Merge branch 'master' into totalwebcasting

youtube_dl/YoutubeDL.py

@@ -24,14 +24,17 @@ import sys
 import time
 import tokenize
 import traceback
+import random
+
+from string import ascii_letters
 
 from .compat import (
     compat_basestring,
     compat_cookiejar,
-    compat_expanduser,
     compat_get_terminal_size,
     compat_http_client,
     compat_kwargs,
+    compat_numeric_types,
     compat_os_name,
     compat_str,
     compat_tokenize_tokenize,
@@ -52,12 +55,17 @@ from .utils import (
     encode_compat_str,
     encodeFilename,
     error_to_compat_str,
+    expand_path,
     ExtractorError,
     format_bytes,
     formatSeconds,
+    GeoRestrictedError,
+    int_or_none,
+    ISO3166Utils,
     locked_file,
     make_HTTPS_handler,
     MaxDownloadsReached,
+    orderedSet,
     PagedList,
     parse_filesize,
     PerRequestProxyHandler,
@@ -85,6 +93,7 @@ from .utils import (
 )
 from .cache import Cache
 from .extractor import get_info_extractor, gen_extractor_classes, _LAZY_LOADER
+from .extractor.openload import PhantomJSwrapper
 from .downloader import get_suitable_downloader
 from .downloader.rtmp import rtmpdump_version
 from .postprocessor import (
@@ -159,6 +168,7 @@ class YoutubeDL(object):
     playlistend:       Playlist item to end at.
     playlist_items:    Specific indices of playlist to download.
     playlistreverse:   Download playlist items in reverse order.
+    playlistrandom:    Download playlist items in random order.
     matchtitle:        Download only matching titles.
     rejecttitle:       Reject downloads for matching titles.
     logger:            Log messages to a logging.Logger instance.
@@ -270,6 +280,12 @@ class YoutubeDL(object):
                        If it returns None, the video is downloaded.
                        match_filter_func in utils.py is one example for this.
     no_color:          Do not emit color codes in output.
+    geo_bypass:        Bypass geographic restriction via faking X-Forwarded-For
+                       HTTP header (experimental)
+    geo_bypass_country:
+                       Two-letter ISO 3166-2 country code that will be used for
+                       explicit geographic restriction bypassing via faking
+                       X-Forwarded-For HTTP header (experimental)
 
     The following options determine which downloader is picked:
     external_downloader: Executable of the external downloader to call.
@@ -289,8 +305,25 @@ class YoutubeDL(object):
                        otherwise prefer avconv.
     postprocessor_args: A list of additional command-line arguments for the
                         postprocessor.
+
+    The following options are used by the Youtube extractor:
+    youtube_include_dash_manifest: If True (default), DASH manifests and related
+                        data will be downloaded and processed by extractor.
+                        You can reduce network I/O by disabling it if you don't
+                        care about DASH.
     """
 
+    _NUMERIC_FIELDS = set((
+        'width', 'height', 'tbr', 'abr', 'asr', 'vbr', 'fps', 'filesize', 'filesize_approx',
+        'timestamp', 'upload_year', 'upload_month', 'upload_day',
+        'duration', 'view_count', 'like_count', 'dislike_count', 'repost_count',
+        'average_rating', 'comment_count', 'age_limit',
+        'start_time', 'end_time',
+        'chapter_number', 'season_number', 'episode_number',
+        'track_number', 'disc_number', 'release_year',
+        'playlist_index',
+    ))
+
     params = None
     _ies = []
     _pps = []
@@ -317,11 +350,21 @@ class YoutubeDL(object):
         self.params.update(params)
         self.cache = Cache(self)
 
-        if self.params.get('cn_verification_proxy') is not None:
-            self.report_warning('--cn-verification-proxy is deprecated. Use --geo-verification-proxy instead.')
+        def check_deprecated(param, option, suggestion):
+            if self.params.get(param) is not None:
+                self.report_warning(
+                    '%s is deprecated. Use %s instead.' % (option, suggestion))
+                return True
+            return False
+
+        if check_deprecated('cn_verification_proxy', '--cn-verification-proxy', '--geo-verification-proxy'):
             if self.params.get('geo_verification_proxy') is None:
                 self.params['geo_verification_proxy'] = self.params['cn_verification_proxy']
 
+        check_deprecated('autonumber_size', '--autonumber-size', 'output template with %(autonumber)0Nd, where N in the number of digits')
+        check_deprecated('autonumber', '--auto-number', '-o "%(autonumber)s-%(title)s.%(ext)s"')
+        check_deprecated('usetitle', '--title', '-o "%(title)s-%(id)s.%(ext)s"')
+
         if params.get('bidi_workaround', False):
             try:
                 import pty
@@ -349,10 +392,10 @@ class YoutubeDL(object):
                 else:
                     raise
 
-        if (sys.version_info >= (3,) and sys.platform != 'win32' and
+        if (sys.platform != 'win32' and
                 sys.getfilesystemencoding() in ['ascii', 'ANSI_X3.4-1968'] and
                 not params.get('restrictfilenames', False)):
-            # On Python 3, the Unicode filesystem API will throw errors (#1474)
+            # Unicode filesystem API will throw errors (#1474, #13027)
             self.report_warning(
                 'Assuming --restrict-filenames since file system encoding '
                 'cannot encode all characters. '
@@ -477,24 +520,25 @@ class YoutubeDL(object):
     def to_console_title(self, message):
         if not self.params.get('consoletitle', False):
             return
-        if compat_os_name == 'nt' and ctypes.windll.kernel32.GetConsoleWindow():
-            # c_wchar_p() might not be necessary if `message` is
-            # already of type unicode()
-            ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
+        if compat_os_name == 'nt':
+            if ctypes.windll.kernel32.GetConsoleWindow():
+                # c_wchar_p() might not be necessary if `message` is
+                # already of type unicode()
+                ctypes.windll.kernel32.SetConsoleTitleW(ctypes.c_wchar_p(message))
         elif 'TERM' in os.environ:
             self._write_string('\033]0;%s\007' % message, self._screen_file)
 
     def save_console_title(self):
         if not self.params.get('consoletitle', False):
             return
-        if 'TERM' in os.environ:
+        if compat_os_name != 'nt' and 'TERM' in os.environ:
             # Save the title on stack
             self._write_string('\033[22;0t', self._screen_file)
 
     def restore_console_title(self):
         if not self.params.get('consoletitle', False):
             return
-        if 'TERM' in os.environ:
+        if compat_os_name != 'nt' and 'TERM' in os.environ:
             # Restore the title from stack
             self._write_string('\033[23;0t', self._screen_file)
 
@@ -583,10 +627,7 @@ class YoutubeDL(object):
             autonumber_size = self.params.get('autonumber_size')
             if autonumber_size is None:
                 autonumber_size = 5
-            autonumber_templ = '%0' + str(autonumber_size) + 'd'
-            template_dict['autonumber'] = autonumber_templ % self._num_downloads
-            if template_dict.get('playlist_index') is not None:
-                template_dict['playlist_index'] = '%0*d' % (len(str(template_dict['n_entries'])), template_dict['playlist_index'])
+            template_dict['autonumber'] = self.params.get('autonumber_start', 1) - 1 + self._num_downloads
             if template_dict.get('resolution') is None:
                 if template_dict.get('width') and template_dict.get('height'):
                     template_dict['resolution'] = '%dx%d' % (template_dict['width'], template_dict['height'])
@@ -598,15 +639,64 @@ class YoutubeDL(object):
             sanitize = lambda k, v: sanitize_filename(
                 compat_str(v),
                 restricted=self.params.get('restrictfilenames'),
-                is_id=(k == 'id'))
-            template_dict = dict((k, sanitize(k, v))
+                is_id=(k == 'id' or k.endswith('_id')))
+            template_dict = dict((k, v if isinstance(v, compat_numeric_types) else sanitize(k, v))
                                  for k, v in template_dict.items()
                                  if v is not None and not isinstance(v, (list, tuple, dict)))
             template_dict = collections.defaultdict(lambda: 'NA', template_dict)
 
             outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
-            tmpl = compat_expanduser(outtmpl)
-            filename = tmpl % template_dict
+
+            # For fields playlist_index and autonumber convert all occurrences
+            # of %(field)s to %(field)0Nd for backward compatibility
+            field_size_compat_map = {
+                'playlist_index': len(str(template_dict['n_entries'])),
+                'autonumber': autonumber_size,
+            }
+            FIELD_SIZE_COMPAT_RE = r'(?<!%)%\((?P<field>autonumber|playlist_index)\)s'
+            mobj = re.search(FIELD_SIZE_COMPAT_RE, outtmpl)
+            if mobj:
+                outtmpl = re.sub(
+                    FIELD_SIZE_COMPAT_RE,
+                    r'%%(\1)0%dd' % field_size_compat_map[mobj.group('field')],
+                    outtmpl)
+
+            # Missing numeric fields used together with integer presentation types
+            # in format specification will break the argument substitution since
+            # string 'NA' is returned for missing fields. We will patch output
+            # template for missing fields to meet string presentation type.
+            for numeric_field in self._NUMERIC_FIELDS:
+                if numeric_field not in template_dict:
+                    # As of [1] format syntax is:
+                    #  %[mapping_key][conversion_flags][minimum_width][.precision][length_modifier]type
+                    # 1. https://docs.python.org/2/library/stdtypes.html#string-formatting
+                    FORMAT_RE = r'''(?x)
+                        (?<!%)
+                        %
+                        \({0}\)  # mapping key
+                        (?:[#0\-+ ]+)?  # conversion flags (optional)
+                        (?:\d+)?  # minimum field width (optional)
+                        (?:\.\d+)?  # precision (optional)
+                        [hlL]?  # length modifier (optional)
+                        [diouxXeEfFgGcrs%]  # conversion type
+                    '''
+                    outtmpl = re.sub(
+                        FORMAT_RE.format(numeric_field),
+                        r'%({0})s'.format(numeric_field), outtmpl)
+
+            # expand_path translates '%%' into '%' and '$$' into '$'
+            # correspondingly that is not what we want since we need to keep
+            # '%%' intact for template dict substitution step. Working around
+            # with boundary-alike separator hack.
+            sep = ''.join([random.choice(ascii_letters) for _ in range(32)])
+            outtmpl = outtmpl.replace('%%', '%{0}%'.format(sep)).replace('$$', '${0}$'.format(sep))
+
+            # outtmpl should be expand_path'ed before template dict substitution
+            # because meta fields may contain env variables we don't want to
+            # be expanded. For example, for outtmpl "%(title)s.%(ext)s" and
+            # title "Hello $PATH", we don't want `$PATH` to be expanded.
+            filename = expand_path(outtmpl).replace(sep, '') % template_dict
+
             # Temporary fix for #4787
             # 'Treat' all problem characters by passing filename through preferredencoding
             # to workaround encoding issues with subprocess on python2 @ Windows
@@ -705,6 +795,14 @@ class YoutubeDL(object):
                     return self.process_ie_result(ie_result, download, extra_info)
                 else:
                     return ie_result
+            except GeoRestrictedError as e:
+                msg = e.msg
+                if e.countries:
+                    msg += '\nThis video is available in %s.' % ', '.join(
+                        map(ISO3166Utils.short2full, e.countries))
+                msg += '\nYou might want to use a VPN or a proxy server (with --proxy) to workaround.'
+                self.report_error(msg)
+                break
             except ExtractorError as e:  # An error we somewhat expected
                 self.report_error(compat_str(e), e.format_traceback())
                 break
@@ -762,19 +860,32 @@ class YoutubeDL(object):
                 ie_result['url'], ie_key=ie_result.get('ie_key'),
                 extra_info=extra_info, download=False, process=False)
 
+            # extract_info may return None when ignoreerrors is enabled and
+            # extraction failed with an error, don't crash and return early
+            # in this case
+            if not info:
+                return info
+
             force_properties = dict(
                 (k, v) for k, v in ie_result.items() if v is not None)
-            for f in ('_type', 'url', 'ie_key'):
+            for f in ('_type', 'url', 'id', 'extractor', 'extractor_key', 'ie_key'):
                 if f in force_properties:
                     del force_properties[f]
             new_result = info.copy()
             new_result.update(force_properties)
 
-            assert new_result.get('_type') != 'url_transparent'
+            # Extracted info may not be a video result (i.e.
+            # info.get('_type', 'video') != video) but rather an url or
+            # url_transparent. In such cases outer metadata (from ie_result)
+            # should be propagated to inner one (info). For this to happen
+            # _type of info should be overridden with url_transparent. This
+            # fixes issue from https://github.com/rg3/youtube-dl/pull/11163.
+            if new_result.get('_type') == 'url':
+                new_result['_type'] = 'url_transparent'
 
             return self.process_ie_result(
                 new_result, download=download, extra_info=extra_info)
-        elif result_type == 'playlist' or result_type == 'multi_video':
+        elif result_type in ('playlist', 'multi_video'):
             # We process each entry in the playlist
             playlist = ie_result.get('title') or ie_result.get('id')
             self.to_screen('[download] Downloading playlist: %s' % playlist)
@@ -798,15 +909,25 @@ class YoutubeDL(object):
                                 yield int(item)
                         else:
                             yield int(string_segment)
-                playlistitems = iter_playlistitems(playlistitems_str)
+                playlistitems = orderedSet(iter_playlistitems(playlistitems_str))
 
             ie_entries = ie_result['entries']
+
+            def make_playlistitems_entries(list_ie_entries):
+                num_entries = len(list_ie_entries)
+                return [
+                    list_ie_entries[i - 1] for i in playlistitems
+                    if -num_entries <= i - 1 < num_entries]
+
+            def report_download(num_entries):
+                self.to_screen(
+                    '[%s] playlist %s: Downloading %d videos' %
+                    (ie_result['extractor'], playlist, num_entries))
+
             if isinstance(ie_entries, list):
                 n_all_entries = len(ie_entries)
                 if playlistitems:
-                    entries = [
-                        ie_entries[i - 1] for i in playlistitems
-                        if -n_all_entries <= i - 1 < n_all_entries]
+                    entries = make_playlistitems_entries(ie_entries)
                 else:
                     entries = ie_entries[playliststart:playlistend]
                 n_entries = len(entries)
@@ -824,31 +945,38 @@ class YoutubeDL(object):
                     entries = ie_entries.getslice(
                         playliststart, playlistend)
                 n_entries = len(entries)
-                self.to_screen(
-                    '[%s] playlist %s: Downloading %d videos' %
-                    (ie_result['extractor'], playlist, n_entries))
+                report_download(n_entries)
             else:  # iterable
                 if playlistitems:
-                    entry_list = list(ie_entries)
-                    entries = [entry_list[i - 1] for i in playlistitems]
+                    entries = make_playlistitems_entries(list(itertools.islice(
+                        ie_entries, 0, max(playlistitems))))
                 else:
                     entries = list(itertools.islice(
                         ie_entries, playliststart, playlistend))
                 n_entries = len(entries)
-                self.to_screen(
-                    '[%s] playlist %s: Downloading %d videos' %
-                    (ie_result['extractor'], playlist, n_entries))
+                report_download(n_entries)
 
             if self.params.get('playlistreverse', False):
                 entries = entries[::-1]
 
+            if self.params.get('playlistrandom', False):
+                random.shuffle(entries)
+
+            x_forwarded_for = ie_result.get('__x_forwarded_for_ip')
+
             for i, entry in enumerate(entries, 1):
                 self.to_screen('[download] Downloading video %s of %s' % (i, n_entries))
+                # This __x_forwarded_for_ip thing is a bit ugly but requires
+                # minimal changes
+                if x_forwarded_for:
+                    entry['__x_forwarded_for_ip'] = x_forwarded_for
                 extra = {
                     'n_entries': n_entries,
                     'playlist': playlist,
                     'playlist_id': ie_result.get('id'),
                     'playlist_title': ie_result.get('title'),
+                    'playlist_uploader': ie_result.get('uploader'),
+                    'playlist_uploader_id': ie_result.get('uploader_id'),
                     'playlist_index': i + playliststart,
                     'extractor': ie_result['extractor'],
                     'webpage_url': ie_result['webpage_url'],
@@ -952,6 +1080,30 @@ class YoutubeDL(object):
             return op(actual_value, comparison_value)
         return _filter
 
+    def _default_format_spec(self, info_dict, download=True):
+
+        def can_merge():
+            merger = FFmpegMergerPP(self)
+            return merger.available and merger.can_merge()
+
+        def prefer_best():
+            if self.params.get('simulate', False):
+                return False
+            if not download:
+                return False
+            if self.params.get('outtmpl', DEFAULT_OUTTMPL) == '-':
+                return True
+            if info_dict.get('is_live'):
+                return True
+            if not can_merge():
+                return True
+            return False
+
+        req_format_list = ['bestvideo+bestaudio', 'best']
+        if prefer_best():
+            req_format_list.reverse()
+        return '/'.join(req_format_list)
+
     def build_format_selector(self, format_spec):
         def syntax_error(note, start):
             message = (
@@ -1228,6 +1380,11 @@ class YoutubeDL(object):
         if cookies:
             res['Cookie'] = cookies
 
+        if 'X-Forwarded-For' not in res:
+            x_forwarded_for_ip = info_dict.get('__x_forwarded_for_ip')
+            if x_forwarded_for_ip:
+                res['X-Forwarded-For'] = x_forwarded_for_ip
+
         return res
 
     def _calc_cookies(self, info_dict):
@@ -1243,9 +1400,28 @@ class YoutubeDL(object):
         if 'title' not in info_dict:
             raise ExtractorError('Missing "title" field in extractor result')
 
-        if not isinstance(info_dict['id'], compat_str):
-            self.report_warning('"id" field is not a string - forcing string conversion')
-            info_dict['id'] = compat_str(info_dict['id'])
+        def report_force_conversion(field, field_not, conversion):
+            self.report_warning(
+                '"%s" field is not %s - forcing %s conversion, there is an error in extractor'
+                % (field, field_not, conversion))
+
+        def sanitize_string_field(info, string_field):
+            field = info.get(string_field)
+            if field is None or isinstance(field, compat_str):
+                return
+            report_force_conversion(string_field, 'a string', 'string')
+            info[string_field] = compat_str(field)
+
+        def sanitize_numeric_fields(info):
+            for numeric_field in self._NUMERIC_FIELDS:
+                field = info.get(numeric_field)
+                if field is None or isinstance(field, compat_numeric_types):
+                    continue
+                report_force_conversion(numeric_field, 'numeric', 'int')
+                info[numeric_field] = int_or_none(field)
+
+        sanitize_string_field(info_dict, 'id')
+        sanitize_numeric_fields(info_dict)
 
         if 'playlist' not in info_dict:
             # It isn't part of a playlist
@@ -1326,16 +1502,28 @@ class YoutubeDL(object):
         if not formats:
             raise ExtractorError('No video formats found!')
 
+        def is_wellformed(f):
+            url = f.get('url')
+            if not url:
+                self.report_warning(
+                    '"url" field is missing or empty - skipping format, '
+                    'there is an error in extractor')
+                return False
+            if isinstance(url, bytes):
+                sanitize_string_field(f, 'url')
+            return True
+
+        # Filter out malformed formats for better extraction robustness
+        formats = list(filter(is_wellformed, formats))
+
         formats_dict = {}
 
         # We check that all the formats have the format and format_id fields
         for i, format in enumerate(formats):
-            if 'url' not in format:
-                raise ExtractorError('Missing "url" key in result (index %d)' % i)
-
+            sanitize_string_field(format, 'format_id')
+            sanitize_numeric_fields(format)
             format['url'] = sanitize_url(format['url'])
-
-            if format.get('format_id') is None:
+            if not format.get('format_id'):
                 format['format_id'] = compat_str(i)
             else:
                 # Sanitize format_id from characters used in format selector expression
@@ -1363,13 +1551,16 @@ class YoutubeDL(object):
                 format['ext'] = determine_ext(format['url']).lower()
             # Automatically determine protocol if missing (useful for format
             # selection purposes)
-            if 'protocol' not in format:
+            if format.get('protocol') is None:
                 format['protocol'] = determine_protocol(format)
             # Add HTTP headers, so that external programs can use them from the
             # json output
             full_format_info = info_dict.copy()
             full_format_info.update(format)
             format['http_headers'] = self._calc_headers(full_format_info)
+        # Remove private housekeeping stuff
+        if '__x_forwarded_for_ip' in info_dict:
+            del info_dict['__x_forwarded_for_ip']
 
         # TODO Central sorting goes here
 
@@ -1385,14 +1576,10 @@ class YoutubeDL(object):
 
         req_format = self.params.get('format')
         if req_format is None:
-            req_format_list = []
-            if (self.params.get('outtmpl', DEFAULT_OUTTMPL) != '-' and
-                    not info_dict.get('is_live')):
-                merger = FFmpegMergerPP(self)
-                if merger.available and merger.can_merge():
-                    req_format_list.append('bestvideo+bestaudio')
-            req_format_list.append('best')
-            req_format = '/'.join(req_format_list)
+            req_format = self._default_format_spec(info_dict, download=download)
+            if self.params.get('verbose'):
+                self.to_stdout('[debug] Default format spec: %s' % req_format)
+
         format_selector = self.build_format_selector(req_format)
 
         # While in format selection we may need to have an access to the original
@@ -1544,12 +1731,17 @@ class YoutubeDL(object):
         if filename is None:
             return
 
