage, 'vuid', group='vuid')
         return xsrft, vuid
 
+    def _set_vimeo_cookie(self, name, value):
+        self._set_cookie('vimeo.com', name, value)
+
 
 class VimeoIE(VimeoBaseInfoExtractor):
     """Information extractor for vimeo.com."""
@@ -133,7 +136,7 @@ class VimeoIE(VimeoBaseInfoExtractor):
                 'uploader_id': 'user18948128',
                 'uploader': 'Jaime Marquínez Ferrándiz',
                 'duration': 10,
-                'description': 'This is "youtube-dl password protected test video" by Jaime Marquínez Ferrándiz on Vimeo, the home for high quality videos and the people who love them.',
+                'description': 'This is "youtube-dl password protected test video" by Jaime Marquínez Ferrándiz on Vimeo, the home for high quality videos and the people\u2026',
             },
             'params': {
                 'videopassword': 'youtube-dl',
@@ -181,6 +184,15 @@ class VimeoIE(VimeoBaseInfoExtractor):
                 'uploader_id': 'user28849593',
             },
         },
+        {
+            'url': 'https://vimeo.com/109815029',
+            'note': 'Video not completely processed, "failed" seed status',
+            'only_matching': True,
+        },
+        {
+            'url': 'https://vimeo.com/groups/travelhd/videos/22439234',
+            'only_matching': True,
+        },
     ]
 
     @staticmethod
@@ -203,17 +215,17 @@ class VimeoIE(VimeoBaseInfoExtractor):
         if password is None:
             raise ExtractorError('This video is protected by a password, use the --video-password option', expected=True)
         token, vuid = self._extract_xsrft_and_vuid(webpage)
-        data = urlencode_postdata({
+        data = urlencode_postdata(encode_dict({
             'password': password,
             'token': token,
-        })
+        }))
         if url.startswith('http://'):
             # vimeo only supports https now, but the user can give an http url
             url = url.replace('http://', 'https://')
-        password_request = compat_urllib_request.Request(url + '/password', data)
+        password_request = sanitized_Request(url + '/password', data)
         password_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
-        password_request.add_header('Cookie', 'clip_test2=1; vuid=%s' % vuid)
         password_request.add_header('Referer', url)
+        self._set_vimeo_cookie('vuid', vuid)
         return self._download_webpage(
             password_request, video_id,
             'Verifying the password', 'Wrong password')
@@ -222,9 +234,9 @@ class VimeoIE(VimeoBaseInfoExtractor):
         password = self._downloader.params.get('videopassword', None)
         if password is None:
             raise ExtractorError('This video is protected by a password, use the --video-password option')
-        data = compat_urllib_parse.urlencode({'password': password})
+        data = urlencode_postdata(encode_dict({'password': password}))
         pass_url = url + '/check-password'
-        password_request = compat_urllib_request.Request(pass_url, data)
+        password_request = sanitized_Request(pass_url, data)
         password_request.add_header('Content-Type', 'application/x-www-form-urlencoded')
         return self._download_json(
             password_request, video_id,
@@ -253,7 +265,7 @@ class VimeoIE(VimeoBaseInfoExtractor):
             url = 'https://vimeo.com/' + video_id
 
         # Retrieve video webpage to extract further information
-        request = compat_urllib_request.Request(url, None, headers)
+        request = sanitized_Request(url, None, headers)
         try:
             webpage = self._download_webpage(request, video_id)
         except ExtractorError as ee:
@@ -273,20 +285,30 @@ class VimeoIE(VimeoBaseInfoExtractor):
         self.report_extraction(video_id)
 
         vimeo_config = self._search_regex(
-            r'vimeo\.config\s*=\s*({.+?});', webpage,
+            r'vimeo\.config\s*=\s*(?:({.+?})|_extend\([^,]+,\s+({.+?})\));', webpage,
             'vimeo config', default=None)
         if vimeo_config:
             seed_status = self._parse_json(vimeo_config, video_id).get('seed_status', {})
             if seed_status.get('state') == 'failed':
                 raise ExtractorError(
-                    '%s returned error: %s' % (self.IE_NAME, seed_status['title']),
+                    '%s said: %s' % (self.IE_NAME, seed_status['title']),
                     expected=True)
 
         # Extract the config JSON
         try:
             try:
                 config_url = self._html_search_regex(
-                    r' data-config-url="(.+?)"', webpage, 'config URL')
+                    r' data-config-url="(.+?)"', webpage,
+                    'config URL', default=None)
+                if not config_url:
+                    # Sometimes new react-based page is served instead of old one that require
+                    # different config URL extraction approach (see
+                    # https://github.com/rg3/youtube-dl/pull/7209)
+                    vimeo_clip_page_config = self._search_regex(
+                        r'vimeo\.clip_page_config\s*=\s*({.+?});', webpage,
+                        'vimeo clip page config')
+                    config_url = self._parse_json(
+                        vimeo_clip_page_config, video_id)['player']['config_url']
                 config_json = self._download_webpage(config_url, video_id)
                 config = json.loads(config_json)
             except RegexNotFoundError:
@@ -369,47 +391,29 @@ class VimeoIE(VimeoBaseInfoExtractor):
             like_count = None
             comment_count = None
 