-        try:
-            dn = os.path.dirname(sanitize_path(encodeFilename(filename)))
-            if dn and not os.path.exists(dn):
-                os.makedirs(dn)
-        except (OSError, IOError) as err:
-            self.report_error('unable to create directory ' + error_to_compat_str(err))
+        def ensure_dir_exists(path):
+            try:
+                dn = os.path.dirname(path)
+                if dn and not os.path.exists(dn):
+                    os.makedirs(dn)
+                return True
+            except (OSError, IOError) as err:
+                self.report_error('unable to create directory ' + error_to_compat_str(err))
+                return False
+
+        if not ensure_dir_exists(sanitize_path(encodeFilename(filename))):
             return
 
         if self.params.get('writedescription', False):
@@ -1592,29 +1784,30 @@ class YoutubeDL(object):
             ie = self.get_info_extractor(info_dict['extractor_key'])
             for sub_lang, sub_info in subtitles.items():
                 sub_format = sub_info['ext']
-                if sub_info.get('data') is not None:
-                    sub_data = sub_info['data']
+                sub_filename = subtitles_filename(filename, sub_lang, sub_format)
+                if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
+                    self.to_screen('[info] Video subtitle %s.%s is already present' % (sub_lang, sub_format))
                 else:
-                    try:
-                        sub_data = ie._download_webpage(
-                            sub_info['url'], info_dict['id'], note=False)
-                    except ExtractorError as err:
-                        self.report_warning('Unable to download subtitle for "%s": %s' %
-                                            (sub_lang, error_to_compat_str(err.cause)))
-                        continue
-                try:
-                    sub_filename = subtitles_filename(filename, sub_lang, sub_format)
-                    if self.params.get('nooverwrites', False) and os.path.exists(encodeFilename(sub_filename)):
-                        self.to_screen('[info] Video subtitle %s.%s is already_present' % (sub_lang, sub_format))
+                    self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
+                    if sub_info.get('data') is not None:
+                        try:
+                            # Use newline='' to prevent conversion of newline characters
+                            # See https://github.com/rg3/youtube-dl/issues/10268
+                            with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
+                                subfile.write(sub_info['data'])
+                        except (OSError, IOError):
+                            self.report_error('Cannot write subtitles file ' + sub_filename)
+                            return
                     else:
-                        self.to_screen('[info] Writing video subtitles to: ' + sub_filename)
-                        # Use newline='' to prevent conversion of newline characters
-                        # See https://github.com/rg3/youtube-dl/issues/10268
-                        with io.open(encodeFilename(sub_filename), 'w', encoding='utf-8', newline='') as subfile:
-                            subfile.write(sub_data)
-                except (OSError, IOError):
-                    self.report_error('Cannot write subtitles file ' + sub_filename)
-                    return
+                        try:
+                            sub_data = ie._request_webpage(
+                                sub_info['url'], info_dict['id'], note=False).read()
+                            with io.open(encodeFilename(sub_filename), 'wb') as subfile:
+                                subfile.write(sub_data)
+                        except (ExtractorError, IOError, OSError, ValueError) as err:
+                            self.report_warning('Unable to download subtitle for "%s": %s' %
+                                                (sub_lang, error_to_compat_str(err)))
+                            continue
 
         if self.params.get('writeinfojson', False):
             infofn = replace_extension(filename, 'info.json', info_dict.get('ext'))
@@ -1687,8 +1880,11 @@ class YoutubeDL(object):
                         for f in requested_formats:
                             new_info = dict(info_dict)
                             new_info.update(f)
-                            fname = self.prepare_filename(new_info)
-                            fname = prepend_extension(fname, 'f%s' % f['format_id'], new_info['ext'])
+                            fname = prepend_extension(
+                                self.prepare_filename(new_info),
+                                'f%s' % f['format_id'], new_info['ext'])
+                            if not ensure_dir_exists(fname):
+                                return
                             downloaded.append(fname)
                             partial_success = dl(fname, new_info)
                             success = success and partial_success
@@ -1755,7 +1951,7 @@ class YoutubeDL(object):
                         info_dict.get('protocol') == 'm3u8' and
                         self.params.get('hls_prefer_native')):
                     if fixup_policy == 'warn':
-                        self.report_warning('%s: malformated aac bitstream.' % (
+                        self.report_warning('%s: malformed AAC bitstream detected.' % (
                             info_dict['id']))
                     elif fixup_policy == 'detect_or_warn':
                         fixup_pp = FFmpegFixupM3u8PP(self)
@@ -1764,7 +1960,7 @@ class YoutubeDL(object):
                             info_dict['__postprocessors'].append(fixup_pp)
                         else:
                             self.report_warning(
-                                '%s: malformated aac bitstream. %s'
+                                '%s: malformed AAC bitstream detected. %s'
                                 % (info_dict['id'], INSTALL_FFMPEG_MESSAGE))
                     else:
                         assert fixup_policy in ('ignore', 'never')
@@ -1780,6 +1976,7 @@ class YoutubeDL(object):
         """Download a given list of URLs."""
         outtmpl = self.params.get('outtmpl', DEFAULT_OUTTMPL)
         if (len(url_list) > 1 and
+                outtmpl != '-' and
                 '%' not in outtmpl and
                 self.params.get('max_downloads') != 1):
             raise SameFileError(outtmpl)
@@ -2036,11 +2233,20 @@ class YoutubeDL(object):
                 sys.exc_clear()
             except Exception:
                 pass
-        self._write_string('[debug] Python version %s - %s\n' % (
-            platform.python_version(), platform_name()))
+
+        def python_implementation():
+            impl_name = platform.python_implementation()
+            if impl_name == 'PyPy' and hasattr(sys, 'pypy_version_info'):
+                return impl_name + ' version %d.%d.%d' % sys.pypy_version_info[:3]
+            return impl_name
+
+        self._write_string('[debug] Python version %s (%s) - %s\n' % (
+            platform.python_version(), python_implementation(),
+            platform_name()))
 
         exe_versions = FFmpegPostProcessor.get_versions(self)
         exe_versions['rtmpdump'] = rtmpdump_version()
+        exe_versions['phantomjs'] = PhantomJSwrapper._version()
         exe_str = ', '.join(
             '%s %s' % (exe, v)
             for exe, v in sorted(exe_versions.items())
@@ -2077,7 +2283,7 @@ class YoutubeDL(object):
         if opts_cookiefile is None:
             self.cookiejar = compat_cookiejar.CookieJar()
         else:
-            opts_cookiefile = compat_expanduser(opts_cookiefile)
+            opts_cookiefile = expand_path(opts_cookiefile)
             self.cookiejar = compat_cookiejar.MozillaCookieJar(
                 opts_cookiefile)
             if os.access(opts_cookiefile, os.R_OK):

youtube_dl/__init__.py

@@ -16,7 +16,6 @@ from .options import (
     parseOpts,
 )
 from .compat import (
-    compat_expanduser,
     compat_getpass,
     compat_shlex_split,
     workaround_optparse_bug9161,
@@ -26,6 +25,7 @@ from .utils import (
     decodeOption,
     DEFAULT_OUTTMPL,
     DownloadError,
+    expand_path,
     match_filter_func,
     MaxDownloadsReached,
     preferredencoding,
@@ -88,7 +88,7 @@ def _real_main(argv=None):
                 batchfd = sys.stdin
             else:
                 batchfd = io.open(
-                    compat_expanduser(opts.batchfile),
+                    expand_path(opts.batchfile),
                     'r', encoding='utf-8', errors='ignore')
             batch_urls = read_batch_urls(batchfd)
             if opts.verbose:
@@ -133,6 +133,12 @@ def _real_main(argv=None):
         parser.error('TV Provider account username missing\n')
     if opts.outtmpl is not None and (opts.usetitle or opts.autonumber or opts.useid):
         parser.error('using output template conflicts with using title, video ID or auto number')
+    if opts.autonumber_size is not None:
+        if opts.autonumber_size <= 0:
+            parser.error('auto number size must be positive')
+    if opts.autonumber_start is not None:
+        if opts.autonumber_start < 0:
+            parser.error('auto number start must be positive or 0')
     if opts.usetitle and opts.useid:
         parser.error('using title conflicts with using video ID')
     if opts.username is not None and opts.password is None:
@@ -190,7 +196,7 @@ def _real_main(argv=None):
     if opts.playlistend not in (-1, None) and opts.playlistend < opts.playliststart:
         raise ValueError('Playlist end must be greater than playlist start')
     if opts.extractaudio:
-        if opts.audioformat not in ['best', 'aac', 'mp3', 'm4a', 'opus', 'vorbis', 'wav']:
+        if opts.audioformat not in ['best', 'aac', 'flac', 'mp3', 'm4a', 'opus', 'vorbis', 'wav']:
             parser.error('invalid audio format specified')
     if opts.audioquality:
         opts.audioquality = opts.audioquality.strip('k').strip('K')
@@ -200,7 +206,7 @@ def _real_main(argv=None):
         if opts.recodevideo not in ['mp4', 'flv', 'webm', 'ogg', 'mkv', 'avi']:
             parser.error('invalid video recode format specified')
     if opts.convertsubtitles is not None:
-        if opts.convertsubtitles not in ['srt', 'vtt', 'ass']:
+        if opts.convertsubtitles not in ['srt', 'vtt', 'ass', 'lrc']:
             parser.error('invalid subtitle format specified')
 
     if opts.date is not None:
@@ -232,18 +238,15 @@ def _real_main(argv=None):
 
     any_getting = opts.geturl or opts.gettitle or opts.getid or opts.getthumbnail or opts.getdescription or opts.getfilename or opts.getformat or opts.getduration or opts.dumpjson or opts.dump_single_json
     any_printing = opts.print_json
-    download_archive_fn = compat_expanduser(opts.download_archive) if opts.download_archive is not None else opts.download_archive
+    download_archive_fn = expand_path(opts.download_archive) if opts.download_archive is not None else opts.download_archive
 
     # PostProcessors
     postprocessors = []
-    # Add the metadata pp first, the other pps will copy it
     if opts.metafromtitle:
         postprocessors.append({
             'key': 'MetadataFromTitle',
             'titleformat': opts.metafromtitle
         })
-    if opts.addmetadata:
-        postprocessors.append({'key': 'FFmpegMetadata'})
     if opts.extractaudio:
         postprocessors.append({
             'key': 'FFmpegExtractAudio',
@@ -256,6 +259,16 @@ def _real_main(argv=None):
             'key': 'FFmpegVideoConvertor',
             'preferedformat': opts.recodevideo,
         })
+    # FFmpegMetadataPP should be run after FFmpegVideoConvertorPP and
+    # FFmpegExtractAudioPP as containers before conversion may not support
+    # metadata (3gp, webm, etc.)
+    # And this post-processor should be placed before other metadata
+    # manipulating post-processors (FFmpegEmbedSubtitle) to prevent loss of
+    # extra metadata. By default ffmpeg preserves metadata applicable for both
+    # source and target containers. From this point the container won't change,
+    # so metadata can be added here.
+    if opts.addmetadata:
+        postprocessors.append({'key': 'FFmpegMetadata'})
     if opts.convertsubtitles:
         postprocessors.append({
             'key': 'FFmpegSubtitlesConvertor',
@@ -321,6 +334,7 @@ def _real_main(argv=None):
         'listformats': opts.listformats,
         'outtmpl': outtmpl,
         'autonumber_size': opts.autonumber_size,
+        'autonumber_start': opts.autonumber_start,
         'restrictfilenames': opts.restrictfilenames,
         'ignoreerrors': opts.ignoreerrors,
         'force_generic_extractor': opts.force_generic_extractor,
@@ -329,6 +343,7 @@ def _real_main(argv=None):
         'retries': opts.retries,
         'fragment_retries': opts.fragment_retries,
         'skip_unavailable_fragments': opts.skip_unavailable_fragments,
+        'keep_fragments': opts.keep_fragments,
         'buffersize': opts.buffersize,
         'noresizebuffer': opts.noresizebuffer,
         'continuedl': opts.continue_dl,
@@ -337,6 +352,7 @@ def _real_main(argv=None):
         'playliststart': opts.playliststart,
         'playlistend': opts.playlistend,
         'playlistreverse': opts.playlist_reverse,
+        'playlistrandom': opts.playlist_random,
         'noplaylist': opts.noplaylist,
         'logtostderr': opts.outtmpl == '-',
         'consoletitle': opts.consoletitle,
@@ -406,6 +422,11 @@ def _real_main(argv=None):
         'cn_verification_proxy': opts.cn_verification_proxy,
         'geo_verification_proxy': opts.geo_verification_proxy,
         'config_location': opts.config_location,
+        'geo_bypass': opts.geo_bypass,
+        'geo_bypass_country': opts.geo_bypass_country,
+        # just for deprecation check
+        'autonumber': opts.autonumber if opts.autonumber is True else None,
+        'usetitle': opts.usetitle if opts.usetitle is True else None,
     }
 
     with YoutubeDL(ydl_opts) as ydl:
@@ -429,7 +450,7 @@ def _real_main(argv=None):
 
         try:
             if opts.load_info_filename is not None:
-                retcode = ydl.download_with_info_file(compat_expanduser(opts.load_info_filename))
+                retcode = ydl.download_with_info_file(expand_path(opts.load_info_filename))
             else:
                 retcode = ydl.download(all_urls)
         except MaxDownloadsReached:

youtube_dl/aes.py

@@ -60,6 +60,34 @@ def aes_cbc_decrypt(data, key, iv):
     return decrypted_data
 
 
+def aes_cbc_encrypt(data, key, iv):
+    """
+    Encrypt with aes in CBC mode. Using PKCS#7 padding
+
+    @param {int[]} data        cleartext
+    @param {int[]} key         16/24/32-Byte cipher key
+    @param {int[]} iv          16-Byte IV
+    @returns {int[]}           encrypted data
+    """
+    expanded_key = key_expansion(key)
+    block_count = int(ceil(float(len(data)) / BLOCK_SIZE_BYTES))
+
+    encrypted_data = []
+    previous_cipher_block = iv
+    for i in range(block_count):
+        block = data[i * BLOCK_SIZE_BYTES: (i + 1) * BLOCK_SIZE_BYTES]
+        remaining_length = BLOCK_SIZE_BYTES - len(block)
+        block += [remaining_length] * remaining_length
+        mixed_block = xor(block, previous_cipher_block)
+
+        encrypted_block = aes_encrypt(mixed_block, expanded_key)
+        encrypted_data += encrypted_block
+
+        previous_cipher_block = encrypted_block
+
+    return encrypted_data
+
+
 def key_expansion(data):
     """
     Generate key schedule

youtube_dl/cache.py

@@ -8,8 +8,11 @@ import re
 import shutil
 import traceback
 
-from .compat import compat_expanduser, compat_getenv
-from .utils import write_json_file
+from .compat import compat_getenv
+from .utils import (
+    expand_path,
+    write_json_file,
+)
 
 
 class Cache(object):
@@ -21,7 +24,7 @@ class Cache(object):
         if res is None:
             cache_root = compat_getenv('XDG_CACHE_HOME', '~/.cache')
             res = os.path.join(cache_root, 'youtube-dl')
-        return compat_expanduser(res)
+        return expand_path(res)
 
     def _get_cache_fn(self, section, key, dtype):
         assert re.match(r'^[a-zA-Z0-9_.-]+$', section), \

youtube_dl/compat.py

@@ -3,11 +3,14 @@ from __future__ import unicode_literals
 
 import binascii
 import collections
+import ctypes
 import email
 import getpass
 import io
+import itertools
 import optparse
 import os
+import platform
 import re
 import shlex
 import shutil
@@ -15,7 +18,6 @@ import socket
 import struct
 import subprocess
 import sys
-import itertools
 import xml.etree.ElementTree
 
 
@@ -2322,6 +2324,19 @@ try:
 except ImportError:  # Python 2
     from HTMLParser import HTMLParser as compat_HTMLParser
 
+try:  # Python 2
+    from HTMLParser import HTMLParseError as compat_HTMLParseError
+except ImportError:  # Python <3.4
+    try:
+        from html.parser import HTMLParseError as compat_HTMLParseError
+    except ImportError:  # Python >3.4
+
+        # HTMLParseError has been deprecated in Python 3.3 and removed in
+        # Python 3.5. Introducing dummy exception for Python >3.5 for compatible
+        # and uniform cross-version exceptiong handling
+        class compat_HTMLParseError(Exception):
+            pass
+
 try:
     from subprocess import DEVNULL
     compat_subprocess_get_DEVNULL = lambda: DEVNULL
@@ -2529,6 +2544,24 @@ else:
                 el.text = el.text.decode('utf-8')
         return doc
 
+if hasattr(etree, 'register_namespace'):
+    compat_etree_register_namespace = etree.register_namespace
+else:
+    def compat_etree_register_namespace(prefix, uri):
+        """Register a namespace prefix.
+        The registry is global, and any existing mapping for either the
+        given prefix or the namespace URI will be removed.
+        *prefix* is the namespace prefix, *uri* is a namespace uri. Tags and
+        attributes in this namespace will be serialized with prefix if possible.
+        ValueError is raised if prefix is reserved or is invalid.
+        """
+        if re.match(r"ns\d+$", prefix):
+            raise ValueError("Prefix format reserved for internal use")
+        for k, v in list(etree._namespace_map.items()):
+            if k == uri or v == prefix:
+                del etree._namespace_map[k]
+        etree._namespace_map[uri] = prefix
+
 if sys.version_info < (2, 7):
     # Here comes the crazy part: In 2.6, if the xpath is a unicode,
     # .//node does not match if a node is a direct child of . !
@@ -2586,14 +2619,22 @@ except ImportError:  # Python 2
                 parsed_result[name] = [value]
         return parsed_result
 
-try:
-    from shlex import quote as compat_shlex_quote
-except ImportError:  # Python < 3.3
+
+compat_os_name = os._name if os.name == 'java' else os.name
+
+
+if compat_os_name == 'nt':
     def compat_shlex_quote(s):
-        if re.match(r'^[-_\w./]+$', s):
-            return s
-        else:
-            return "'" + s.replace("'", "'\"'\"'") + "'"
+        return s if re.match(r'^[-_\w./]+$', s) else '"%s"' % s.replace('"', '\\"')
+else:
+    try:
+        from shlex import quote as compat_shlex_quote
+    except ImportError:  # Python < 3.3
+        def compat_shlex_quote(s):
+            if re.match(r'^[-_\w./]+$', s):
+                return s
+            else:
+                return "'" + s.replace("'", "'\"'\"'") + "'"
 
 
 try:
@@ -2618,9 +2659,6 @@ def compat_ord(c):
         return ord(c)
 
 
-compat_os_name = os._name if os.name == 'java' else os.name
-
-
 if sys.version_info >= (3, 0):
     compat_getenv = os.getenv
     compat_expanduser = os.path.expanduser
@@ -2674,7 +2712,7 @@ else:
                 userhome = pwent.pw_dir
             userhome = userhome.rstrip('/')
             return (userhome + path[i:]) or '/'
-    elif compat_os_name == 'nt' or compat_os_name == 'ce':
+    elif compat_os_name in ('nt', 'ce'):
         def compat_expanduser(path):
             """Expand ~ and ~user constructs.
 
@@ -2742,6 +2780,12 @@ else:
     compat_kwargs = lambda kwargs: kwargs
 
 
+try:
+    compat_numeric_types = (int, float, long, complex)
+except NameError:  # Python 3
+    compat_numeric_types = (int, float, complex)
+
+
 if sys.version_info < (2, 7):
     def compat_socket_create_connection(address, timeout, source_address=None):
         host, port = address
@@ -2856,15 +2900,43 @@ else:
     compat_struct_pack = struct.pack
     compat_struct_unpack = struct.unpack
 
+try:
+    from future_builtins import zip as compat_zip
+except ImportError:  # not 2.6+ or is 3.x
+    try:
+        from itertools import izip as compat_zip  # < 2.5 or 3.x
+    except ImportError:
+        compat_zip = zip
+
+if platform.python_implementation() == 'PyPy' and sys.pypy_version_info < (5, 4, 0):
+    # PyPy2 prior to version 5.4.0 expects byte strings as Windows function
+    # names, see the original PyPy issue [1] and the youtube-dl one [2].
+    # 1. https://bitbucket.org/pypy/pypy/issues/2360/windows-ctypescdll-typeerror-function-name
+    # 2. https://github.com/rg3/youtube-dl/pull/4392
+    def compat_ctypes_WINFUNCTYPE(*args, **kwargs):
+        real = ctypes.WINFUNCTYPE(*args, **kwargs)
+
+        def resf(tpl, *args, **kwargs):
+            funcname, dll = tpl
+            return real((str(funcname), dll), *args, **kwargs)
+
+        return resf
+else:
+    def compat_ctypes_WINFUNCTYPE(*args, **kwargs):
+        return ctypes.WINFUNCTYPE(*args, **kwargs)
+
 
 __all__ = [
+    'compat_HTMLParseError',
     'compat_HTMLParser',
     'compat_HTTPError',
     'compat_basestring',
     'compat_chr',
     'compat_cookiejar',
     'compat_cookies',
+    'compat_ctypes_WINFUNCTYPE',
     'compat_etree_fromstring',
+    'compat_etree_register_namespace',
     'compat_expanduser',
     'compat_get_terminal_size',
     'compat_getenv',
@@ -2876,6 +2948,7 @@ __all__ = [
     'compat_input',
     'compat_itertools_count',
     'compat_kwargs',
+    'compat_numeric_types',
     'compat_ord',
     'compat_os_name',
     'compat_parse_qs',
@@ -2903,5 +2976,6 @@ __all__ = [
     'compat_urlretrieve',
     'compat_xml_parse_error',
     'compat_xpath',
+    'compat_zip',
     'workaround_optparse_bug9161',
 ]

youtube_dl/downloader/__init__.py

@@ -43,6 +43,9 @@ def get_suitable_downloader(info_dict, params={}):
         if ed.can_download(info_dict):
             return ed
 
+    if protocol.startswith('m3u8') and info_dict.get('is_live'):
+        return FFmpegFD
+
     if protocol == 'm3u8' and params.get('hls_prefer_native') is True:
         return HlsFD
 

youtube_dl/downloader/common.py

@@ -8,10 +8,11 @@ import random
 
 from ..compat import compat_os_name
 from ..utils import (
+    decodeArgument,
     encodeFilename,
     error_to_compat_str,
-    decodeArgument,
     format_bytes,
+    shell_quote,
     timeconvert,
 )
 
@@ -187,6 +188,9 @@ class FileDownloader(object):
             return filename[:-len('.part')]
         return filename
 
+    def ytdl_filename(self, filename):
+        return filename + '.ytdl'
+
     def try_rename(self, old_filename, new_filename):
         try:
             if old_filename == new_filename:
@@ -300,11 +304,11 @@ class FileDownloader(object):
         """Report attempt to resume at given byte."""
         self.to_screen('[download] Resuming download at byte %s' % resume_len)
 
-    def report_retry(self, count, retries):
+    def report_retry(self, err, count, retries):
         """Report retry in case of HTTP error 5xx"""
         self.to_screen(
-            '[download] Got server HTTP error. Retrying (attempt %d of %s)...'
-            % (count, self.format_retries(retries)))
+            '[download] Got server HTTP error: %s. Retrying (attempt %d of %s)...'
+            % (error_to_compat_str(err), count, self.format_retries(retries)))
 
     def report_file_already_downloaded(self, file_name):
         """Report file has already been fully downloaded."""
@@ -327,27 +331,31 @@ class FileDownloader(object):
             os.path.exists(encodeFilename(filename))
         )
 
-        continuedl_and_exists = (
-            self.params.get('continuedl', True) and
-            os.path.isfile(encodeFilename(filename)) and
-            not self.params.get('nopart', False)
-        )
+        if not hasattr(filename, 'write'):
+            continuedl_and_exists = (
+                self.params.get('continuedl', True) and
+                os.path.isfile(encodeFilename(filename)) and
+                not self.params.get('nopart', False)
+            )
 
-        # Check file already present
-        if filename != '-' and (nooverwrites_and_exists or continuedl_and_exists):
-            self.report_file_already_downloaded(filename)
-            self._hook_progress({
-                'filename': filename,
-                'status': 'finished',
-                'total_bytes': os.path.getsize(encodeFilename(filename)),
-            })
-            return True
+            # Check file already present
+            if filename != '-' and (nooverwrites_and_exists or continuedl_and_exists):
+                self.report_file_already_downloaded(filename)
+                self._hook_progress({
+                    'filename': filename,
+                    'status': 'finished',
+                    'total_bytes': os.path.getsize(encodeFilename(filename)),
+                })
+                return True
 
         min_sleep_interval = self.params.get('sleep_interval')
         if min_sleep_interval:
             max_sleep_interval = self.params.get('max_sleep_interval', min_sleep_interval)
             sleep_interval = random.uniform(min_sleep_interval, max_sleep_interval)
-            self.to_screen('[download] Sleeping %s seconds...' % sleep_interval)
+            self.to_screen(
+                '[download] Sleeping %s seconds...' % (
+                    int(sleep_interval) if sleep_interval.is_integer()
+                    else '%.2f' % sleep_interval))
             time.sleep(sleep_interval)
 
         return self.real_download(filename, info_dict)
@@ -374,10 +382,5 @@ class FileDownloader(object):
         if exe is None:
             exe = os.path.basename(str_args[0])
 
-        try:
-            import pipes
-            shell_quote = lambda args: ' '.join(map(pipes.quote, str_args))
-        except ImportError:
-            shell_quote = repr
         self.to_screen('[debug] %s command line: %s' % (
             exe, shell_quote(str_args)))

youtube_dl/downloader/dash.py

@@ -1,13 +1,8 @@
 from __future__ import unicode_literals
 
-import os
-
 from .fragment import FragmentFD
 from ..compat import compat_urllib_error
-from ..utils import (
-    sanitize_open,
-    encodeFilename,
-)
+from ..utils import urljoin
 
 
 class DashSegmentsFD(FragmentFD):
@@ -18,38 +13,39 @@ class DashSegmentsFD(FragmentFD):
     FD_NAME = 'dashsegments'
 
     def real_download(self, filename, info_dict):
-        segments = info_dict['fragments'][:1] if self.params.get(
+        fragment_base_url = info_dict.get('fragment_base_url')
+        fragments = info_dict['fragments'][:1] if self.params.get(
             'test', False) else info_dict['fragments']
 
         ctx = {
             'filename': filename,
-            'total_frags': len(segments),
+            'total_frags': len(fragments),
         }
 
         self._prepare_and_start_frag_download(ctx)
 
-        segments_filenames = []
-
         fragment_retries = self.params.get('fragment_retries', 0)
         skip_unavailable_fragments = self.params.get('skip_unavailable_fragments', True)
 
-        def process_segment(segment, tmp_filename, num):
-            segment_url = segment['url']
-            segment_name = 'Frag%d' % num
-            target_filename = '%s-%s' % (tmp_filename, segment_name)
+        frag_index = 0
+        for i, fragment in enumerate(fragments):
+            frag_index += 1
+            if frag_index <= ctx['fragment_index']:
+                continue
             # In DASH, the first segment contains necessary headers to
             # generate a valid MP4 file, so always abort for the first segment
-            fatal = num == 0 or not skip_unavailable_fragments
+            fatal = i == 0 or not skip_unavailable_fragments
             count = 0
             while count <= fragment_retries:
                 try:
-                    success = ctx['dl'].download(target_filename, {'url': segment_url})
+                    fragment_url = fragment.get('url')
+                    if not fragment_url:
+                        assert fragment_base_url
+                        fragment_url = urljoin(fragment_base_url, fragment['path'])
+                    success, frag_content = self._download_fragment(ctx, fragment_url, info_dict)
                     if not success:
                         return False
-                    down, target_sanitized = sanitize_open(target_filename, 'rb')
-                    ctx['dest_stream'].write(down.read())
-                    down.close()
-                    segments_filenames.append(target_sanitized)
+                    self._append_fragment(ctx, frag_content)
                     break
                 except compat_urllib_error.HTTPError as err:
                     # YouTube may often return 404 HTTP error for a fragment causing the
@@ -60,22 +56,14 @@ class DashSegmentsFD(FragmentFD):
                     # HTTP error.
                     count += 1
                     if count <= fragment_retries:
-                        self.report_retry_fragment(err, segment_name, count, fragment_retries)
+                        self.report_retry_fragment(err, frag_index, count, fragment_retries)
             if count > fragment_retries:
                 if not fatal:
-                    self.report_skip_fragment(segment_name)
-                    return True
+                    self.report_skip_fragment(frag_index)
+                    continue
                 self.report_error('giving up after %s fragment retries' % fragment_retries)
                 return False
-            return True
-
-        for i, segment in enumerate(segments):
-            if not process_segment(segment, ctx['tmpfilename'], i):
-                return False
 
         self._finish_frag_download(ctx)
 
-        for segment_file in segments_filenames:
-            os.remove(encodeFilename(segment_file))
-
         return True

youtube_dl/downloader/external.py

@@ -6,7 +6,10 @@ import sys
 import re
 
 from .common import FileDownloader
-from ..compat import compat_setenv
+from ..compat import (
+    compat_setenv,
+    compat_str,
+)
 from ..postprocessor.ffmpeg import FFmpegPostProcessor, EXT_TO_OUT_FORMATS
 from ..utils import (
     cli_option,
@@ -17,6 +20,7 @@ from ..utils import (
     encodeArgument,
     handle_youtubedl_headers,
     check_executable,
+    is_outdated_version,
 )
 
 
@@ -25,7 +29,17 @@ class ExternalFD(FileDownloader):
         self.report_destination(filename)
         tmpfilename = self.temp_name(filename)
 
-        retval = self._call_downloader(tmpfilename, info_dict)
+        try:
+            retval = self._call_downloader(tmpfilename, info_dict)
+        except KeyboardInterrupt:
+            if not info_dict.get('is_live'):
+                raise
+            # Live stream downloading cancellation should be considered as
+            # correct and expected termination thus all postprocessing
+            # should take place
+            retval = 0
+            self.to_screen('[%s] Interrupted by user' % self.get_basename())
+
         if retval == 0:
             fsize = os.path.getsize(encodeFilename(tmpfilename))
             self.to_screen('\r[%s] Downloaded %s bytes' % (self.get_basename(), fsize))
@@ -198,6 +212,20 @@ class FFmpegFD(ExternalFD):
 
         args = [ffpp.executable, '-y']
 
+        for log_level in ('quiet', 'verbose'):
+            if self.params.get(log_level, False):
+                args += ['-loglevel', log_level]
+                break
+
+        seekable = info_dict.get('_seekable')
+        if seekable is not None:
+            # setting -seekable prevents ffmpeg from guessing if the server
+            # supports seeking(by adding the header `Range: bytes=0-`), which
+            # can cause problems in some cases
+            # https://github.com/rg3/youtube-dl/issues/11800#issuecomment-275037127
+            # http://trac.ffmpeg.org/ticket/6125#comment:10
+            args += ['-seekable', '1' if seekable else '0']
+
         args += self._configuration_args()
 
         # start_time = info_dict.get('start_time') or 0
@@ -260,11 +288,17 @@ class FFmpegFD(ExternalFD):
                 args += ['-rtmp_live', 'live']
 
         args += ['-i', url, '-c', 'copy']
+
+        if self.params.get('test', False):
+            args += ['-fs', compat_str(self._TEST_FILE_SIZE)]
+
         if protocol in ('m3u8', 'm3u8_native'):
             if self.params.get('hls_use_mpegts', False) or tmpfilename == '-':
                 args += ['-f', 'mpegts']
             else:
-                args += ['-f', 'mp4', '-bsf:a', 'aac_adtstoasc']
+                args += ['-f', 'mp4']
+                if (ffpp.basename == 'ffmpeg' and is_outdated_version(ffpp._versions['ffmpeg'], '3.2', False)) and (not info_dict.get('acodec') or info_dict['acodec'].split('.')[0] in ('aac', 'mp4a')):
+                    args += ['-bsf:a', 'aac_adtstoasc']
         elif protocol == 'rtmp':
             args += ['-f', 'flv']
         else:

youtube_dl/downloader/f4m.py

@@ -3,7 +3,6 @@ from __future__ import division, unicode_literals
 import base64
 import io
 import itertools
-import os
 import time
 
 from .fragment import FragmentFD
@@ -16,9 +15,7 @@ from ..compat import (
     compat_struct_unpack,
 )
 from ..utils import (
-    encodeFilename,
     fix_xml_ampersands,
-    sanitize_open,
     xpath_text,
 )
 
@@ -246,8 +243,17 @@ def remove_encrypted_media(media):
                        media))
 
 
-def _add_ns(prop):
-    return '{http://ns.adobe.com/f4m/1.0}%s' % prop
+def _add_ns(prop, ver=1):
+    return '{http://ns.adobe.com/f4m/%d.0}%s' % (ver, prop)
+
+
+def get_base_url(manifest):
+    base_url = xpath_text(
+        manifest, [_add_ns('baseURL'), _add_ns('baseURL', 2)],
+        'base URL', default=None)
+    if base_url:
+        base_url = base_url.strip()
+    return base_url
 
 
 class F4mFD(FragmentFD):
@@ -333,13 +339,13 @@ class F4mFD(FragmentFD):
             rate, media = list(filter(
                 lambda f: int(f[0]) == requested_bitrate, formats))[0]
 
-        base_url = compat_urlparse.urljoin(man_url, media.attrib['url'])
+        # Prefer baseURL for relative URLs as per 11.2 of F4M 3.0 spec.
+        man_base_url = get_base_url(doc) or man_url
+
+        base_url = compat_urlparse.urljoin(man_base_url, media.attrib['url'])
         bootstrap_node = doc.find(_add_ns('bootstrapInfo'))
-        # From Adobe F4M 3.0 spec:
-        # The <baseURL> element SHALL be the base URL for all relative
-        # (HTTP-based) URLs in the manifest. If <baseURL> is not present, said
-        # URLs should be relative to the location of the containing document.
-        boot_info, bootstrap_url = self._parse_bootstrap_node(bootstrap_node, man_url)
+        boot_info, bootstrap_url = self._parse_bootstrap_node(
+            bootstrap_node, man_base_url)
         live = boot_info['live']
         metadata_node = media.find(_add_ns('metadata'))
         if metadata_node is not None:
@@ -366,17 +372,21 @@ class F4mFD(FragmentFD):
 
         dest_stream = ctx['dest_stream']
 
-        write_flv_header(dest_stream)
-        if not live:
-            write_metadata_tag(dest_stream, metadata)
+        if ctx['complete_frags_downloaded_bytes'] == 0:
+            write_flv_header(dest_stream)
+            if not live:
+                write_metadata_tag(dest_stream, metadata)
 
         base_url_parsed = compat_urllib_parse_urlparse(base_url)
 
         self._start_frag_download(ctx)
 
-        frags_filenames = []
+        frag_index = 0
         while fragments_list:
             seg_i, frag_i = fragments_list.pop(0)
+            frag_index += 1
+            if frag_index <= ctx['fragment_index']:
+                continue
             name = 'Seg%d-Frag%d' % (seg_i, frag_i)
             query = []
             if base_url_parsed.query:
@@ -386,17 +396,10 @@ class F4mFD(FragmentFD):
             if info_dict.get('extra_param_to_segment_url'):
                 query.append(info_dict['extra_param_to_segment_url'])
             url_parsed = base_url_parsed._replace(path=base_url_parsed.path + name, query='&'.join(query))
-            frag_filename = '%s-%s' % (ctx['tmpfilename'], name)
             try:
-                success = ctx['dl'].download(frag_filename, {
-                    'url': url_parsed.geturl(),
-                    'http_headers': info_dict.get('http_headers'),
-                })
+                success, down_data = self._download_fragment(ctx, url_parsed.geturl(), info_dict)
                 if not success:
                     return False
-                (down, frag_sanitized) = sanitize_open(frag_filename, 'rb')
-                down_data = down.read()
-                down.close()
                 reader = FlvReader(down_data)
                 while True:
                     try:
@@ -411,12 +414,8 @@ class F4mFD(FragmentFD):
                             break
                         raise
                     if box_type == b'mdat':
-                        dest_stream.write(box_data)
+                        self._append_fragment(ctx, box_data)
                         break
-                if live:
-                    os.remove(encodeFilename(frag_sanitized))
-                else:
-                    frags_filenames.append(frag_sanitized)
             except (compat_urllib_error.HTTPError, ) as err:
                 if live and (err.code == 404 or err.code == 410):
                     # We didn't keep up with the live window. Continue
@@ -436,7 +435,4 @@ class F4mFD(FragmentFD):
 
         self._finish_frag_download(ctx)
 
-        for frag_file in frags_filenames:
-            os.remove(encodeFilename(frag_file))
-
         return True

youtube_dl/downloader/fragment.py

@@ -2,6 +2,7 @@ from __future__ import division, unicode_literals
 
 import os
 import time
+import json
 
 from .common import FileDownloader
 from .http import HttpFD
@@ -28,15 +29,37 @@ class FragmentFD(FileDownloader):
                         and hlsnative only)
     skip_unavailable_fragments:
                         Skip unavailable fragments (DASH and hlsnative only)
+    keep_fragments:     Keep downloaded fragments on disk after downloading is
+                        finished
+
+    For each incomplete fragment download youtube-dl keeps on disk a special
+    bookkeeping file with download state and metadata (in future such files will
+    be used for any incomplete download handled by youtube-dl). This file is
+    used to properly handle resuming, check download file consistency and detect
+    potential errors. The file has a .ytdl extension and represents a standard
+    JSON file of the following format:
+
+    extractor:
+        Dictionary of extractor related data. TBD.
+
+    downloader:
+        Dictionary of downloader related data. May contain following data:
+            current_fragment:
+                Dictionary with current (being downloaded) fragment data:
+                index:  0-based index of current fragment among all fragments
+            fragment_count:
+                Total count of fragments
+
+    This feature is experimental and file format may change in future.
     """
 
-    def report_retry_fragment(self, err, fragment_name, count, retries):
+    def report_retry_fragment(self, err, frag_index, count, retries):
         self.to_screen(
-            '[download] Got server HTTP error: %s. Retrying fragment %s (attempt %d of %s)...'
-            % (error_to_compat_str(err), fragment_name, count, self.format_retries(retries)))
+            '[download] Got server HTTP error: %s. Retrying fragment %d (attempt %d of %s)...'
+            % (error_to_compat_str(err), frag_index, count, self.format_retries(retries)))
 
-    def report_skip_fragment(self, fragment_name):
-        self.to_screen('[download] Skipping fragment %s...' % fragment_name)
+    def report_skip_fragment(self, frag_index):
+        self.to_screen('[download] Skipping fragment %d...' % frag_index)
 
     def _prepare_url(self, info_dict, url):
         headers = info_dict.get('http_headers')
@@ -46,12 +69,64 @@ class FragmentFD(FileDownloader):
         self._prepare_frag_download(ctx)
         self._start_frag_download(ctx)
 
+    @staticmethod
+    def __do_ytdl_file(ctx):
+        return not ctx['live'] and not ctx['tmpfilename'] == '-'
+
+    def _read_ytdl_file(self, ctx):
+        stream, _ = sanitize_open(self.ytdl_filename(ctx['filename']), 'r')
+        ctx['fragment_index'] = json.loads(stream.read())['downloader']['current_fragment']['index']
+        stream.close()
+
+    def _write_ytdl_file(self, ctx):
+        frag_index_stream, _ = sanitize_open(self.ytdl_filename(ctx['filename']), 'w')
+        downloader = {
+            'current_fragment': {
+                'index': ctx['fragment_index'],
+            },
+        }
+        if ctx.get('fragment_count') is not None:
+            downloader['fragment_count'] = ctx['fragment_count']
+        frag_index_stream.write(json.dumps({'downloader': downloader}))
+        frag_index_stream.close()
+
+    def _download_fragment(self, ctx, frag_url, info_dict, headers=None):
+        fragment_filename = '%s-Frag%d' % (ctx['tmpfilename'], ctx['fragment_index'])
+        success = ctx['dl'].download(fragment_filename, {
+            'url': frag_url,
+            'http_headers': headers or info_dict.get('http_headers'),
+        })
+        if not success:
+            return False, None
+        down, frag_sanitized = sanitize_open(fragment_filename, 'rb')
+        ctx['fragment_filename_sanitized'] = frag_sanitized
+        frag_content = down.read()
+        down.close()
+        return True, frag_content
+
+    def _append_fragment(self, ctx, frag_content):
+        try:
+            ctx['dest_stream'].write(frag_content)
+            ctx['dest_stream'].flush()
+        finally:
+            if self.__do_ytdl_file(ctx):
+                self._write_ytdl_file(ctx)
+            if not self.params.get('keep_fragments', False):
+                os.remove(encodeFilename(ctx['fragment_filename_sanitized']))
+            del ctx['fragment_filename_sanitized']
+
     def _prepare_frag_download(self, ctx):
         if 'live' not in ctx:
             ctx['live'] = False
+        if not ctx['live']:
+            total_frags_str = '%d' % ctx['total_frags']
+            ad_frags = ctx.get('ad_frags', 0)
+            if ad_frags:
+                total_frags_str += ' (not including %d ad)' % ad_frags
+        else:
+            total_frags_str = 'unknown (live)'
         self.to_screen(
-            '[%s] Total fragments: %s'
-            % (self.FD_NAME, ctx['total_frags'] if not ctx['live'] else 'unknown (live)'))
+            '[%s] Total fragments: %s' % (self.FD_NAME, total_frags_str))
         self.report_destination(ctx['filename'])
         dl = HttpQuietDownloader(
             self.ydl,
@@ -61,15 +136,46 @@ class FragmentFD(FileDownloader):
                 'noprogress': True,
                 'ratelimit': self.params.get('ratelimit'),
                 'retries': self.params.get('retries', 0),
+                'nopart': self.params.get('nopart', False),
                 'test': self.params.get('test', False),
             }
         )
         tmpfilename = self.temp_name(ctx['filename'])
-        dest_stream, tmpfilename = sanitize_open(tmpfilename, 'wb')
+        open_mode = 'wb'
+        resume_len = 0
+
+        # Establish possible resume length
+        if os.path.isfile(encodeFilename(tmpfilename)):
+            open_mode = 'ab'
+            resume_len = os.path.getsize(encodeFilename(tmpfilename))
+
+        # Should be initialized before ytdl file check
+        ctx.update({
+            'tmpfilename': tmpfilename,
+            'fragment_index': 0,
+        })
+
+        if self.__do_ytdl_file(ctx):
+            if os.path.isfile(encodeFilename(self.ytdl_filename(ctx['filename']))):
+                self._read_ytdl_file(ctx)
+                if ctx['fragment_index'] > 0 and resume_len == 0:
+                    self.report_warning(
+                        'Inconsistent state of incomplete fragment download. '
+                        'Restarting from the beginning...')
+                    ctx['fragment_index'] = resume_len = 0
+                    self._write_ytdl_file(ctx)
+            else:
+                self._write_ytdl_file(ctx)
+                assert ctx['fragment_index'] == 0
+
+        dest_stream, tmpfilename = sanitize_open(tmpfilename, open_mode)
+
         ctx.update({
             'dl': dl,
             'dest_stream': dest_stream,
             'tmpfilename': tmpfilename,
+            # Total complete fragments downloaded so far in bytes
+            'complete_frags_downloaded_bytes': resume_len,
         })
 
     def _start_frag_download(self, ctx):
@@ -78,9 +184,9 @@ class FragmentFD(FileDownloader):
         # hook
         state = {
             'status': 'downloading',
-            'downloaded_bytes': 0,
-            'frag_index': 0,
-            'frag_count': total_frags,
+            'downloaded_bytes': ctx['complete_frags_downloaded_bytes'],
+            'fragment_index': ctx['fragment_index'],
+            'fragment_count': total_frags,
             'filename': ctx['filename'],
             'tmpfilename': ctx['tmpfilename'],
         }
@@ -88,8 +194,6 @@ class FragmentFD(FileDownloader):
         start = time.time()
         ctx.update({
             'started': start,
-            # Total complete fragments downloaded so far in bytes
-            'complete_frags_downloaded_bytes': 0,
             # Amount of fragment's bytes downloaded by the time of the previous
             # frag progress hook invocation
             'prev_frag_downloaded_bytes': 0,
@@ -105,11 +209,12 @@ class FragmentFD(FileDownloader):
             if not ctx['live']:
                 estimated_size = (
                     (ctx['complete_frags_downloaded_bytes'] + frag_total_bytes) /
-                    (state['frag_index'] + 1) * total_frags)
+                    (state['fragment_index'] + 1) * total_frags)
                 state['total_bytes_estimate'] = estimated_size
 
             if s['status'] == 'finished':
-                state['frag_index'] += 1
+                state['fragment_index'] += 1
+                ctx['fragment_index'] = state['fragment_index']
                 state['downloaded_bytes'] += frag_total_bytes - ctx['prev_frag_downloaded_bytes']
                 ctx['complete_frags_downloaded_bytes'] = state['downloaded_bytes']
                 ctx['prev_frag_downloaded_bytes'] = 0
@@ -131,6 +236,10 @@ class FragmentFD(FileDownloader):
 
     def _finish_frag_download(self, ctx):
         ctx['dest_stream'].close()
+        if self.__do_ytdl_file(ctx):
+            ytdl_filename = encodeFilename(self.ytdl_filename(ctx['filename']))
+            if os.path.isfile(ytdl_filename):
+                os.remove(ytdl_filename)
         elapsed = time.time() - ctx['started']
         self.try_rename(ctx['tmpfilename'], ctx['filename'])
         fsize = os.path.getsize(encodeFilename(ctx['filename']))

youtube_dl/downloader/hls.py

@@ -1,6 +1,5 @@
 from __future__ import unicode_literals
 
-import os.path
 import re
 import binascii
 try:
@@ -18,8 +17,6 @@ from ..compat import (
     compat_struct_pack,
 )
 from ..utils import (
-    encodeFilename,
-    sanitize_open,
     parse_m3u8_attributes,
     update_url_query,
 )
@@ -34,7 +31,7 @@ class HlsFD(FragmentFD):
     def can_download(manifest, info_dict):
         UNSUPPORTED_FEATURES = (
             r'#EXT-X-KEY:METHOD=(?!NONE|AES-128)',  # encrypted streams [1]
-            r'#EXT-X-BYTERANGE',  # playlists composed of byte ranges of media files [2]
+            # r'#EXT-X-BYTERANGE',  # playlists composed of byte ranges of media files [2]
 