-        # Vimeo specific: extract request signature and timestamp
-        sig = config['request']['signature']
-        timestamp = config['request']['timestamp']
-
-        # Vimeo specific: extract video codec and quality information
-        # First consider quality, then codecs, then take everything
-        codecs = [('vp6', 'flv'), ('vp8', 'flv'), ('h264', 'mp4')]
-        files = {'hd': [], 'sd': [], 'other': []}
-        config_files = config["video"].get("files") or config["request"].get("files")
-        for codec_name, codec_extension in codecs:
-            for quality in config_files.get(codec_name, []):
-                format_id = '-'.join((codec_name, quality)).lower()
-                key = quality if quality in files else 'other'
-                video_url = None
-                if isinstance(config_files[codec_name], dict):
-                    file_info = config_files[codec_name][quality]
-                    video_url = file_info.get('url')
-                else:
-                    file_info = {}
-                if video_url is None:
-                    video_url = "http://player.vimeo.com/play_redirect?clip_id=%s&sig=%s&time=%s&quality=%s&codecs=%s&type=moogaloop_local&embed_location=" \
-                        % (video_id, sig, timestamp, quality, codec_name.upper())
-
-                files[key].append({
-                    'ext': codec_extension,
-                    'url': video_url,
-                    'format_id': format_id,
-                    'width': int_or_none(file_info.get('width')),
-                    'height': int_or_none(file_info.get('height')),
-                    'tbr': int_or_none(file_info.get('bitrate')),
-                })
         formats = []
-        m3u8_url = config_files.get('hls', {}).get('all')
+        config_files = config['video'].get('files') or config['request'].get('files', {})
+        for f in config_files.get('progressive', []):
+            video_url = f.get('url')
+            if not video_url:
+                continue
+            formats.append({
+                'url': video_url,
+                'format_id': 'http-%s' % f.get('quality'),
+                'width': int_or_none(f.get('width')),
+                'height': int_or_none(f.get('height')),
+                'fps': int_or_none(f.get('fps')),
+                'tbr': int_or_none(f.get('bitrate')),
+            })
+        m3u8_url = config_files.get('hls', {}).get('url')
         if m3u8_url:
             m3u8_formats = self._extract_m3u8_formats(
                 m3u8_url, video_id, 'mp4', 'm3u8_native', 0, 'hls', fatal=False)
             if m3u8_formats:
                 formats.extend(m3u8_formats)
-        for key in ('other', 'sd', 'hd'):
-            formats += files[key]
-        self._sort_formats(formats)
+        # Bitrates are completely broken. Single m3u8 may contain entries in kbps and bps
+        # at the same time without actual units specified. This lead to wrong sorting.
+        self._sort_formats(formats, field_preference=('height', 'width', 'fps', 'format_id'))
 
         subtitles = {}
         text_tracks = config['request'].get('text_tracks')
@@ -473,21 +477,20 @@ class VimeoChannelIE(VimeoBaseInfoExtractor):
         token, vuid = self._extract_xsrft_and_vuid(webpage)
         fields['token'] = token
         fields['password'] = password
-        post = urlencode_postdata(fields)
+        post = urlencode_postdata(encode_dict(fields))
         password_path = self._search_regex(
             r'action="([^"]+)"', login_form, 'password URL')
         password_url = compat_urlparse.urljoin(page_url, password_path)
-        password_request = compat_urllib_request.Request(password_url, post)
+        password_request = sanitized_Request(password_url, post)
         password_request.add_header('Content-type', 'application/x-www-form-urlencoded')
-        password_request.add_header('Cookie', 'vuid=%s' % vuid)
-        self._set_cookie('vimeo.com', 'xsrft', token)
+        self._set_vimeo_cookie('vuid', vuid)
+        self._set_vimeo_cookie('xsrft', token)
 
         return self._download_webpage(
             password_request, list_id,
             'Verifying the password', 'Wrong password')
 
-    def _extract_videos(self, list_id, base_url):
-        video_ids = []
+    def _title_and_entries(self, list_id, base_url):
         for pagenum in itertools.count(1):
             page_url = self._page_url(base_url, pagenum)
             webpage = self._download_webpage(
@@ -496,18 +499,18 @@ class VimeoChannelIE(VimeoBaseInfoExtractor):
 
             if pagenum == 1:
                 webpage = self._login_list_password(page_url, list_id, webpage)
+                yield self._extract_list_title(webpage)
+
+            for video_id in re.findall(r'id="clip_(\d+?)"', webpage):
+                yield self.url_result('https://vimeo.com/%s' % video_id, 'Vimeo')
 
-            video_ids.extend(re.findall(r'id="clip_(\d+?)"', webpage))
             if re.search(self._MORE_PAGES_INDICATOR, webpage, re.DOTALL) is None:
                 break
 
-        entries = [self.url_result('https://vimeo.com/%s' % video_id, 'Vimeo')
-                   for video_id in video_ids]
-        return {'_type': 'playlist',
-                'id': list_id,
-                'title': self._extract_list_title(webpage),
-                'entries': entries,
-                }
+    def _extract_videos(self, list_id, base_url):
+        title_and_entries = self._title_and_entries(list_id, base_url)
+        list_title = next(title_and_entries)
+        return self.playlist_result(title_and_entries, list_id, list_title)
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
@@ -568,7 +571,7 @@ class VimeoAlbumIE(VimeoChannelIE):
 
 class VimeoGroupsIE(VimeoAlbumIE):
     IE_NAME = 'vimeo:group'
-    _VALID_URL = r'https://vimeo\.com/groups/(?P<name>[^/]+)'
+    _VALID_URL = r'https://vimeo\.com/groups/(?P<name>[^/]+)(?:/(?!videos?/\d+)|$)'
     _TESTS = [{
         'url': 'https://vimeo.com/groups/rolexawards',
         'info_dict': {
@@ -637,7 +640,7 @@ class VimeoWatchLaterIE(VimeoChannelIE):
 
     def _page_url(self, base_url, pagenum):
         url = '%s/page:%d/' % (base_url, pagenum)
-        request = compat_urllib_request.Request(url)
+        request = sanitized_Request(url)
         # Set the header to get a partial html page with the ids,
         # the normal page doesn't contain them.
         request.add_header('X-Requested-With', 'XMLHttpRequest')