             # Live streams heuristic does not always work (e.g. geo restricted to Germany
             # http://hls-geo.daserste.de/i/videoportal/Film/c_620000/622873/format,716451,716457,716450,716458,716459,.mp4.csmil/index_4_av.m3u8?null=0)
@@ -52,7 +49,9 @@ class HlsFD(FragmentFD):
             # 4. https://tools.ietf.org/html/draft-pantos-http-live-streaming-17#section-4.3.3.5
         )
         check_results = [not re.search(feature, manifest) for feature in UNSUPPORTED_FEATURES]
-        check_results.append(can_decrypt_frag or '#EXT-X-KEY:METHOD=AES-128' not in manifest)
+        is_aes128_enc = '#EXT-X-KEY:METHOD=AES-128' in manifest
+        check_results.append(can_decrypt_frag or not is_aes128_enc)
+        check_results.append(not (is_aes128_enc and r'#EXT-X-BYTERANGE' in manifest))
         check_results.append(not info_dict.get('is_live'))
         return all(check_results)
 
@@ -60,9 +59,9 @@ class HlsFD(FragmentFD):
         man_url = info_dict['url']
         self.to_screen('[%s] Downloading m3u8 manifest' % self.FD_NAME)
 
-        manifest = self.ydl.urlopen(self._prepare_url(info_dict, man_url)).read()
-
-        s = manifest.decode('utf-8', 'ignore')
+        urlh = self.ydl.urlopen(self._prepare_url(info_dict, man_url))
+        man_url = urlh.geturl()
+        s = urlh.read().decode('utf-8', 'ignore')
 
         if not self.can_download(s, info_dict):
             if info_dict.get('extra_param_to_segment_url'):
@@ -76,15 +75,30 @@ class HlsFD(FragmentFD):
                 fd.add_progress_hook(ph)
             return fd.real_download(filename, info_dict)
 
-        total_frags = 0
+        def anvato_ad(s):
+            return s.startswith('#ANVATO-SEGMENT-INFO') and 'type=ad' in s
+
+        media_frags = 0
+        ad_frags = 0
+        ad_frag_next = False
         for line in s.splitlines():
             line = line.strip()
-            if line and not line.startswith('#'):
-                total_frags += 1
+            if not line:
+                continue
+            if line.startswith('#'):
+                if anvato_ad(line):
+                    ad_frags += 1
+                    ad_frag_next = True
+                continue
+            if ad_frag_next:
+                ad_frag_next = False
+                continue
+            media_frags += 1
 
         ctx = {
             'filename': filename,
-            'total_frags': total_frags,
+            'total_frags': media_frags,
+            'ad_frags': ad_frags,
         }
 
         self._prepare_and_start_frag_download(ctx)
@@ -100,31 +114,35 @@ class HlsFD(FragmentFD):
         i = 0
         media_sequence = 0
         decrypt_info = {'METHOD': 'NONE'}
-        frags_filenames = []
+        byte_range = {}
+        frag_index = 0
+        ad_frag_next = False
         for line in s.splitlines():
             line = line.strip()
             if line:
                 if not line.startswith('#'):
+                    if ad_frag_next:
+                        ad_frag_next = False
+                        continue
+                    frag_index += 1
+                    if frag_index <= ctx['fragment_index']:
+                        continue
                     frag_url = (
                         line
                         if re.match(r'^https?://', line)
                         else compat_urlparse.urljoin(man_url, line))
-                    frag_name = 'Frag%d' % i
-                    frag_filename = '%s-%s' % (ctx['tmpfilename'], frag_name)
                     if extra_query:
                         frag_url = update_url_query(frag_url, extra_query)
                     count = 0
+                    headers = info_dict.get('http_headers', {})
+                    if byte_range:
+                        headers['Range'] = 'bytes=%d-%d' % (byte_range['start'], byte_range['end'])
                     while count <= fragment_retries:
                         try:
-                            success = ctx['dl'].download(frag_filename, {
-                                'url': frag_url,
-                                'http_headers': info_dict.get('http_headers'),
-                            })
+                            success, frag_content = self._download_fragment(
+                                ctx, frag_url, info_dict, headers)
                             if not success:
                                 return False
-                            down, frag_sanitized = sanitize_open(frag_filename, 'rb')
-                            frag_content = down.read()
-                            down.close()
                             break
                         except compat_urllib_error.HTTPError as err:
                             # Unavailable (possibly temporary) fragments may be served.
@@ -133,28 +151,30 @@ class HlsFD(FragmentFD):
                             # https://github.com/rg3/youtube-dl/issues/10448).
                             count += 1
                             if count <= fragment_retries:
-                                self.report_retry_fragment(err, frag_name, count, fragment_retries)
+                                self.report_retry_fragment(err, frag_index, count, fragment_retries)
                     if count > fragment_retries:
                         if skip_unavailable_fragments:
                             i += 1
                             media_sequence += 1
-                            self.report_skip_fragment(frag_name)
+                            self.report_skip_fragment(frag_index)
                             continue
                         self.report_error(
                             'giving up after %s fragment retries' % fragment_retries)
                         return False
                     if decrypt_info['METHOD'] == 'AES-128':
                         iv = decrypt_info.get('IV') or compat_struct_pack('>8xq', media_sequence)
+                        decrypt_info['KEY'] = decrypt_info.get('KEY') or self.ydl.urlopen(
+                            self._prepare_url(info_dict, decrypt_info['URI'])).read()
                         frag_content = AES.new(
                             decrypt_info['KEY'], AES.MODE_CBC, iv).decrypt(frag_content)
-                    ctx['dest_stream'].write(frag_content)
-                    frags_filenames.append(frag_sanitized)
+                    self._append_fragment(ctx, frag_content)
                     # We only download the first fragment during the test
                     if test:
                         break
                     i += 1
                     media_sequence += 1
                 elif line.startswith('#EXT-X-KEY'):
+                    decrypt_url = decrypt_info.get('URI')
                     decrypt_info = parse_m3u8_attributes(line[11:])
                     if decrypt_info['METHOD'] == 'AES-128':
                         if 'IV' in decrypt_info:
@@ -164,13 +184,20 @@ class HlsFD(FragmentFD):
                                 man_url, decrypt_info['URI'])
                         if extra_query:
                             decrypt_info['URI'] = update_url_query(decrypt_info['URI'], extra_query)
-                        decrypt_info['KEY'] = self.ydl.urlopen(decrypt_info['URI']).read()
+                        if decrypt_url != decrypt_info['URI']:
+                            decrypt_info['KEY'] = None
                 elif line.startswith('#EXT-X-MEDIA-SEQUENCE'):
                     media_sequence = int(line[22:])
+                elif line.startswith('#EXT-X-BYTERANGE'):
+                    splitted_byte_range = line[17:].split('@')
+                    sub_range_start = int(splitted_byte_range[1]) if len(splitted_byte_range) == 2 else byte_range['end']
+                    byte_range = {
+                        'start': sub_range_start,
+                        'end': sub_range_start + int(splitted_byte_range[0]),
+                    }
+                elif anvato_ad(line):
+                    ad_frag_next = True
 
         self._finish_frag_download(ctx)
 
-        for frag_file in frags_filenames:
-            os.remove(encodeFilename(frag_file))
-
         return True

youtube_dl/downloader/http.py

@@ -22,8 +22,16 @@ from ..utils import (
 class HttpFD(FileDownloader):
     def real_download(self, filename, info_dict):
         url = info_dict['url']
-        tmpfilename = self.temp_name(filename)
-        stream = None
+
+        class DownloadContext(dict):
+            __getattr__ = dict.get
+            __setattr__ = dict.__setitem__
+            __delattr__ = dict.__delitem__
+
+        ctx = DownloadContext()
+        ctx.filename = filename
+        ctx.tmpfilename = self.temp_name(filename)
+        ctx.stream = None
 
         # Do not include the Accept-Encoding header
         headers = {'Youtubedl-no-compression': 'True'}
@@ -38,46 +46,51 @@ class HttpFD(FileDownloader):
         if is_test:
             request.add_header('Range', 'bytes=0-%s' % str(self._TEST_FILE_SIZE - 1))
 
-        # Establish possible resume length
-        if os.path.isfile(encodeFilename(tmpfilename)):
-            resume_len = os.path.getsize(encodeFilename(tmpfilename))
-        else:
-            resume_len = 0
+        ctx.open_mode = 'wb'
+        ctx.resume_len = 0
 
-        open_mode = 'wb'
-        if resume_len != 0:
-            if self.params.get('continuedl', True):
-                self.report_resuming_byte(resume_len)
-                request.add_header('Range', 'bytes=%d-' % resume_len)
-                open_mode = 'ab'
-            else:
-                resume_len = 0
+        if self.params.get('continuedl', True):
+            # Establish possible resume length
+            if os.path.isfile(encodeFilename(ctx.tmpfilename)):
+                ctx.resume_len = os.path.getsize(encodeFilename(ctx.tmpfilename))
 
         count = 0
         retries = self.params.get('retries', 0)
-        while count <= retries:
+
+        class SucceedDownload(Exception):
+            pass
+
+        class RetryDownload(Exception):
+            def __init__(self, source_error):
+                self.source_error = source_error
+
+        def establish_connection():
+            if ctx.resume_len != 0:
+                self.report_resuming_byte(ctx.resume_len)
+                request.add_header('Range', 'bytes=%d-' % ctx.resume_len)
+                ctx.open_mode = 'ab'
             # Establish connection
             try:
-                data = self.ydl.urlopen(request)
+                ctx.data = self.ydl.urlopen(request)
                 # When trying to resume, Content-Range HTTP header of response has to be checked
                 # to match the value of requested Range HTTP header. This is due to a webservers
                 # that don't support resuming and serve a whole file with no Content-Range
                 # set in response despite of requested Range (see
                 # https://github.com/rg3/youtube-dl/issues/6057#issuecomment-126129799)
-                if resume_len > 0:
-                    content_range = data.headers.get('Content-Range')
+                if ctx.resume_len > 0:
+                    content_range = ctx.data.headers.get('Content-Range')
                     if content_range:
                         content_range_m = re.search(r'bytes (\d+)-', content_range)
                         # Content-Range is present and matches requested Range, resume is possible
-                        if content_range_m and resume_len == int(content_range_m.group(1)):
-                            break
+                        if content_range_m and ctx.resume_len == int(content_range_m.group(1)):
+                            return
                     # Content-Range is either not present or invalid. Assuming remote webserver is
                     # trying to send the whole file, resume is not possible, so wiping the local file
                     # and performing entire redownload
                     self.report_unable_to_resume()
-                    resume_len = 0
-                    open_mode = 'wb'
-                break
+                    ctx.resume_len = 0
+                    ctx.open_mode = 'wb'
+                return
             except (compat_urllib_error.HTTPError, ) as err:
                 if (err.code < 500 or err.code >= 600) and err.code != 416:
                     # Unexpected HTTP error
@@ -86,15 +99,15 @@ class HttpFD(FileDownloader):
                     # Unable to resume (requested range not satisfiable)
                     try:
                         # Open the connection again without the range header
-                        data = self.ydl.urlopen(basic_request)
-                        content_length = data.info()['Content-Length']
+                        ctx.data = self.ydl.urlopen(basic_request)
+                        content_length = ctx.data.info()['Content-Length']
                     except (compat_urllib_error.HTTPError, ) as err:
                         if err.code < 500 or err.code >= 600:
                             raise
                     else:
                         # Examine the reported length
                         if (content_length is not None and
-                                (resume_len - 100 < int(content_length) < resume_len + 100)):
+                                (ctx.resume_len - 100 < int(content_length) < ctx.resume_len + 100)):
                             # The file had already been fully downloaded.
                             # Explanation to the above condition: in issue #175 it was revealed that
                             # YouTube sometimes adds or removes a few bytes from the end of the file,
@@ -102,152 +115,183 @@ class HttpFD(FileDownloader):
                             # I decided to implement a suggested change and consider the file
                             # completely downloaded if the file size differs less than 100 bytes from
                             # the one in the hard drive.
-                            self.report_file_already_downloaded(filename)
-                            self.try_rename(tmpfilename, filename)
+                            self.report_file_already_downloaded(ctx.filename)
+                            self.try_rename(ctx.tmpfilename, ctx.filename)
                             self._hook_progress({
-                                'filename': filename,
+                                'filename': ctx.filename,
                                 'status': 'finished',
-                                'downloaded_bytes': resume_len,
-                                'total_bytes': resume_len,
+                                'downloaded_bytes': ctx.resume_len,
+                                'total_bytes': ctx.resume_len,
                             })
-                            return True
+                            raise SucceedDownload()
                         else:
                             # The length does not match, we start the download over
                             self.report_unable_to_resume()
-                            resume_len = 0
-                            open_mode = 'wb'
-                            break
-            except socket.error as e:
-                if e.errno != errno.ECONNRESET:
+                            ctx.resume_len = 0
+                            ctx.open_mode = 'wb'
+                            return
+                raise RetryDownload(err)
+            except socket.error as err:
+                if err.errno != errno.ECONNRESET:
                     # Connection reset is no problem, just retry
                     raise
+                raise RetryDownload(err)
 
-            # Retry
-            count += 1
-            if count <= retries:
-                self.report_retry(count, retries)
+        def download():
+            data_len = ctx.data.info().get('Content-length', None)
 
-        if count > retries:
-            self.report_error('giving up after %s retries' % retries)
-            return False
+            # Range HTTP header may be ignored/unsupported by a webserver
+            # (e.g. extractor/scivee.py, extractor/bambuser.py).
+            # However, for a test we still would like to download just a piece of a file.
+            # To achieve this we limit data_len to _TEST_FILE_SIZE and manually control
+            # block size when downloading a file.
+            if is_test and (data_len is None or int(data_len) > self._TEST_FILE_SIZE):
+                data_len = self._TEST_FILE_SIZE
 
-        data_len = data.info().get('Content-length', None)
-
-        # Range HTTP header may be ignored/unsupported by a webserver
-        # (e.g. extractor/scivee.py, extractor/bambuser.py).
-        # However, for a test we still would like to download just a piece of a file.
-        # To achieve this we limit data_len to _TEST_FILE_SIZE and manually control
-        # block size when downloading a file.
-        if is_test and (data_len is None or int(data_len) > self._TEST_FILE_SIZE):
-            data_len = self._TEST_FILE_SIZE
-
-        if data_len is not None:
-            data_len = int(data_len) + resume_len
-            min_data_len = self.params.get('min_filesize')
-            max_data_len = self.params.get('max_filesize')
-            if min_data_len is not None and data_len < min_data_len:
-                self.to_screen('\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (data_len, min_data_len))
-                return False
-            if max_data_len is not None and data_len > max_data_len:
-                self.to_screen('\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
-                return False
-
-        byte_counter = 0 + resume_len
-        block_size = self.params.get('buffersize', 1024)
-        start = time.time()
-
-        # measure time over whole while-loop, so slow_down() and best_block_size() work together properly
-        now = None  # needed for slow_down() in the first loop run
-        before = start  # start measuring
-        while True:
-
-            # Download and write
-            data_block = data.read(block_size if not is_test else min(block_size, data_len - byte_counter))
-            byte_counter += len(data_block)
-
-            # exit loop when download is finished
-            if len(data_block) == 0:
-                break
-
-            # Open destination file just in time
-            if stream is None:
-                try:
-                    (stream, tmpfilename) = sanitize_open(tmpfilename, open_mode)
-                    assert stream is not None
-                    filename = self.undo_temp_name(tmpfilename)
-                    self.report_destination(filename)
-                except (OSError, IOError) as err:
-                    self.report_error('unable to open for writing: %s' % str(err))
+            if data_len is not None:
+                data_len = int(data_len) + ctx.resume_len
+                min_data_len = self.params.get('min_filesize')
+                max_data_len = self.params.get('max_filesize')
+                if min_data_len is not None and data_len < min_data_len:
+                    self.to_screen('\r[download] File is smaller than min-filesize (%s bytes < %s bytes). Aborting.' % (data_len, min_data_len))
+                    return False
+                if max_data_len is not None and data_len > max_data_len:
+                    self.to_screen('\r[download] File is larger than max-filesize (%s bytes > %s bytes). Aborting.' % (data_len, max_data_len))
                     return False
 
-                if self.params.get('xattr_set_filesize', False) and data_len is not None:
+            byte_counter = 0 + ctx.resume_len
+            block_size = self.params.get('buffersize', 1024)
+            start = time.time()
+
+            # measure time over whole while-loop, so slow_down() and best_block_size() work together properly
+            now = None  # needed for slow_down() in the first loop run
+            before = start  # start measuring
+
+            def retry(e):
+                if ctx.tmpfilename != '-':
+                    ctx.stream.close()
+                ctx.stream = None
+                ctx.resume_len = os.path.getsize(encodeFilename(ctx.tmpfilename))
+                raise RetryDownload(e)
+
+            while True:
+                try:
+                    # Download and write
+                    data_block = ctx.data.read(block_size if not is_test else min(block_size, data_len - byte_counter))
+                # socket.timeout is a subclass of socket.error but may not have
+                # errno set
+                except socket.timeout as e:
+                    retry(e)
+                except socket.error as e:
+                    if e.errno not in (errno.ECONNRESET, errno.ETIMEDOUT):
+                        raise
+                    retry(e)
+
+                byte_counter += len(data_block)
+
+                # exit loop when download is finished
+                if len(data_block) == 0:
+                    break
+
+                # Open destination file just in time
+                if ctx.stream is None:
                     try:
-                        write_xattr(tmpfilename, 'user.ytdl.filesize', str(data_len).encode('utf-8'))
-                    except (XAttrUnavailableError, XAttrMetadataError) as err:
-                        self.report_error('unable to set filesize xattr: %s' % str(err))
+                        ctx.stream, ctx.tmpfilename = sanitize_open(
+                            ctx.tmpfilename, ctx.open_mode)
+                        assert ctx.stream is not None
+                        ctx.filename = self.undo_temp_name(ctx.tmpfilename)
+                        self.report_destination(ctx.filename)
+                    except (OSError, IOError) as err:
+                        self.report_error('unable to open for writing: %s' % str(err))
+                        return False
 
-            try:
-                stream.write(data_block)
-            except (IOError, OSError) as err:
+                    if self.params.get('xattr_set_filesize', False) and data_len is not None:
+                        try:
+                            write_xattr(ctx.tmpfilename, 'user.ytdl.filesize', str(data_len).encode('utf-8'))
+                        except (XAttrUnavailableError, XAttrMetadataError) as err:
+                            self.report_error('unable to set filesize xattr: %s' % str(err))
+
+                try:
+                    ctx.stream.write(data_block)
+                except (IOError, OSError) as err:
+                    self.to_stderr('\n')
+                    self.report_error('unable to write data: %s' % str(err))
+                    return False
+
+                # Apply rate limit
+                self.slow_down(start, now, byte_counter - ctx.resume_len)
+
+                # end measuring of one loop run
+                now = time.time()
+                after = now
+
+                # Adjust block size
+                if not self.params.get('noresizebuffer', False):
+                    block_size = self.best_block_size(after - before, len(data_block))
+
+                before = after
+
+                # Progress message
+                speed = self.calc_speed(start, now, byte_counter - ctx.resume_len)
+                if data_len is None:
+                    eta = None
+                else:
+                    eta = self.calc_eta(start, time.time(), data_len - ctx.resume_len, byte_counter - ctx.resume_len)
+
+                self._hook_progress({
+                    'status': 'downloading',
+                    'downloaded_bytes': byte_counter,
+                    'total_bytes': data_len,
+                    'tmpfilename': ctx.tmpfilename,
+                    'filename': ctx.filename,
+                    'eta': eta,
+                    'speed': speed,
+                    'elapsed': now - start,
+                })
+
+                if is_test and byte_counter == data_len:
+                    break
+
+            if ctx.stream is None:
                 self.to_stderr('\n')
-                self.report_error('unable to write data: %s' % str(err))
+                self.report_error('Did not get any data blocks')
                 return False
+            if ctx.tmpfilename != '-':
+                ctx.stream.close()
 
-            # Apply rate limit
-            self.slow_down(start, now, byte_counter - resume_len)
+            if data_len is not None and byte_counter != data_len:
+                err = ContentTooShortError(byte_counter, int(data_len))
+                if count <= retries:
+                    retry(err)
+                raise err
 
-            # end measuring of one loop run
-            now = time.time()
-            after = now
+            self.try_rename(ctx.tmpfilename, ctx.filename)
 
-            # Adjust block size
-            if not self.params.get('noresizebuffer', False):
-                block_size = self.best_block_size(after - before, len(data_block))
-
-            before = after
-
-            # Progress message
-            speed = self.calc_speed(start, now, byte_counter - resume_len)
-            if data_len is None:
-                eta = None
-            else:
-                eta = self.calc_eta(start, time.time(), data_len - resume_len, byte_counter - resume_len)
+            # Update file modification time
+            if self.params.get('updatetime', True):
+                info_dict['filetime'] = self.try_utime(ctx.filename, ctx.data.info().get('last-modified', None))
 
             self._hook_progress({
-                'status': 'downloading',
                 'downloaded_bytes': byte_counter,
-                'total_bytes': data_len,
-                'tmpfilename': tmpfilename,
-                'filename': filename,
-                'eta': eta,
-                'speed': speed,
-                'elapsed': now - start,
+                'total_bytes': byte_counter,
+                'filename': ctx.filename,
+                'status': 'finished',
+                'elapsed': time.time() - start,
             })
 
-            if is_test and byte_counter == data_len:
-                break
+            return True
 
-        if stream is None:
-            self.to_stderr('\n')
-            self.report_error('Did not get any data blocks')
-            return False
-        if tmpfilename != '-':
-            stream.close()
+        while count <= retries:
+            try:
+                establish_connection()
+                return download()
+            except RetryDownload as e:
+                count += 1
+                if count <= retries:
+                    self.report_retry(e.source_error, count, retries)
+                continue
+            except SucceedDownload:
+                return True
 
-        if data_len is not None and byte_counter != data_len:
-            raise ContentTooShortError(byte_counter, int(data_len))
-        self.try_rename(tmpfilename, filename)
-
-        # Update file modification time
-        if self.params.get('updatetime', True):
-            info_dict['filetime'] = self.try_utime(filename, data.info().get('last-modified', None))
-
-        self._hook_progress({
-            'downloaded_bytes': byte_counter,
-            'total_bytes': byte_counter,
-            'filename': filename,
-            'status': 'finished',
-            'elapsed': time.time() - start,
-        })
-
-        return True
+        self.report_error('giving up after %s retries' % retries)
+        return False

youtube_dl/downloader/ism.py

@@ -1,6 +1,5 @@
 from __future__ import unicode_literals
 
-import os
 import time
 import struct
 import binascii
@@ -8,10 +7,6 @@ import io
 
 from .fragment import FragmentFD
 from ..compat import compat_urllib_error
-from ..utils import (
-    sanitize_open,
-    encodeFilename,
-)
 
 
 u8 = struct.Struct(b'>B')
@@ -103,7 +98,7 @@ def write_piff_header(stream, params):
 
     if is_audio:
         smhd_payload = s88.pack(0)  # balance
-        smhd_payload = u16.pack(0)  # reserved
+        smhd_payload += u16.pack(0)  # reserved
         media_header_box = full_box(b'smhd', 0, 0, smhd_payload)  # Sound Media Header
     else:
         vmhd_payload = u16.pack(0)  # graphics mode
@@ -131,7 +126,6 @@ def write_piff_header(stream, params):
         if fourcc == 'AACL':
             sample_entry_box = box(b'mp4a', sample_entry_payload)
     else:
-        sample_entry_payload = sample_entry_payload
         sample_entry_payload += u16.pack(0)  # pre defined
         sample_entry_payload += u16.pack(0)  # reserved
         sample_entry_payload += u32.pack(0) * 3  # pre defined
@@ -225,47 +219,39 @@ class IsmFD(FragmentFD):
 
         self._prepare_and_start_frag_download(ctx)
 
-        segments_filenames = []
-
         fragment_retries = self.params.get('fragment_retries', 0)
         skip_unavailable_fragments = self.params.get('skip_unavailable_fragments', True)
 
         track_written = False
+        frag_index = 0
         for i, segment in enumerate(segments):
-            segment_url = segment['url']
-            segment_name = 'Frag%d' % i
-            target_filename = '%s-%s' % (ctx['tmpfilename'], segment_name)
+            frag_index += 1
+            if frag_index <= ctx['fragment_index']:
+                continue
             count = 0
             while count <= fragment_retries:
                 try:
-                    success = ctx['dl'].download(target_filename, {'url': segment_url})
+                    success, frag_content = self._download_fragment(ctx, segment['url'], info_dict)
                     if not success:
                         return False
-                    down, target_sanitized = sanitize_open(target_filename, 'rb')
-                    down_data = down.read()
                     if not track_written:
-                        tfhd_data = extract_box_data(down_data, [b'moof', b'traf', b'tfhd'])
+                        tfhd_data = extract_box_data(frag_content, [b'moof', b'traf', b'tfhd'])
                         info_dict['_download_params']['track_id'] = u32.unpack(tfhd_data[4:8])[0]
                         write_piff_header(ctx['dest_stream'], info_dict['_download_params'])
                         track_written = True
-                    ctx['dest_stream'].write(down_data)
-                    down.close()
-                    segments_filenames.append(target_sanitized)
+                    self._append_fragment(ctx, frag_content)
                     break
                 except compat_urllib_error.HTTPError as err:
                     count += 1
                     if count <= fragment_retries:
-                        self.report_retry_fragment(err, segment_name, count, fragment_retries)
+                        self.report_retry_fragment(err, frag_index, count, fragment_retries)
             if count > fragment_retries:
                 if skip_unavailable_fragments:
-                    self.report_skip_fragment(segment_name)
+                    self.report_skip_fragment(frag_index)
                     continue
                 self.report_error('giving up after %s fragment retries' % fragment_retries)
                 return False
 
         self._finish_frag_download(ctx)
 
-        for segment_file in segments_filenames:
-            os.remove(encodeFilename(segment_file))
-
         return True

youtube_dl/downloader/rtmp.py

@@ -169,7 +169,7 @@ class RtmpFD(FileDownloader):
             self.report_error('[rtmpdump] Could not connect to RTMP server.')
             return False
 
-        while (retval == RD_INCOMPLETE or retval == RD_FAILED) and not test and not live:
+        while retval in (RD_INCOMPLETE, RD_FAILED) and not test and not live:
             prevsize = os.path.getsize(encodeFilename(tmpfilename))
             self.to_screen('[rtmpdump] %s bytes' % prevsize)
             time.sleep(5.0)  # This seems to be needed

youtube_dl/extractor/abc.py

@@ -1,13 +1,19 @@
 from __future__ import unicode_literals
 
+import hashlib
+import hmac
 import re
+import time
 
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
     ExtractorError,
     js_to_json,
     int_or_none,
     parse_iso8601,
+    try_get,
+    update_url_query,
 )
 
 
@@ -99,21 +105,24 @@ class ABCIE(InfoExtractor):
 class ABCIViewIE(InfoExtractor):
     IE_NAME = 'abc.net.au:iview'
     _VALID_URL = r'https?://iview\.abc\.net\.au/programs/[^/]+/(?P<id>[^/?#]+)'
+    _GEO_COUNTRIES = ['AU']
 
     # ABC iview programs are normally available for 14 days only.
     _TESTS = [{
-        'url': 'http://iview.abc.net.au/programs/diaries-of-a-broken-mind/ZX9735A001S00',
+        'url': 'http://iview.abc.net.au/programs/call-the-midwife/ZW0898A003S00',
         'md5': 'cde42d728b3b7c2b32b1b94b4a548afc',
         'info_dict': {
-            'id': 'ZX9735A001S00',
+            'id': 'ZW0898A003S00',
             'ext': 'mp4',
-            'title': 'Diaries Of A Broken Mind',
-            'description': 'md5:7de3903874b7a1be279fe6b68718fc9e',
-            'upload_date': '20161010',
-            'uploader_id': 'abc2',
-            'timestamp': 1476064920,
+            'title': 'Series 5 Ep 3',
+            'description': 'md5:e0ef7d4f92055b86c4f33611f180ed79',
+            'upload_date': '20171228',
+            'uploader_id': 'abc1',
+            'timestamp': 1514499187,
+        },
+        'params': {
+            'skip_download': True,
         },
-        'skip': 'Video gone',
     }]
 
     def _real_extract(self, url):
@@ -124,7 +133,30 @@ class ABCIViewIE(InfoExtractor):
         title = video_params.get('title') or video_params['seriesTitle']
         stream = next(s for s in video_params['playlist'] if s.get('type') == 'program')
 
-        formats = self._extract_akamai_formats(stream['hds-unmetered'], video_id)
+        house_number = video_params.get('episodeHouseNumber')
+        path = '/auth/hls/sign?ts={0}&hn={1}&d=android-mobile'.format(
+            int(time.time()), house_number)
+        sig = hmac.new(
+            'android.content.res.Resources'.encode('utf-8'),
+            path.encode('utf-8'), hashlib.sha256).hexdigest()
+        token = self._download_webpage(
+            'http://iview.abc.net.au{0}&sig={1}'.format(path, sig), video_id)
+
+        def tokenize_url(url, token):
+            return update_url_query(url, {
+                'hdnea': token,
+            })
+
+        for sd in ('sd', 'sd-low'):
+            sd_url = try_get(
+                stream, lambda x: x['streams']['hls'][sd], compat_str)
+            if not sd_url:
+                continue
+            formats = self._extract_m3u8_formats(
+                tokenize_url(sd_url, token), video_id, 'mp4',
+                entry_protocol='m3u8_native', m3u8_id='hls', fatal=False)
+            if formats:
+                break
         self._sort_formats(formats)
 
         subtitles = {}

youtube_dl/extractor/abcnews.py

@@ -7,12 +7,21 @@ import time
 
 from .amp import AMPIE
 from .common import InfoExtractor
+from .youtube import YoutubeIE
 from ..compat import compat_urlparse
 
 
 class AbcNewsVideoIE(AMPIE):
     IE_NAME = 'abcnews:video'
-    _VALID_URL = r'https?://abcnews\.go\.com/[^/]+/video/(?P<display_id>[0-9a-z-]+)-(?P<id>\d+)'
+    _VALID_URL = r'''(?x)
+                    https?://
+                        abcnews\.go\.com/
+                        (?:
+                            [^/]+/video/(?P<display_id>[0-9a-z-]+)-|
+                            video/embed\?.*?\bid=
+                        )
+                        (?P<id>\d+)
+                    '''
 
     _TESTS = [{
         'url': 'http://abcnews.go.com/ThisWeek/video/week-exclusive-irans-foreign-minister-zarif-20411932',
@@ -29,6 +38,9 @@ class AbcNewsVideoIE(AMPIE):
             # m3u8 download
             'skip_download': True,
         },
+    }, {
+        'url': 'http://abcnews.go.com/video/embed?id=46979033',
+        'only_matching': True,
     }, {
         'url': 'http://abcnews.go.com/2020/video/2020-husband-stands-teacher-jail-student-affairs-26119478',
         'only_matching': True,
@@ -97,9 +109,7 @@ class AbcNewsIE(InfoExtractor):
             r'window\.abcnvideo\.url\s*=\s*"([^"]+)"', webpage, 'video URL')
         full_video_url = compat_urlparse.urljoin(url, video_url)
 
-        youtube_url = self._html_search_regex(
-            r'<iframe[^>]+src="(https://www\.youtube\.com/embed/[^"]+)"',
-            webpage, 'YouTube URL', default=None)
+        youtube_url = YoutubeIE._extract_url(webpage)
 
         timestamp = None
         date_str = self._html_search_regex(
@@ -129,7 +139,7 @@ class AbcNewsIE(InfoExtractor):
         }
 
         if youtube_url:
-            entries = [entry, self.url_result(youtube_url, 'Youtube')]
+            entries = [entry, self.url_result(youtube_url, ie=YoutubeIE.ie_key())]
             return self.playlist_result(entries)
 
         return entry

youtube_dl/extractor/abcotvs.py

@@ -22,7 +22,7 @@ class ABCOTVSIE(InfoExtractor):
                 'display_id': 'east-bay-museum-celebrates-vintage-synthesizers',
                 'ext': 'mp4',
                 'title': 'East Bay museum celebrates vintage synthesizers',
-                'description': 'md5:a4f10fb2f2a02565c1749d4adbab4b10',
+                'description': 'md5:24ed2bd527096ec2a5c67b9d5a9005f3',
                 'thumbnail': r're:^https?://.*\.jpg$',
                 'timestamp': 1421123075,
                 'upload_date': '20150113',

youtube_dl/extractor/acast.py

@@ -8,7 +8,7 @@ from .common import InfoExtractor
 from ..compat import compat_str
 from ..utils import (
     int_or_none,
-    parse_iso8601,
+    unified_timestamp,
     OnDemandPagedList,
 )
 
@@ -32,7 +32,7 @@ class ACastIE(InfoExtractor):
     }, {
         # test with multiple blings
         'url': 'https://www.acast.com/sparpodcast/2.raggarmordet-rosterurdetforflutna',
-        'md5': '55c0097badd7095f494c99a172f86501',
+        'md5': 'e87d5b8516cd04c0d81b6ee1caca28d0',
         'info_dict': {
             'id': '2a92b283-1a75-4ad8-8396-499c641de0d9',
             'ext': 'mp3',
@@ -40,23 +40,24 @@ class ACastIE(InfoExtractor):
             'timestamp': 1477346700,
             'upload_date': '20161024',
             'description': 'md5:4f81f6d8cf2e12ee21a321d8bca32db4',
-            'duration': 2797,
+            'duration': 2766,
         }
     }]
 
     def _real_extract(self, url):
         channel, display_id = re.match(self._VALID_URL, url).groups()
         cast_data = self._download_json(
-            'https://embed.acast.com/api/acasts/%s/%s' % (channel, display_id), display_id)
+            'https://play-api.acast.com/splash/%s/%s' % (channel, display_id), display_id)
+        e = cast_data['result']['episode']
         return {
-            'id': compat_str(cast_data['id']),
+            'id': compat_str(e['id']),
             'display_id': display_id,
-            'url': [b['audio'] for b in cast_data['blings'] if b['type'] == 'BlingAudio'][0],
-            'title': cast_data['name'],
-            'description': cast_data.get('description'),
-            'thumbnail': cast_data.get('image'),
-            'timestamp': parse_iso8601(cast_data.get('publishingDate')),
-            'duration': int_or_none(cast_data.get('duration')),
+            'url': e['mediaUrl'],
+            'title': e['name'],
+            'description': e.get('description'),
+            'thumbnail': e.get('image'),
+            'timestamp': unified_timestamp(e.get('publishingDate')),
+            'duration': int_or_none(e.get('duration')),
         }
 
 

youtube_dl/extractor/addanime.py

@@ -25,7 +25,8 @@ class AddAnimeIE(InfoExtractor):
             'ext': 'mp4',
             'description': 'One Piece 606',
             'title': 'One Piece 606',
-        }
+        },
+        'skip': 'Video is gone',
     }, {
         'url': 'http://add-anime.net/video/MDUGWYKNGBD8/One-Piece-687',
         'only_matching': True,

youtube_dl/extractor/adn.py

@@ -0,0 +1,150 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import base64
+import json
+import os
+
+from .common import InfoExtractor
+from ..aes import aes_cbc_decrypt
+from ..compat import compat_ord
+from ..utils import (
+    bytes_to_intlist,
+    ExtractorError,
+    float_or_none,
+    intlist_to_bytes,
+    srt_subtitles_timecode,
+    strip_or_none,
+    urljoin,
+)
+
+
+class ADNIE(InfoExtractor):
+    IE_DESC = 'Anime Digital Network'
+    _VALID_URL = r'https?://(?:www\.)?animedigitalnetwork\.fr/video/[^/]+/(?P<id>\d+)'
+    _TEST = {
+        'url': 'http://animedigitalnetwork.fr/video/blue-exorcist-kyoto-saga/7778-episode-1-debut-des-hostilites',
+        'md5': 'e497370d847fd79d9d4c74be55575c7a',
+        'info_dict': {
+            'id': '7778',
+            'ext': 'mp4',
+            'title': 'Blue Exorcist - Kyôto Saga - Épisode 1',
+            'description': 'md5:2f7b5aa76edbc1a7a92cedcda8a528d5',
+        }
+    }
+    _BASE_URL = 'http://animedigitalnetwork.fr'
+
+    def _get_subtitles(self, sub_path, video_id):
+        if not sub_path:
+            return None
+
+        enc_subtitles = self._download_webpage(
+            urljoin(self._BASE_URL, sub_path),
+            video_id, fatal=False, headers={
+                'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64; rv:53.0) Gecko/20100101 Firefox/53.0',
+            })
+        if not enc_subtitles:
+            return None
+
+        # http://animedigitalnetwork.fr/components/com_vodvideo/videojs/adn-vjs.min.js
+        dec_subtitles = intlist_to_bytes(aes_cbc_decrypt(
+            bytes_to_intlist(base64.b64decode(enc_subtitles[24:])),
+            bytes_to_intlist(b'\x1b\xe0\x29\x61\x38\x94\x24\x00\x12\xbd\xc5\x80\xac\xce\xbe\xb0'),
+            bytes_to_intlist(base64.b64decode(enc_subtitles[:24]))
+        ))
+        subtitles_json = self._parse_json(
+            dec_subtitles[:-compat_ord(dec_subtitles[-1])].decode(),
+            None, fatal=False)
+        if not subtitles_json:
+            return None
+
+        subtitles = {}
+        for sub_lang, sub in subtitles_json.items():
+            srt = ''
+            for num, current in enumerate(sub):
+                start, end, text = (
+                    float_or_none(current.get('startTime')),
+                    float_or_none(current.get('endTime')),
+                    current.get('text'))
+                if start is None or end is None or text is None:
+                    continue
+                srt += os.linesep.join(
+                    (
+                        '%d' % num,
+                        '%s --> %s' % (
+                            srt_subtitles_timecode(start),
+                            srt_subtitles_timecode(end)),
+                        text,
+                        os.linesep,
+                    ))
+
+            if sub_lang == 'vostf':
+                sub_lang = 'fr'
+            subtitles.setdefault(sub_lang, []).extend([{
+                'ext': 'json',
+                'data': json.dumps(sub),
+            }, {
+                'ext': 'srt',
+                'data': srt,
+            }])
+        return subtitles
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        player_config = self._parse_json(self._search_regex(
+            r'playerConfig\s*=\s*({.+});', webpage, 'player config'), video_id)
+
+        video_info = {}
+        video_info_str = self._search_regex(
+            r'videoInfo\s*=\s*({.+});', webpage,
+            'video info', fatal=False)
+        if video_info_str:
+            video_info = self._parse_json(
+                video_info_str, video_id, fatal=False) or {}
+
+        options = player_config.get('options') or {}
+        metas = options.get('metas') or {}
+        title = metas.get('title') or video_info['title']
+        links = player_config.get('links') or {}
+        error = None
+        if not links:
+            links_url = player_config['linksurl']
+            links_data = self._download_json(urljoin(
+                self._BASE_URL, links_url), video_id)
+            links = links_data.get('links') or {}
+            error = links_data.get('error')
+
+        formats = []
+        for format_id, qualities in links.items():
+            if not isinstance(qualities, dict):
+                continue
+            for load_balancer_url in qualities.values():
+                load_balancer_data = self._download_json(
+                    load_balancer_url, video_id, fatal=False) or {}
+                m3u8_url = load_balancer_data.get('location')
+                if not m3u8_url:
+                    continue
+                m3u8_formats = self._extract_m3u8_formats(
+                    m3u8_url, video_id, 'mp4', 'm3u8_native',
+                    m3u8_id=format_id, fatal=False)
+                if format_id == 'vf':
+                    for f in m3u8_formats:
+                        f['language'] = 'fr'
+                formats.extend(m3u8_formats)
+        if not error:
+            error = options.get('error')
+        if not formats and error:
+            raise ExtractorError('%s said: %s' % (self.IE_NAME, error), expected=True)
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': strip_or_none(metas.get('summary') or video_info.get('resume')),
+            'thumbnail': video_info.get('image'),
+            'formats': formats,
+            'subtitles': self.extract_subtitles(player_config.get('subtitles'), video_id),
+            'episode': metas.get('subtitle') or video_info.get('videoTitle'),
+            'series': video_info.get('playlistTitle'),
+        }

youtube_dl/extractor/adobepass.py

@@ -6,12 +6,16 @@ import time
 import xml.etree.ElementTree as etree
 
 from .common import InfoExtractor
-from ..compat import compat_urlparse
+from ..compat import (
+    compat_kwargs,
+    compat_urlparse,
+)
 from ..utils import (
     unescapeHTML,
     urlencode_postdata,
     unified_timestamp,
     ExtractorError,
+    NO_DEFAULT,
 )
 
 
@@ -21,6 +25,11 @@ MSO_INFO = {
         'username_field': 'username',
         'password_field': 'password',
     },
+    'ATTOTT': {
+        'name': 'DIRECTV NOW',
+        'username_field': 'email',
+        'password_field': 'loginpassword',
+    },
     'Rogers': {
         'name': 'Rogers',
         'username_field': 'UserName',
@@ -31,6 +40,26 @@ MSO_INFO = {
         'username_field': 'user',
         'password_field': 'passwd',
     },
+    'TWC': {
+        'name': 'Time Warner Cable | Spectrum',
+        'username_field': 'Ecom_User_ID',
+        'password_field': 'Ecom_Password',
+    },
+    'Brighthouse': {
+        'name': 'Bright House Networks | Spectrum',
+        'username_field': 'j_username',
+        'password_field': 'j_password',
+    },
+    'Charter_Direct': {
+        'name': 'Charter Spectrum',
+        'username_field': 'IDToken1',
+        'password_field': 'IDToken2',
+    },
+    'Verizon': {
+        'name': 'Verizon FiOS',
+        'username_field': 'IDToken1',
+        'password_field': 'IDToken2',
+    },
     'thr030': {
         'name': '3 Rivers Communications'
     },
@@ -1293,6 +1322,15 @@ class AdobePassIE(InfoExtractor):
     _USER_AGENT = 'Mozilla/5.0 (X11; Linux i686; rv:47.0) Gecko/20100101 Firefox/47.0'
     _MVPD_CACHE = 'ap-mvpd'
 
+    _DOWNLOADING_LOGIN_PAGE = 'Downloading Provider Login Page'
+
+    def _download_webpage_handle(self, *args, **kwargs):
+        headers = kwargs.get('headers', {})
+        headers.update(self.geo_verification_headers())
+        kwargs['headers'] = headers
+        return super(AdobePassIE, self)._download_webpage_handle(
+            *args, **compat_kwargs(kwargs))
+
     @staticmethod
     def _get_mvpd_resource(provider_id, title, guid, rating):
         channel = etree.Element('channel')
@@ -1335,6 +1373,21 @@ class AdobePassIE(InfoExtractor):
                 'Use --ap-mso to specify Adobe Pass Multiple-system operator Identifier '
                 'and --ap-username and --ap-password or --netrc to provide account credentials.', expected=True)
 
+        def extract_redirect_url(html, url=None, fatal=False):
+            # TODO: eliminate code duplication with generic extractor and move
+            # redirection code into _download_webpage_handle
+            REDIRECT_REGEX = r'[0-9]{,2};\s*(?:URL|url)=\'?([^\'"]+)'
+            redirect_url = self._search_regex(
+                r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
+                r'(?:[a-z-]+="[^"]+"\s+)*?content="%s' % REDIRECT_REGEX,
+                html, 'meta refresh redirect',
+                default=NO_DEFAULT if fatal else None, fatal=fatal)
+            if not redirect_url:
+                return None
+            if url:
+                redirect_url = compat_urlparse.urljoin(url, unescapeHTML(redirect_url))
+            return redirect_url
+
         mvpd_headers = {
             'ap_42': 'anonymous',
             'ap_11': 'Linux i686',
@@ -1374,42 +1427,82 @@ class AdobePassIE(InfoExtractor):
                     # Comcast page flow varies by video site and whether you
                     # are on Comcast's network.
                     provider_redirect_page, urlh = provider_redirect_page_res
-                    # Check for Comcast auto login
                     if 'automatically signing you in' in provider_redirect_page:
                         oauth_redirect_url = self._html_search_regex(
                             r'window\.location\s*=\s*[\'"]([^\'"]+)',
                             provider_redirect_page, 'oauth redirect')
-                        # Just need to process the request. No useful data comes back
                         self._download_webpage(
                             oauth_redirect_url, video_id, 'Confirming auto login')
                     else:
                         if '<form name="signin"' in provider_redirect_page:
-                            # already have the form, just fill it
                             provider_login_page_res = provider_redirect_page_res
                         elif 'http-equiv="refresh"' in provider_redirect_page:
-                            # redirects to the login page
-                            oauth_redirect_url = self._html_search_regex(
-                                r'content="0;\s*url=([^\'"]+)',
-                                provider_redirect_page, 'meta refresh redirect')
+                            oauth_redirect_url = extract_redirect_url(
+                                provider_redirect_page, fatal=True)
                             provider_login_page_res = self._download_webpage_handle(
-                                oauth_redirect_url,
-                                video_id, 'Downloading Provider Login Page')
+                                oauth_redirect_url, video_id,
+                                self._DOWNLOADING_LOGIN_PAGE)
                         else:
                             provider_login_page_res = post_form(
-                                provider_redirect_page_res, 'Downloading Provider Login Page')
+                                provider_redirect_page_res,
+                                self._DOWNLOADING_LOGIN_PAGE)
 
-                        mvpd_confirm_page_res = post_form(provider_login_page_res, 'Logging in', {
-                            mso_info.get('username_field', 'username'): username,
-                            mso_info.get('password_field', 'password'): password,
-                        })
+                        mvpd_confirm_page_res = post_form(
+                            provider_login_page_res, 'Logging in', {
+                                mso_info['username_field']: username,
+                                mso_info['password_field']: password,
+                            })
                         mvpd_confirm_page, urlh = mvpd_confirm_page_res
                         if '<button class="submit" value="Resume">Resume</button>' in mvpd_confirm_page:
                             post_form(mvpd_confirm_page_res, 'Confirming Login')
-
+                elif mso_id == 'Verizon':
+                    # In general, if you're connecting from a Verizon-assigned IP,
+                    # you will not actually pass your credentials.
+                    provider_redirect_page, urlh = provider_redirect_page_res
+                    if 'Please wait ...' in provider_redirect_page:
+                        saml_redirect_url = self._html_search_regex(
+                            r'self\.parent\.location=(["\'])(?P<url>.+?)\1',
+                            provider_redirect_page,
+                            'SAML Redirect URL', group='url')
+                        saml_login_page = self._download_webpage(
+                            saml_redirect_url, video_id,
+                            'Downloading SAML Login Page')
+                    else:
+                        saml_login_page_res = post_form(
+                            provider_redirect_page_res, 'Logging in', {
+                                mso_info['username_field']: username,
+                                mso_info['password_field']: password,
+                            })
+                        saml_login_page, urlh = saml_login_page_res
+                        if 'Please try again.' in saml_login_page:
+                            raise ExtractorError(
+                                'We\'re sorry, but either the User ID or Password entered is not correct.')
+                    saml_login_url = self._search_regex(
+                        r'xmlHttp\.open\("POST"\s*,\s*(["\'])(?P<url>.+?)\1',
+                        saml_login_page, 'SAML Login URL', group='url')
+                    saml_response_json = self._download_json(
+                        saml_login_url, video_id, 'Downloading SAML Response',
+                        headers={'Content-Type': 'text/xml'})
+                    self._download_webpage(
+                        saml_response_json['targetValue'], video_id,
+                        'Confirming Login', data=urlencode_postdata({
+                            'SAMLResponse': saml_response_json['SAMLResponse'],
+                            'RelayState': saml_response_json['RelayState']
+                        }), headers={
+                            'Content-Type': 'application/x-www-form-urlencoded'
+                        })
                 else:
-                    # Normal, non-Comcast flow
+                    # Some providers (e.g. DIRECTV NOW) have another meta refresh
+                    # based redirect that should be followed.
+                    provider_redirect_page, urlh = provider_redirect_page_res
+                    provider_refresh_redirect_url = extract_redirect_url(
+                        provider_redirect_page, url=urlh.geturl())
+                    if provider_refresh_redirect_url:
+                        provider_redirect_page_res = self._download_webpage_handle(
+                            provider_refresh_redirect_url, video_id,
+                            'Downloading Provider Redirect Page (meta refresh)')
                     provider_login_page_res = post_form(
-                        provider_redirect_page_res, 'Downloading Provider Login Page')
+                        provider_redirect_page_res, self._DOWNLOADING_LOGIN_PAGE)
                     mvpd_confirm_page_res = post_form(provider_login_page_res, 'Logging in', {
                         mso_info.get('username_field', 'username'): username,
                         mso_info.get('password_field', 'password'): password,
@@ -1448,6 +1541,8 @@ class AdobePassIE(InfoExtractor):
                     self._downloader.cache.store(self._MVPD_CACHE, requestor_id, {})
                     count += 1
                     continue
+                if '<error' in authorize:
+                    raise ExtractorError(xml_text(authorize, 'details'), expected=True)
                 authz_token = unescapeHTML(xml_text(authorize, 'authzToken'))
                 requestor_info[guid] = authz_token
                 self._downloader.cache.store(self._MVPD_CACHE, requestor_id, requestor_info)

youtube_dl/extractor/adultswim.py

@@ -5,91 +5,52 @@ import re
 
 from .turner import TurnerBaseIE
 from ..utils import (
-    ExtractorError,
     int_or_none,
+    strip_or_none,
 )
 
 
 class AdultSwimIE(TurnerBaseIE):
-    _VALID_URL = r'https?://(?:www\.)?adultswim\.com/videos/(?P<is_playlist>playlists/)?(?P<show_path>[^/]+)/(?P<episode_path>[^/?#]+)/?'
+    _VALID_URL = r'https?://(?:www\.)?adultswim\.com/videos/(?P<show_path>[^/?#]+)(?:/(?P<episode_path>[^/?#]+))?'
 
     _TESTS = [{
         'url': 'http://adultswim.com/videos/rick-and-morty/pilot',
-        'playlist': [
-            {
-                'md5': '247572debc75c7652f253c8daa51a14d',
-                'info_dict': {
-                    'id': 'rQxZvXQ4ROaSOqq-or2Mow-0',
-                    'ext': 'flv',
-                    'title': 'Rick and Morty - Pilot Part 1',
-                    'description': "Rick moves in with his daughter's family and establishes himself as a bad influence on his grandson, Morty. "
-                },
-            },
-            {
-                'md5': '77b0e037a4b20ec6b98671c4c379f48d',
-                'info_dict': {
-                    'id': 'rQxZvXQ4ROaSOqq-or2Mow-3',
-                    'ext': 'flv',
-                    'title': 'Rick and Morty - Pilot Part 4',
-                    'description': "Rick moves in with his daughter's family and establishes himself as a bad influence on his grandson, Morty. "
-                },
-            },
-        ],
         'info_dict': {
             'id': 'rQxZvXQ4ROaSOqq-or2Mow',
+            'ext': 'mp4',
             'title': 'Rick and Morty - Pilot',
-            'description': "Rick moves in with his daughter's family and establishes himself as a bad influence on his grandson, Morty. "
-        },
-        'skip': 'This video is only available for registered users',
-    }, {
-        'url': 'http://www.adultswim.com/videos/playlists/american-parenting/putting-francine-out-of-business/',
-        'playlist': [
-            {
-                'md5': '2eb5c06d0f9a1539da3718d897f13ec5',
-                'info_dict': {
-                    'id': '-t8CamQlQ2aYZ49ItZCFog-0',
-                    'ext': 'flv',
-                    'title': 'American Dad - Putting Francine Out of Business',
-                    'description': 'Stan hatches a plan to get Francine out of the real estate business.Watch more American Dad on [adult swim].'
-                },
-            }
-        ],
-        'info_dict': {
-            'id': '-t8CamQlQ2aYZ49ItZCFog',
-            'title': 'American Dad - Putting Francine Out of Business',
-            'description': 'Stan hatches a plan to get Francine out of the real estate business.Watch more American Dad on [adult swim].'
-        },
-    }, {
-        'url': 'http://www.adultswim.com/videos/tim-and-eric-awesome-show-great-job/dr-steve-brule-for-your-wine/',
-        'playlist': [
-            {
-                'md5': '3e346a2ab0087d687a05e1e7f3b3e529',
-                'info_dict': {
-                    'id': 'sY3cMUR_TbuE4YmdjzbIcQ-0',
-                    'ext': 'mp4',
-                    'title': 'Tim and Eric Awesome Show Great Job! - Dr. Steve Brule, For Your Wine',
-                    'description': 'Dr. Brule reports live from Wine Country with a special report on wines.  \r\nWatch Tim and Eric Awesome Show Great Job! episode #20, "Embarrassed" on Adult Swim.\r\n\r\n',
-                },
-            }
-        ],
-        'info_dict': {
-            'id': 'sY3cMUR_TbuE4YmdjzbIcQ',
-            'title': 'Tim and Eric Awesome Show Great Job! - Dr. Steve Brule, For Your Wine',
-            'description': 'Dr. Brule reports live from Wine Country with a special report on wines.  \r\nWatch Tim and Eric Awesome Show Great Job! episode #20, "Embarrassed" on Adult Swim.\r\n\r\n',
+            'description': 'Rick moves in with his daughter\'s family and establishes himself as a bad influence on his grandson, Morty.',
+            'timestamp': 1493267400,
+            'upload_date': '20170427',
         },
         'params': {
             # m3u8 download
             'skip_download': True,
-        }
+        },
+        'expected_warnings': ['Unable to download f4m manifest'],
+    }, {
+        'url': 'http://www.adultswim.com/videos/tim-and-eric-awesome-show-great-job/dr-steve-brule-for-your-wine/',
+        'info_dict': {
+            'id': 'sY3cMUR_TbuE4YmdjzbIcQ',
+            'ext': 'mp4',
+            'title': 'Tim and Eric Awesome Show Great Job! - Dr. Steve Brule, For Your Wine',
+            'description': 'Dr. Brule reports live from Wine Country with a special report on wines.  \nWatch Tim and Eric Awesome Show Great Job! episode #20, "Embarrassed" on Adult Swim.',
+            'upload_date': '20080124',
+            'timestamp': 1201150800,
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
     }, {
-        # heroMetadata.trailer
         'url': 'http://www.adultswim.com/videos/decker/inside-decker-a-new-hero/',
         'info_dict': {
             'id': 'I0LQFQkaSUaFp8PnAWHhoQ',
             'ext': 'mp4',
             'title': 'Decker - Inside Decker: A New Hero',
-            'description': 'md5:c916df071d425d62d70c86d4399d3ee0',
-            'duration': 249.008,
+            'description': 'The guys recap the conclusion of the season. They announce a new hero, take a peek into the Victorville Film Archive and welcome back the talented James Dean.',
+            'timestamp': 1469480460,
+            'upload_date': '20160725',
         },
         'params': {
             # m3u8 download
@@ -97,136 +58,102 @@ class AdultSwimIE(TurnerBaseIE):
         },
         'expected_warnings': ['Unable to download f4m manifest'],
     }, {
-        'url': 'http://www.adultswim.com/videos/toonami/friday-october-14th-2016/',
+        'url': 'http://www.adultswim.com/videos/attack-on-titan',
         'info_dict': {
-            'id': 'eYiLsKVgQ6qTC6agD67Sig',
-            'title': 'Toonami - Friday, October 14th, 2016',
-            'description': 'md5:99892c96ffc85e159a428de85c30acde',
+            'id': 'b7A69dzfRzuaXIECdxW8XQ',
+            'title': 'Attack on Titan',
+            'description': 'md5:6c8e003ea0777b47013e894767f5e114',
+        },
+        'playlist_mincount': 12,
+    }, {
+        'url': 'http://www.adultswim.com/videos/streams/williams-stream',
+        'info_dict': {
+            'id': 'd8DEBj7QRfetLsRgFnGEyg',
+            'ext': 'mp4',
+            'title': r're:^Williams Stream \d{4}-\d{2}-\d{2} \d{2}:\d{2}$',
+            'description': 'original programming',
         },
-        'playlist': [{
-            'md5': '',
-            'info_dict': {
-                'id': 'eYiLsKVgQ6qTC6agD67Sig',
-                'ext': 'mp4',
-                'title': 'Toonami - Friday, October 14th, 2016',
-                'description': 'md5:99892c96ffc85e159a428de85c30acde',
-            },
-        }],
         'params': {
             # m3u8 download
             'skip_download': True,
         },
-        'expected_warnings': ['Unable to download f4m manifest'],
     }]
 
-    @staticmethod
-    def find_video_info(collection, slug):
-        for video in collection.get('videos'):
-            if video.get('slug') == slug:
-                return video
-
-    @staticmethod
-    def find_collection_by_linkURL(collections, linkURL):
-        for collection in collections:
-            if collection.get('linkURL') == linkURL:
-                return collection
-
-    @staticmethod
-    def find_collection_containing_video(collections, slug):
-        for collection in collections:
-            for video in collection.get('videos'):
-                if video.get('slug') == slug:
-                    return collection, video
-        return None, None
-
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        show_path = mobj.group('show_path')
-        episode_path = mobj.group('episode_path')
-        is_playlist = True if mobj.group('is_playlist') else False
+        show_path, episode_path = re.match(self._VALID_URL, url).groups()
+        display_id = episode_path or show_path
+        webpage = self._download_webpage(url, display_id)
+        initial_data = self._parse_json(self._search_regex(
+            r'AS_INITIAL_DATA(?:__)?\s*=\s*({.+?});',
+            webpage, 'initial data'), display_id)
 
-        webpage = self._download_webpage(url, episode_path)
+        is_stream = show_path == 'streams'
+        if is_stream:
+            if not episode_path:
+                episode_path = 'live-stream'
 
-        # Extract the value of `bootstrappedData` from the Javascript in the page.
-        bootstrapped_data = self._parse_json(self._search_regex(
-            r'var bootstrappedData = ({.*});', webpage, 'bootstraped data'), episode_path)
+            video_data = next(stream for stream_path, stream in initial_data['streams'].items() if stream_path == episode_path)
+            video_id = video_data.get('stream')
 
-        # Downloading videos from a /videos/playlist/ URL needs to be handled differently.
-        # NOTE: We are only downloading one video (the current one) not the playlist
-        if is_playlist:
-            collections = bootstrapped_data['playlists']['collections']
-            collection = self.find_collection_by_linkURL(collections, show_path)
-            video_info = self.find_video_info(collection, episode_path)
-
-            show_title = video_info['showTitle']
-            segment_ids = [video_info['videoPlaybackID']]
+            if not video_id:
+                entries = []
+                for episode in video_data.get('archiveEpisodes', []):
+                    episode_url = episode.get('url')
+                    if not episode_url:
+                        continue
+                    entries.append(self.url_result(
+                        episode_url, 'AdultSwim', episode.get('id')))
+                return self.playlist_result(
+                    entries, video_data.get('id'), video_data.get('title'),
+                    strip_or_none(video_data.get('description')))
         else:
-            collections = bootstrapped_data['show']['collections']
-            collection, video_info = self.find_collection_containing_video(collections, episode_path)
-            # Video wasn't found in the collections, let's try `slugged_video`.
-            if video_info is None:
-                if bootstrapped_data.get('slugged_video', {}).get('slug') == episode_path:
-                    video_info = bootstrapped_data['slugged_video']
-            if not video_info:
-                video_info = bootstrapped_data.get(
-                    'heroMetadata', {}).get('trailer', {}).get('video')
-            if not video_info:
-                video_info = bootstrapped_data.get('onlineOriginals', [None])[0]
-            if not video_info:
-                raise ExtractorError('Unable to find video info')
+            show_data = initial_data['show']
 
-            show = bootstrapped_data['show']
-            show_title = show['title']
-            stream = video_info.get('stream')
-            if stream and stream.get('videoPlaybackID'):
-                segment_ids = [stream['videoPlaybackID']]
-            elif video_info.get('clips'):
-                segment_ids = [clip['videoPlaybackID'] for clip in video_info['clips']]
-            elif video_info.get('videoPlaybackID'):
-                segment_ids = [video_info['videoPlaybackID']]
-            elif video_info.get('id'):
-                segment_ids = [video_info['id']]
-            else:
-                if video_info.get('auth') is True:
-                    raise ExtractorError(
-                        'This video is only available via cable service provider subscription that'
-                        ' is not currently supported. You may want to use --cookies.', expected=True)
-                else:
-                    raise ExtractorError('Unable to find stream or clips')
+            if not episode_path:
+                entries = []
+                for video in show_data.get('videos', []):
+                    slug = video.get('slug')
+                    if not slug:
+                        continue
+                    entries.append(self.url_result(
+                        'http://adultswim.com/videos/%s/%s' % (show_path, slug),
+                        'AdultSwim', video.get('id')))
+                return self.playlist_result(
+                    entries, show_data.get('id'), show_data.get('title'),
+                    strip_or_none(show_data.get('metadata', {}).get('description')))
 
-        episode_id = video_info['id']
-        episode_title = video_info['title']
-        episode_description = video_info.get('description')
-        episode_duration = int_or_none(video_info.get('duration'))
-        view_count = int_or_none(video_info.get('views'))
+            video_data = show_data['sluggedVideo']
+            video_id = video_data['id']
 
-        entries = []
-        for part_num, segment_id in enumerate(segment_ids):
-            segement_info = self._extract_cvp_info(
-                'http://www.adultswim.com/videos/api/v0/assets?id=%s&platform=desktop' % segment_id,
-                segment_id, {
-                    'secure': {
-                        'media_src': 'http://androidhls-secure.cdn.turner.com/adultswim/big',
-                        'tokenizer_src': 'http://www.adultswim.com/astv/mvpd/processors/services/token_ipadAdobe.do',
-                    },
-                })
-            segment_title = '%s - %s' % (show_title, episode_title)
-            if len(segment_ids) > 1:
-                segment_title += ' Part %d' % (part_num + 1)
-            segement_info.update({
-                'id': segment_id,
-                'title': segment_title,
-                'description': episode_description,
+        info = self._extract_cvp_info(
+            'http://www.adultswim.com/videos/api/v0/assets?platform=desktop&id=' + video_id,
+            video_id, {
+                'secure': {
+                    'media_src': 'http://androidhls-secure.cdn.turner.com/adultswim/big',
+                    'tokenizer_src': 'http://www.adultswim.com/astv/mvpd/processors/services/token_ipadAdobe.do',
+                },
+            }, {
+                'url': url,
+                'site_name': 'AdultSwim',
+                'auth_required': video_data.get('auth'),
             })
-            entries.append(segement_info)
 
-        return {
-            '_type': 'playlist',
-            'id': episode_id,
-            'display_id': episode_path,
-            'entries': entries,
-            'title': '%s - %s' % (show_title, episode_title),
-            'description': episode_description,
-            'duration': episode_duration,
-            'view_count': view_count,
-        }
+        info.update({
+            'id': video_id,
+            'display_id': display_id,
+            'description': info.get('description') or strip_or_none(video_data.get('description')),
+        })
+        if not is_stream:
+            info.update({
+                'duration': info.get('duration') or int_or_none(video_data.get('duration')),
+                'timestamp': info.get('timestamp') or int_or_none(video_data.get('launch_date')),
+                'season_number': info.get('season_number') or int_or_none(video_data.get('season_number')),
+                'episode': info['title'],
+                'episode_number': info.get('episode_number') or int_or_none(video_data.get('episode_number')),
+            })
+
+            info['series'] = video_data.get('collection_title') or info.get('series')
+            if info['series'] and info['series'] != info['title']:
+                info['title'] = '%s - %s' % (info['series'], info['title'])
+
+        return info

youtube_dl/extractor/aenetworks.py

@@ -23,7 +23,19 @@ class AENetworksBaseIE(ThePlatformIE):
 class AENetworksIE(AENetworksBaseIE):
     IE_NAME = 'aenetworks'
     IE_DESC = 'A+E Networks: A&E, Lifetime, History.com, FYI Network'
-    _VALID_URL = r'https?://(?:www\.)?(?P<domain>(?:history|aetv|mylifetime)\.com|fyi\.tv)/(?:shows/(?P<show_path>[^/]+(?:/[^/]+){0,2})|movies/(?P<movie_display_id>[^/]+)/full-movie)'
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:www\.)?
+                        (?P<domain>
+                            (?:history|aetv|mylifetime|lifetimemovieclub)\.com|
+                            fyi\.tv
+                        )/
+                        (?:
+                            shows/(?P<show_path>[^/]+(?:/[^/]+){0,2})|
+                            movies/(?P<movie_display_id>[^/]+)(?:/full-movie)?|
+                            specials/(?P<special_display_id>[^/]+)/full-special
+                        )
+                    '''
     _TESTS = [{
         'url': 'http://www.history.com/shows/mountain-men/season-1/episode-1',
         'md5': 'a97a65f7e823ae10e9244bc5433d5fe6',
@@ -62,17 +74,24 @@ class AENetworksIE(AENetworksBaseIE):
     }, {
         'url': 'http://www.mylifetime.com/movies/center-stage-on-pointe/full-movie',
         'only_matching': True
+    }, {
+        'url': 'https://www.lifetimemovieclub.com/movies/a-killer-among-us',
+        'only_matching': True
+    }, {
+        'url': 'http://www.history.com/specials/sniper-into-the-kill-zone/full-special',
+        'only_matching': True
     }]
     _DOMAIN_TO_REQUESTOR_ID = {
         'history.com': 'HISTORY',
         'aetv.com': 'AETV',
         'mylifetime.com': 'LIFETIME',
+        'lifetimemovieclub.com': 'LIFETIMEMOVIECLUB',
         'fyi.tv': 'FYI',
     }
 
     def _real_extract(self, url):
-        domain, show_path, movie_display_id = re.match(self._VALID_URL, url).groups()
-        display_id = show_path or movie_display_id
+        domain, show_path, movie_display_id, special_display_id = re.match(self._VALID_URL, url).groups()
+        display_id = show_path or movie_display_id or special_display_id
         webpage = self._download_webpage(url, display_id)
         if show_path:
             url_parts = show_path.split('/')
@@ -82,18 +101,22 @@ class AENetworksIE(AENetworksBaseIE):
                 for season_url_path in re.findall(r'(?s)<li[^>]+data-href="(/shows/%s/season-\d+)"' % url_parts[0], webpage):
                     entries.append(self.url_result(
                         compat_urlparse.urljoin(url, season_url_path), 'AENetworks'))
-                return self.playlist_result(
-                    entries, self._html_search_meta('aetn:SeriesId', webpage),
-                    self._html_search_meta('aetn:SeriesTitle', webpage))
-            elif url_parts_len == 2:
+                if entries:
+                    return self.playlist_result(
+                        entries, self._html_search_meta('aetn:SeriesId', webpage),
+                        self._html_search_meta('aetn:SeriesTitle', webpage))
+                else:
+                    # single season
+                    url_parts_len = 2
+            if url_parts_len == 2:
                 entries = []
-                for episode_item in re.findall(r'(?s)<div[^>]+class="[^"]*episode-item[^"]*"[^>]*>', webpage):
+                for episode_item in re.findall(r'(?s)<[^>]+class="[^"]*(?:episode|program)-item[^"]*"[^>]*>', webpage):
                     episode_attributes = extract_attributes(episode_item)
                     episode_url = compat_urlparse.urljoin(
                         url, episode_attributes['data-canonical'])
                     entries.append(self.url_result(
                         episode_url, 'AENetworks',
-                        episode_attributes['data-videoid']))
+                        episode_attributes.get('data-videoid') or episode_attributes.get('data-video-id')))
                 return self.playlist_result(
                     entries, self._html_search_meta('aetn:SeasonId', webpage))
 
@@ -103,9 +126,12 @@ class AENetworksIE(AENetworksBaseIE):
         }
         video_id = self._html_search_meta('aetn:VideoID', webpage)
         media_url = self._search_regex(
-            r"media_url\s*=\s*'([^']+)'", webpage, 'video url')
+            [r"media_url\s*=\s*'(?P<url>[^']+)'",
+             r'data-media-url=(?P<url>(?:https?:)?//[^\s>]+)',
+             r'data-media-url=(["\'])(?P<url>(?:(?!\1).)+?)\1'],
+            webpage, 'video url', group='url')
         theplatform_metadata = self._download_theplatform_metadata(self._search_regex(
-            r'https?://link.theplatform.com/s/([^?]+)', media_url, 'theplatform_path'), video_id)
+            r'https?://link\.theplatform\.com/s/([^?]+)', media_url, 'theplatform_path'), video_id)
         info = self._parse_theplatform_metadata(theplatform_metadata)
         if theplatform_metadata.get('AETN$isBehindWall'):
             requestor_id = self._DOMAIN_TO_REQUESTOR_ID[domain]

youtube_dl/extractor/afreecatv.py

@@ -4,20 +4,17 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import (
-    compat_urllib_parse_urlparse,
-    compat_urlparse,
-)
+from ..compat import compat_xpath
 from ..utils import (
+    determine_ext,
     ExtractorError,
     int_or_none,
-    update_url_query,
-    xpath_element,
     xpath_text,
 )
 
 
 class AfreecaTVIE(InfoExtractor):
+    IE_NAME = 'afreecatv'
     IE_DESC = 'afreecatv.com'
     _VALID_URL = r'''(?x)
                     https?://
@@ -42,7 +39,8 @@ class AfreecaTVIE(InfoExtractor):
             'uploader': 'dailyapril',
             'uploader_id': 'dailyapril',
             'upload_date': '20160503',
-        }
+        },
+        'skip': 'Video is gone',
     }, {
         'url': 'http://afbbs.afreecatv.com:8080/app/read_ucc_bbs.cgi?nStationNo=16711924&nTitleNo=36153164&szBjId=dailyapril&nBbsNo=18605867',
         'info_dict': {
@@ -70,6 +68,93 @@ class AfreecaTVIE(InfoExtractor):
                 'upload_date': '20160502',
             },
         }],
+        'skip': 'Video is gone',
+    }, {
+        'url': 'http://vod.afreecatv.com/PLAYER/STATION/18650793',
+        'info_dict': {
+            'id': '18650793',
+            'ext': 'mp4',
+            'title': '오늘은 다르다! 쏘님의 우월한 위아래~ 댄스리액션!',
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'uploader': '윈아디',
+            'uploader_id': 'badkids',
+            'duration': 107,
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }, {
+        'url': 'http://vod.afreecatv.com/PLAYER/STATION/10481652',
+        'info_dict': {
+            'id': '10481652',
+            'title': "BJ유트루와 함께하는 '팅커벨 메이크업!'",
+            'thumbnail': 're:^https?://(?:video|st)img.afreecatv.com/.*$',
+            'uploader': 'dailyapril',
+            'uploader_id': 'dailyapril',
+            'duration': 6492,
+        },
+        'playlist_count': 2,
+        'playlist': [{
+            'md5': 'd8b7c174568da61d774ef0203159bf97',
+            'info_dict': {
+                'id': '20160502_c4c62b9d_174361386_1',
+                'ext': 'mp4',
+                'title': "BJ유트루와 함께하는 '팅커벨 메이크업!' (part 1)",
+                'thumbnail': 're:^https?://(?:video|st)img.afreecatv.com/.*$',
+                'uploader': 'dailyapril',
+                'uploader_id': 'dailyapril',
+                'upload_date': '20160502',
+                'duration': 3601,
+            },
+        }, {
+            'md5': '58f2ce7f6044e34439ab2d50612ab02b',
+            'info_dict': {
+                'id': '20160502_39e739bb_174361386_2',
+                'ext': 'mp4',
+                'title': "BJ유트루와 함께하는 '팅커벨 메이크업!' (part 2)",
+                'thumbnail': 're:^https?://(?:video|st)img.afreecatv.com/.*$',
+                'uploader': 'dailyapril',
+                'uploader_id': 'dailyapril',
+                'upload_date': '20160502',
+                'duration': 2891,
+            },
+        }],
+        'params': {
+            'skip_download': True,
+        },
+    }, {
+        # non standard key
+        'url': 'http://vod.afreecatv.com/PLAYER/STATION/20515605',
+        'info_dict': {
+            'id': '20170411_BE689A0E_190960999_1_2_h',
+            'ext': 'mp4',
+            'title': '혼자사는여자집',
+            'thumbnail': 're:^https?://(?:video|st)img.afreecatv.com/.*$',
+            'uploader': '♥이슬이',
+            'uploader_id': 'dasl8121',
+            'upload_date': '20170411',
+            'duration': 213,
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }, {
+        # adult video
+        'url': 'http://vod.afreecatv.com/PLAYER/STATION/26542731',
+        'info_dict': {
+            'id': '20171001_F1AE1711_196617479_1',
+            'ext': 'mp4',
+            'title': '[생]서아 초심 찾기 방송 (part 1)',
+            'thumbnail': 're:^https?://(?:video|st)img.afreecatv.com/.*$',
+            'uploader': 'BJ서아',
+            'uploader_id': 'bjdyrksu',
+            'upload_date': '20171001',
+            'duration': 3600,
+            'age_limit': 18,
+        },
+        'params': {
+            'skip_download': True,
+        },
     }, {
         'url': 'http://www.afreecatv.com/player/Player.swf?szType=szBjId=djleegoon&nStationNo=11273158&nBbsNo=13161095&nTitleNo=36327652',
         'only_matching': True,
@@ -84,42 +169,93 @@ class AfreecaTVIE(InfoExtractor):
         m = re.match(r'^(?P<upload_date>\d{8})_\w+_(?P<part>\d+)$', key)
         if m:
             video_key['upload_date'] = m.group('upload_date')
-            video_key['part'] = m.group('part')
+            video_key['part'] = int(m.group('part'))
         return video_key
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        parsed_url = compat_urllib_parse_urlparse(url)
-        info_url = compat_urlparse.urlunparse(parsed_url._replace(
-            netloc='afbbs.afreecatv.com:8080',
-            path='/api/video/get_video_info.php'))
 
         video_xml = self._download_xml(
-            update_url_query(info_url, {'nTitleNo': video_id}), video_id)
+            'http://afbbs.afreecatv.com:8080/api/video/get_video_info.php',
+            video_id, query={
+                'nTitleNo': video_id,
+                'partialView': 'SKIP_ADULT',
+            })
 
-        if xpath_element(video_xml, './track/video/file') is None:
+        flag = xpath_text(video_xml, './track/flag', 'flag', default=None)
+        if flag and flag != 'SUCCEED':
+            raise ExtractorError(
+                '%s said: %s' % (self.IE_NAME, flag), expected=True)
+
+        video_element = video_xml.findall(compat_xpath('./track/video'))[1]
+        if video_element is None or video_element.text is None:
             raise ExtractorError('Specified AfreecaTV video does not exist',
                                  expected=True)
 
-        title = xpath_text(video_xml, './track/title', 'title')
+        video_url = video_element.text.strip()
+
+        title = xpath_text(video_xml, './track/title', 'title', fatal=True)
+
         uploader = xpath_text(video_xml, './track/nickname', 'uploader')
         uploader_id = xpath_text(video_xml, './track/bj_id', 'uploader id')
-        duration = int_or_none(xpath_text(video_xml, './track/duration',
-                                          'duration'))
+        duration = int_or_none(xpath_text(
+            video_xml, './track/duration', 'duration'))
         thumbnail = xpath_text(video_xml, './track/titleImage', 'thumbnail')
 
-        entries = []
-        for i, video_file in enumerate(video_xml.findall('./track/video/file')):
-            video_key = self.parse_video_key(video_file.get('key', ''))
-            if not video_key:
-                continue
-            entries.append({
-                'id': '%s_%s' % (video_id, video_key.get('part', i + 1)),
-                'title': title,
-                'upload_date': video_key.get('upload_date'),
-                'duration': int_or_none(video_file.get('duration')),
-                'url': video_file.text,
+        common_entry = {
+            'uploader': uploader,
+            'uploader_id': uploader_id,
+            'thumbnail': thumbnail,
+        }
+
+        info = common_entry.copy()
+        info.update({
+            'id': video_id,
+            'title': title,
+            'duration': duration,
+        })
+
+        if not video_url:
+            entries = []
+            file_elements = video_element.findall(compat_xpath('./file'))
+            one = len(file_elements) == 1
+            for file_num, file_element in enumerate(file_elements, start=1):
+                file_url = file_element.text
+                if not file_url:
+                    continue
+                key = file_element.get('key', '')
+                upload_date = self._search_regex(
+                    r'^(\d{8})_', key, 'upload date', default=None)
+                file_duration = int_or_none(file_element.get('duration'))
+                format_id = key if key else '%s_%s' % (video_id, file_num)
+                if determine_ext(file_url) == 'm3u8':
+                    formats = self._extract_m3u8_formats(
+                        file_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                        m3u8_id='hls',
+                        note='Downloading part %d m3u8 information' % file_num)
+                else:
+                    formats = [{
+                        'url': file_url,
+                        'format_id': 'http',
+                    }]
+                if not formats:
+                    continue
+                self._sort_formats(formats)
+                file_info = common_entry.copy()
+                file_info.update({
+                    'id': format_id,
+                    'title': title if one else '%s (part %d)' % (title, file_num),
+                    'upload_date': upload_date,
+                    'duration': file_duration,
+                    'formats': formats,
+                })
+                entries.append(file_info)
+            entries_info = info.copy()
+            entries_info.update({
+                '_type': 'multi_video',
+                'entries': entries,
             })
+            return entries_info
 
         info = {
             'id': video_id,
@@ -130,16 +266,17 @@ class AfreecaTVIE(InfoExtractor):
             'thumbnail': thumbnail,
         }
 
-        if len(entries) > 1:
-            info['_type'] = 'multi_video'
-            info['entries'] = entries
-        elif len(entries) == 1:
-            info['url'] = entries[0]['url']
-            info['upload_date'] = entries[0].get('upload_date')
+        if determine_ext(video_url) == 'm3u8':
+            info['formats'] = self._extract_m3u8_formats(
+                video_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                m3u8_id='hls')
         else:
-            raise ExtractorError(
-                'No files found for the specified AfreecaTV video, either'
-                ' the URL is incorrect or the video has been made private.',
-                expected=True)
+            app, playpath = video_url.split('mp4:')
+            info.update({
+                'url': app,
+                'ext': 'flv',
+                'play_path': 'mp4:' + playpath,
+                'rtmp_live': True,  # downloading won't end without this
+            })
 
         return info

youtube_dl/extractor/airmozilla.py

@@ -15,12 +15,12 @@ class AirMozillaIE(InfoExtractor):
     _VALID_URL = r'https?://air\.mozilla\.org/(?P<id>[0-9a-z-]+)/?'
     _TEST = {
         'url': 'https://air.mozilla.org/privacy-lab-a-meetup-for-privacy-minded-people-in-san-francisco/',
-        'md5': '2e3e7486ba5d180e829d453875b9b8bf',
+        'md5': '8d02f53ee39cf006009180e21df1f3ba',
         'info_dict': {
             'id': '6x4q2w',
             'ext': 'mp4',
             'title': 'Privacy Lab - a meetup for privacy minded people in San Francisco',
-            'thumbnail': r're:https?://vid\.ly/(?P<id>[0-9a-z-]+)/poster',
+            'thumbnail': r're:https?://.*/poster\.jpg',
             'description': 'Brings together privacy professionals and others interested in privacy at for-profits, non-profits, and NGOs in an effort to contribute to the state of the ecosystem...',
             'timestamp': 1422487800,
             'upload_date': '20150128',
@@ -34,21 +34,13 @@ class AirMozillaIE(InfoExtractor):
     def _real_extract(self, url):
         display_id = self._match_id(url)
         webpage = self._download_webpage(url, display_id)
-        video_id = self._html_search_regex(r'//vid.ly/(.*?)/embed', webpage, 'id')
+        video_id = self._html_search_regex(r'//vid\.ly/(.*?)/embed', webpage, 'id')
 
         embed_script = self._download_webpage('https://vid.ly/{0}/embed'.format(video_id), video_id)
-        jwconfig = self._search_regex(r'\svar jwconfig = (\{.*?\});\s', embed_script, 'metadata')
-        metadata = self._parse_json(jwconfig, video_id)
-
-        formats = [{
-            'url': source['file'],
-            'ext': source['type'],
-            'format_id': self._search_regex(r'&format=(.*)$', source['file'], 'video format'),
-            'format': source['label'],
-            'height': int(source['label'].rstrip('p')),
-        } for source in metadata['playlist'][0]['sources']]
-        self._sort_formats(formats)
+        jwconfig = self._parse_json(self._search_regex(
+            r'initCallback\((.*)\);', embed_script, 'metadata'), video_id)['config']
 
+        info_dict = self._parse_jwplayer_data(jwconfig, video_id)
         view_count = int_or_none(self._html_search_regex(
             r'Views since archived: ([0-9]+)',
             webpage, 'view count', fatal=False))
@@ -58,17 +50,17 @@ class AirMozillaIE(InfoExtractor):
             r'Duration:\s*(\d+\s*hours?\s*\d+\s*minutes?)',
             webpage, 'duration', fatal=False))
 
-        return {
+        info_dict.update({
             'id': video_id,
             'title': self._og_search_title(webpage),
-            'formats': formats,
             'url': self._og_search_url(webpage),
             'display_id': display_id,
-            'thumbnail': metadata['playlist'][0].get('image'),
             'description': self._og_search_description(webpage),
             'timestamp': timestamp,
             'location': self._html_search_regex(r'Location: (.*)', webpage, 'location', default=None),
             'duration': duration,
             'view_count': view_count,
             'categories': re.findall(r'<a href=".*?" class="channel">(.*?)</a>', webpage),
-        }
+        })
+
+        return info_dict

youtube_dl/extractor/aliexpress.py

@@ -0,0 +1,53 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..compat import compat_str
+from ..utils import (
+    float_or_none,
+    try_get,
+)
+
+
+class AliExpressLiveIE(InfoExtractor):
+    _VALID_URL = r'https?://live\.aliexpress\.com/live/(?P<id>\d+)'
+    _TEST = {
+        'url': 'https://live.aliexpress.com/live/2800002704436634',
+        'md5': 'e729e25d47c5e557f2630eaf99b740a5',
+        'info_dict': {
+            'id': '2800002704436634',
+            'ext': 'mp4',
+            'title': 'CASIMA7.22',
+            'thumbnail': r're:http://.*\.jpg',
+            'uploader': 'CASIMA Official Store',
+            'timestamp': 1500717600,
+            'upload_date': '20170722',
+        },
+    }
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, video_id)
+
+        data = self._parse_json(
+            self._search_regex(
+                r'(?s)runParams\s*=\s*({.+?})\s*;?\s*var',
+                webpage, 'runParams'),
+            video_id)
+
+        title = data['title']
+
+        formats = self._extract_m3u8_formats(
+            data['replyStreamUrl'], video_id, 'mp4',
+            entry_protocol='m3u8_native', m3u8_id='hls')
+
+        return {
+            'id': video_id,
+            'title': title,
+            'thumbnail': data.get('coverUrl'),
+            'uploader': try_get(
+                data, lambda x: x['followBar']['name'], compat_str),
+            'timestamp': float_or_none(data.get('startTimeLong'), scale=1000),
+            'formats': formats,
+        }

youtube_dl/extractor/aljazeera.py

@@ -4,9 +4,9 @@ from .common import InfoExtractor
 
 
 class AlJazeeraIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?aljazeera\.com/programmes/.*?/(?P<id>[^/]+)\.html'
+    _VALID_URL = r'https?://(?:www\.)?aljazeera\.com/(?:programmes|video)/.*?/(?P<id>[^/]+)\.html'
 
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.aljazeera.com/programmes/the-slum/2014/08/deliverance-201482883754237240.html',
         'info_dict': {
             'id': '3792260579001',
@@ -19,7 +19,10 @@ class AlJazeeraIE(InfoExtractor):
         },
         'add_ie': ['BrightcoveNew'],
         'skip': 'Not accessible from Travis CI server',
-    }
+    }, {
+        'url': 'http://www.aljazeera.com/video/news/2017/05/sierra-leone-709-carat-diamond-auctioned-170511100111930.html',
+        'only_matching': True,
+    }]
     BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/665003303001/default_default/index.html?videoId=%s'
 
     def _real_extract(self, url):

youtube_dl/extractor/allocine.py

@@ -2,9 +2,13 @@
 from __future__ import unicode_literals
 
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
-    remove_end,
+    int_or_none,
     qualities,
+    remove_end,
+    try_get,
+    unified_timestamp,
     url_basename,
 )
 
@@ -22,6 +26,10 @@ class AllocineIE(InfoExtractor):
             'title': 'Astérix - Le Domaine des Dieux Teaser VF',
             'description': 'md5:4a754271d9c6f16c72629a8a993ee884',
             'thumbnail': r're:http://.*\.jpg',
+            'duration': 39,
+            'timestamp': 1404273600,
+            'upload_date': '20140702',
+            'view_count': int,
         },
     }, {
         'url': 'http://www.allocine.fr/video/player_gen_cmedia=19540403&cfilm=222257.html',
@@ -33,6 +41,10 @@ class AllocineIE(InfoExtractor):
             'title': 'Planes 2 Bande-annonce VF',
             'description': 'Regardez la bande annonce du film Planes 2 (Planes 2 Bande-annonce VF). Planes 2, un film de Roberts Gannaway',
             'thumbnail': r're:http://.*\.jpg',
+            'duration': 69,
+            'timestamp': 1385659800,
+            'upload_date': '20131128',
+            'view_count': int,
         },
     }, {
         'url': 'http://www.allocine.fr/video/player_gen_cmedia=19544709&cfilm=181290.html',
@@ -44,6 +56,10 @@ class AllocineIE(InfoExtractor):
             'title': 'Dragons 2 - Bande annonce finale VF',
             'description': 'md5:6cdd2d7c2687d4c6aafe80a35e17267a',
             'thumbnail': r're:http://.*\.jpg',
+            'duration': 144,
+            'timestamp': 1397589900,
+            'upload_date': '20140415',
+            'view_count': int,
         },
     }, {
         'url': 'http://www.allocine.fr/video/video-19550147/',
@@ -69,34 +85,37 @@ class AllocineIE(InfoExtractor):
             r'data-model="([^"]+)"', webpage, 'data model', default=None)
         if model:
             model_data = self._parse_json(model, display_id)
-
-            for video_url in model_data['sources'].values():
+            video = model_data['videos'][0]
+            title = video['title']
+            for video_url in video['sources'].values():
                 video_id, format_id = url_basename(video_url).split('_')[:2]
                 formats.append({
                     'format_id': format_id,
                     'quality': quality(format_id),
                     'url': video_url,
                 })
-
-            title = model_data['title']
+            duration = int_or_none(video.get('duration'))
+            view_count = int_or_none(video.get('view_count'))
+            timestamp = unified_timestamp(try_get(
+                video, lambda x: x['added_at']['date'], compat_str))
         else:
             video_id = display_id
             media_data = self._download_json(
                 'http://www.allocine.fr/ws/AcVisiondataV5.ashx?media=%s' % video_id, display_id)
+            title = remove_end(
+                self._html_search_regex(
+                    r'(?s)<title>(.+?)</title>', webpage, 'title').strip(),
+                ' - AlloCiné')
             for key, value in media_data['video'].items():
                 if not key.endswith('Path'):
                     continue
-
                 format_id = key[:-len('Path')]
                 formats.append({
                     'format_id': format_id,
                     'quality': quality(format_id),
                     'url': value,
                 })
-
-            title = remove_end(self._html_search_regex(
-                r'(?s)<title>(.+?)</title>', webpage, 'title'
-            ).strip(), ' - AlloCiné')
+            duration, view_count, timestamp = [None] * 3
 
         self._sort_formats(formats)
 
@@ -104,7 +123,10 @@ class AllocineIE(InfoExtractor):
             'id': video_id,
             'display_id': display_id,
             'title': title,
-            'thumbnail': self._og_search_thumbnail(webpage),
-            'formats': formats,
             'description': self._og_search_description(webpage),
+            'thumbnail': self._og_search_thumbnail(webpage),
+            'duration': duration,
+            'timestamp': timestamp,
+            'view_count': view_count,
+            'formats': formats,
         }

youtube_dl/extractor/amcnetworks.py

@@ -3,14 +3,15 @@ from __future__ import unicode_literals
 
 from .theplatform import ThePlatformIE
 from ..utils import (
-    update_url_query,
-    parse_age_limit,
     int_or_none,
+    parse_age_limit,
+    try_get,
+    update_url_query,
 )
 
 
 class AMCNetworksIE(ThePlatformIE):
-    _VALID_URL = r'https?://(?:www\.)?(?:amc|bbcamerica|ifc|wetv)\.com/(?:movies/|shows/[^/]+/(?:full-episodes/)?[^/]+/episode-\d+(?:-(?:[^/]+/)?|/))(?P<id>[^/?#]+)'
+    _VALID_URL = r'https?://(?:www\.)?(?:amc|bbcamerica|ifc|wetv)\.com/(?:movies|shows(?:/[^/]+)+)/(?P<id>[^/?#]+)'
     _TESTS = [{
         'url': 'http://www.ifc.com/shows/maron/season-04/episode-01/step-1',
         'md5': '',
@@ -44,6 +45,12 @@ class AMCNetworksIE(ThePlatformIE):
     }, {
         'url': 'http://www.bbcamerica.com/shows/doctor-who/full-episodes/the-power-of-the-daleks/episode-01-episode-1-color-version',
         'only_matching': True,
+    }, {
+        'url': 'http://www.wetv.com/shows/mama-june-from-not-to-hot/full-episode/season-01/thin-tervention',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.wetv.com/shows/la-hair/videos/season-05/episode-09-episode-9-2/episode-9-sneak-peek-3',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
@@ -53,20 +60,31 @@ class AMCNetworksIE(ThePlatformIE):
             'mbr': 'true',
             'manifest': 'm3u',
         }
-        media_url = self._search_regex(r'window\.platformLinkURL\s*=\s*[\'"]([^\'"]+)', webpage, 'media url')
+        media_url = self._search_regex(
+            r'window\.platformLinkURL\s*=\s*[\'"]([^\'"]+)',
+            webpage, 'media url')
         theplatform_metadata = self._download_theplatform_metadata(self._search_regex(
-            r'https?://link.theplatform.com/s/([^?]+)', media_url, 'theplatform_path'), display_id)
+            r'link\.theplatform\.com/s/([^?]+)',
+            media_url, 'theplatform_path'), display_id)
         info = self._parse_theplatform_metadata(theplatform_metadata)
         video_id = theplatform_metadata['pid']
         title = theplatform_metadata['title']
-        rating = theplatform_metadata['ratings'][0]['rating']
-        auth_required = self._search_regex(r'window\.authRequired\s*=\s*(true|false);', webpage, 'auth required')
+        rating = try_get(
+            theplatform_metadata, lambda x: x['ratings'][0]['rating'])
+        auth_required = self._search_regex(
+            r'window\.authRequired\s*=\s*(true|false);',
+            webpage, 'auth required')
         if auth_required == 'true':
-            requestor_id = self._search_regex(r'window\.requestor_id\s*=\s*[\'"]([^\'"]+)', webpage, 'requestor id')
-            resource = self._get_mvpd_resource(requestor_id, title, video_id, rating)
-            query['auth'] = self._extract_mvpd_auth(url, video_id, requestor_id, resource)
+            requestor_id = self._search_regex(
+                r'window\.requestor_id\s*=\s*[\'"]([^\'"]+)',
+                webpage, 'requestor id')
+            resource = self._get_mvpd_resource(
+                requestor_id, title, video_id, rating)
+            query['auth'] = self._extract_mvpd_auth(
+                url, video_id, requestor_id, resource)
         media_url = update_url_query(media_url, query)
-        formats, subtitles = self._extract_theplatform_smil(media_url, video_id)
+        formats, subtitles = self._extract_theplatform_smil(
+            media_url, video_id)
         self._sort_formats(formats)
         info.update({
             'id': video_id,
@@ -78,9 +96,11 @@ class AMCNetworksIE(ThePlatformIE):
         if ns_keys:
             ns = list(ns_keys)[0]
             series = theplatform_metadata.get(ns + '$show')
-            season_number = int_or_none(theplatform_metadata.get(ns + '$season'))
+            season_number = int_or_none(
+                theplatform_metadata.get(ns + '$season'))
             episode = theplatform_metadata.get(ns + '$episodeTitle')
-            episode_number = int_or_none(theplatform_metadata.get(ns + '$episode'))
+            episode_number = int_or_none(
+                theplatform_metadata.get(ns + '$episode'))
             if season_number:
                 title = 'Season %d - %s' % (season_number, title)
             if series:

youtube_dl/extractor/americastestkitchen.py

@@ -0,0 +1,85 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    clean_html,
+    int_or_none,
+    try_get,
+    unified_strdate,
+)
+
+
+class AmericasTestKitchenIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?americastestkitchen\.com/(?:episode|videos)/(?P<id>\d+)'
+    _TESTS = [{
+        'url': 'https://www.americastestkitchen.com/episode/548-summer-dinner-party',
+        'md5': 'b861c3e365ac38ad319cfd509c30577f',
+        'info_dict': {
+            'id': '1_5g5zua6e',
+            'title': 'Summer Dinner Party',
+            'ext': 'mp4',
+            'description': 'md5:858d986e73a4826979b6a5d9f8f6a1ec',
+            'thumbnail': r're:^https?://.*\.jpg',
+            'timestamp': 1497285541,
+            'upload_date': '20170612',
+            'uploader_id': 'roger.metcalf@americastestkitchen.com',
+            'release_date': '20170617',
+            'series': "America's Test Kitchen",
+            'season_number': 17,
+            'episode': 'Summer Dinner Party',
+            'episode_number': 24,
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }, {
+        'url': 'https://www.americastestkitchen.com/videos/3420-pan-seared-salmon',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, video_id)
+
+        partner_id = self._search_regex(
+            r'src=["\'](?:https?:)?//(?:[^/]+\.)kaltura\.com/(?:[^/]+/)*(?:p|partner_id)/(\d+)',
+            webpage, 'kaltura partner id')
+
+        video_data = self._parse_json(
+            self._search_regex(
+                r'window\.__INITIAL_STATE__\s*=\s*({.+?})\s*;\s*</script>',
+                webpage, 'initial context'),
+            video_id)
+
+        ep_data = try_get(
+            video_data,
+            (lambda x: x['episodeDetail']['content']['data'],
+             lambda x: x['videoDetail']['content']['data']), dict)
+        ep_meta = ep_data.get('full_video', {})
+        external_id = ep_data.get('external_id') or ep_meta['external_id']
+
+        title = ep_data.get('title') or ep_meta.get('title')
+        description = clean_html(ep_meta.get('episode_description') or ep_data.get(
+            'description') or ep_meta.get('description'))
+        thumbnail = try_get(ep_meta, lambda x: x['photo']['image_url'])
+        release_date = unified_strdate(ep_data.get('aired_at'))
+
+        season_number = int_or_none(ep_meta.get('season_number'))
+        episode = ep_meta.get('title')
+        episode_number = int_or_none(ep_meta.get('episode_number'))
+
+        return {
+            '_type': 'url_transparent',
+            'url': 'kaltura:%s:%s' % (partner_id, external_id),
+            'ie_key': 'Kaltura',
+            'title': title,
+            'description': description,
+            'thumbnail': thumbnail,
+            'release_date': release_date,
+            'series': "America's Test Kitchen",
+            'season_number': season_number,
+            'episode': episode,
+            'episode_number': episode_number,
+        }

youtube_dl/extractor/amp.py

@@ -7,15 +7,19 @@ from ..utils import (
     parse_iso8601,
     mimetype2ext,
     determine_ext,
+    ExtractorError,
 )
 
 
 class AMPIE(InfoExtractor):
     # parse Akamai Adaptive Media Player feed
     def _extract_feed_info(self, url):
-        item = self._download_json(
+        feed = self._download_json(
             url, None, 'Downloading Akamai AMP feed',
-            'Unable to download Akamai AMP feed')['channel']['item']
+            'Unable to download Akamai AMP feed')
+        item = feed.get('channel', {}).get('item')
+        if not item:
+            raise ExtractorError('%s said: %s' % (self.IE_NAME, feed['error']))
 
         video_id = item['guid']
 
@@ -30,9 +34,12 @@ class AMPIE(InfoExtractor):
             if isinstance(media_thumbnail, dict):
                 media_thumbnail = [media_thumbnail]
             for thumbnail_data in media_thumbnail:
-                thumbnail = thumbnail_data['@attributes']
+                thumbnail = thumbnail_data.get('@attributes', {})
+                thumbnail_url = thumbnail.get('url')
+                if not thumbnail_url:
+                    continue
                 thumbnails.append({
-                    'url': self._proto_relative_url(thumbnail['url'], 'http:'),
+                    'url': self._proto_relative_url(thumbnail_url, 'http:'),
                     'width': int_or_none(thumbnail.get('width')),
                     'height': int_or_none(thumbnail.get('height')),
                 })
@@ -43,9 +50,14 @@ class AMPIE(InfoExtractor):
             if isinstance(media_subtitle, dict):
                 media_subtitle = [media_subtitle]
             for subtitle_data in media_subtitle:
-                subtitle = subtitle_data['@attributes']
-                lang = subtitle.get('lang') or 'en'
-                subtitles[lang] = [{'url': subtitle['href']}]
+                subtitle = subtitle_data.get('@attributes', {})
+                subtitle_href = subtitle.get('href')
+                if not subtitle_href:
+                    continue
+                subtitles.setdefault(subtitle.get('lang') or 'en', []).append({
+                    'url': subtitle_href,
+                    'ext': mimetype2ext(subtitle.get('type')) or determine_ext(subtitle_href),
+                })
 
         formats = []
         media_content = get_media_node('content')

youtube_dl/extractor/animeondemand.py

@@ -3,16 +3,13 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import (
-    compat_urlparse,
-    compat_str,
-)
+from ..compat import compat_str
 from ..utils import (
     determine_ext,
     extract_attributes,
     ExtractorError,
-    sanitized_Request,
     urlencode_postdata,
+    urljoin,
 )
 
 
@@ -21,6 +18,8 @@ class AnimeOnDemandIE(InfoExtractor):
     _LOGIN_URL = 'https://www.anime-on-demand.de/users/sign_in'
     _APPLY_HTML5_URL = 'https://www.anime-on-demand.de/html5apply'
     _NETRC_MACHINE = 'animeondemand'
+    # German-speaking countries of Europe
+    _GEO_COUNTRIES = ['AT', 'CH', 'DE', 'LI', 'LU']
     _TESTS = [{
         # jap, OmU
         'url': 'https://www.anime-on-demand.de/anime/161',
@@ -46,6 +45,10 @@ class AnimeOnDemandIE(InfoExtractor):
         # Full length film, non-series, ger/jap, Dub/OmU, account required
         'url': 'https://www.anime-on-demand.de/anime/185',
         'only_matching': True,
+    }, {
+        # Flash videos
+        'url': 'https://www.anime-on-demand.de/anime/12',
+        'only_matching': True,
     }]
 
     def _login(self):
@@ -72,19 +75,18 @@ class AnimeOnDemandIE(InfoExtractor):
             'post url', default=self._LOGIN_URL, group='url')
 
         if not post_url.startswith('http'):
-            post_url = compat_urlparse.urljoin(self._LOGIN_URL, post_url)
-
-        request = sanitized_Request(
-            post_url, urlencode_postdata(login_form))
-        request.add_header('Referer', self._LOGIN_URL)
+            post_url = urljoin(self._LOGIN_URL, post_url)
 
         response = self._download_webpage(
-            request, None, 'Logging in as %s' % username)
+            post_url, None, 'Logging in',
+            data=urlencode_postdata(login_form), headers={
+                'Referer': self._LOGIN_URL,
+            })
 
         if all(p not in response for p in ('>Logout<', 'href="/users/sign_out"')):
             error = self._search_regex(
-                r'<p class="alert alert-danger">(.+?)</p>',
-                response, 'error', default=None)
+                r'<p[^>]+\bclass=(["\'])(?:(?!\1).)*\balert\b(?:(?!\1).)*\1[^>]*>(?P<error>.+?)</p>',
+                response, 'error', default=None, group='error')
             if error:
                 raise ExtractorError('Unable to login: %s' % error, expected=True)
             raise ExtractorError('Unable to log in')
@@ -120,10 +122,11 @@ class AnimeOnDemandIE(InfoExtractor):
             formats = []
 
             for input_ in re.findall(
-                    r'<input[^>]+class=["\'].*?streamstarter_html5[^>]+>', html):
+                    r'<input[^>]+class=["\'].*?streamstarter[^>]+>', html):
                 attributes = extract_attributes(input_)
+                title = attributes.get('data-dialog-header')
                 playlist_urls = []
-                for playlist_key in ('data-playlist', 'data-otherplaylist'):
+                for playlist_key in ('data-playlist', 'data-otherplaylist', 'data-stream'):
                     playlist_url = attributes.get(playlist_key)
                     if isinstance(playlist_url, compat_str) and re.match(
                             r'/?[\da-zA-Z]+', playlist_url):
@@ -147,19 +150,38 @@ class AnimeOnDemandIE(InfoExtractor):
                         format_id_list.append(compat_str(num))
                     format_id = '-'.join(format_id_list)
                     format_note = ', '.join(filter(None, (kind, lang_note)))
-                    request = sanitized_Request(
-                        compat_urlparse.urljoin(url, playlist_url),
+                    item_id_list = []
+                    if format_id:
+                        item_id_list.append(format_id)
+                    item_id_list.append('videomaterial')
+                    playlist = self._download_json(
+                        urljoin(url, playlist_url), video_id,
+                        'Downloading %s JSON' % ' '.join(item_id_list),
                         headers={
                             'X-Requested-With': 'XMLHttpRequest',
                             'X-CSRF-Token': csrf_token,
                             'Referer': url,
                             'Accept': 'application/json, text/javascript, */*; q=0.01',
-                        })
-                    playlist = self._download_json(
-                        request, video_id, 'Downloading %s playlist JSON' % format_id,
-                        fatal=False)
+                        }, fatal=False)
                     if not playlist:
                         continue
+                    stream_url = playlist.get('streamurl')
+                    if stream_url:
+                        rtmp = re.search(
+                            r'^(?P<url>rtmpe?://(?P<host>[^/]+)/(?P<app>.+/))(?P<playpath>mp[34]:.+)',
+                            stream_url)
+                        if rtmp:
+                            formats.append({
+                                'url': rtmp.group('url'),
+                                'app': rtmp.group('app'),
+                                'play_path': rtmp.group('playpath'),
+                                'page_url': url,
+                                'player_url': 'https://www.anime-on-demand.de/assets/jwplayer.flash-55abfb34080700304d49125ce9ffb4a6.swf',
+                                'rtmp_real_time': True,
+                                'format_id': 'rtmp',
+                                'ext': 'flv',
+                            })
+                            continue
                     start_video = playlist.get('startvideo', 0)
                     playlist = playlist.get('playlist')
                     if not playlist or not isinstance(playlist, list):
@@ -222,7 +244,7 @@ class AnimeOnDemandIE(InfoExtractor):
                     f.update({
                         'id': '%s-%s' % (f['id'], m.group('kind').lower()),
                         'title': m.group('title'),
-                        'url': compat_urlparse.urljoin(url, m.group('href')),
+                        'url': urljoin(url, m.group('href')),
                     })
                     entries.append(f)
 

youtube_dl/extractor/anvato.py

@@ -5,6 +5,7 @@ import base64
 import hashlib
 import json
 import random
+import re
 import time
 
 from .common import InfoExtractor
@@ -16,6 +17,8 @@ from ..utils import (
     intlist_to_bytes,
     int_or_none,
     strip_jsonp,
+    unescapeHTML,
+    unsmuggle_url,
 )
 
 
@@ -26,6 +29,8 @@ def md5_text(s):
 
 
 class AnvatoIE(InfoExtractor):
+    _VALID_URL = r'anvato:(?P<access_key_or_mcp>[^:]+):(?P<id>\d+)'
+
     # Copied from anvplayer.min.js
     _ANVACK_TABLE = {
         'nbcu_nbcd_desktop_web_prod_93d8ead38ce2024f8f544b78306fbd15895ae5e6': 'NNemUkySjxLyPTKvZRiGntBIjEyK8uqicjMakIaQ',
@@ -114,6 +119,22 @@ class AnvatoIE(InfoExtractor):
         'nbcu_nbcd_desktop_web_prod_93d8ead38ce2024f8f544b78306fbd15895ae5e6_secure': 'NNemUkySjxLyPTKvZRiGntBIjEyK8uqicjMakIaQ'
     }
 
+    _MCP_TO_ACCESS_KEY_TABLE = {
+        'qa': 'anvato_mcpqa_demo_web_stage_18b55e00db5a13faa8d03ae6e41f6f5bcb15b922',
+        'lin': 'anvato_mcp_lin_web_prod_4c36fbfd4d8d8ecae6488656e21ac6d1ac972749',
+        'univison': 'anvato_mcp_univision_web_prod_37fe34850c99a3b5cdb71dab10a417dd5cdecafa',
+        'uni': 'anvato_mcp_univision_web_prod_37fe34850c99a3b5cdb71dab10a417dd5cdecafa',
+        'dev': 'anvato_mcp_fs2go_web_prod_c7b90a93e171469cdca00a931211a2f556370d0a',
+        'sps': 'anvato_mcp_sps_web_prod_54bdc90dd6ba21710e9f7074338365bba28da336',
+        'spsstg': 'anvato_mcp_sps_web_prod_54bdc90dd6ba21710e9f7074338365bba28da336',
+        'anv': 'anvato_mcp_anv_web_prod_791407490f4c1ef2a4bcb21103e0cb1bcb3352b3',
+        'gray': 'anvato_mcp_gray_web_prod_4c10f067c393ed8fc453d3930f8ab2b159973900',
+        'hearst': 'anvato_mcp_hearst_web_prod_5356c3de0fc7c90a3727b4863ca7fec3a4524a99',
+        'cbs': 'anvato_mcp_cbs_web_prod_02f26581ff80e5bda7aad28226a8d369037f2cbe',
+        'telemundo': 'anvato_mcp_telemundo_web_prod_c5278d51ad46fda4b6ca3d0ea44a7846a054f582'
+    }
+
+    _ANVP_RE = r'<script[^>]+\bdata-anvp\s*=\s*(["\'])(?P<anvp>(?:(?!\1).)+)\1'
     _AUTH_KEY = b'\x31\xc2\x42\x84\x9e\x73\xa0\xce'
 
     def __init__(self, *args, **kwargs):
@@ -177,17 +198,16 @@ class AnvatoIE(InfoExtractor):
                 'tbr': tbr if tbr != 0 else None,
             }
 
-            if ext == 'm3u8' or media_format in ('m3u8', 'm3u8-variant'):
-                # Not using _extract_m3u8_formats here as individual media
-                # playlists are also included in published_urls.
-                if tbr is None:
-                    formats.append(self._m3u8_meta_format(video_url, ext='mp4', m3u8_id='hls'))
-                    continue
-                else:
-                    a_format.update({
-                        'format_id': '-'.join(filter(None, ['hls', compat_str(tbr)])),
-                        'ext': 'mp4',
-                    })
+            if media_format == 'm3u8' and tbr is not None:
+                a_format.update({
+                    'format_id': '-'.join(filter(None, ['hls', compat_str(tbr)])),
+                    'ext': 'mp4',
+                })
+            elif media_format == 'm3u8-variant' or ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    video_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    m3u8_id='hls', fatal=False))
+                continue
             elif ext == 'mp3' or media_format == 'mp3':
                 a_format['vcodec'] = 'none'
             else:
@@ -222,9 +242,45 @@ class AnvatoIE(InfoExtractor):
             'subtitles': subtitles,
         }
 
+    @staticmethod
+    def _extract_urls(ie, webpage, video_id):
+        entries = []
+        for mobj in re.finditer(AnvatoIE._ANVP_RE, webpage):
+            anvplayer_data = ie._parse_json(
+                mobj.group('anvp'), video_id, transform_source=unescapeHTML,
+                fatal=False)
+            if not anvplayer_data:
+                continue
+            video = anvplayer_data.get('video')
+            if not isinstance(video, compat_str) or not video.isdigit():
+                continue
+            access_key = anvplayer_data.get('accessKey')
+            if not access_key:
+                mcp = anvplayer_data.get('mcp')
+                if mcp:
+                    access_key = AnvatoIE._MCP_TO_ACCESS_KEY_TABLE.get(
+                        mcp.lower())
+            if not access_key:
+                continue
+            entries.append(ie.url_result(
+                'anvato:%s:%s' % (access_key, video), ie=AnvatoIE.ie_key(),
+                video_id=video))
+        return entries
+
     def _extract_anvato_videos(self, webpage, video_id):
-        anvplayer_data = self._parse_json(self._html_search_regex(
-            r'<script[^>]+data-anvp=\'([^\']+)\'', webpage,
-            'Anvato player data'), video_id)
+        anvplayer_data = self._parse_json(
+            self._html_search_regex(
+                self._ANVP_RE, webpage, 'Anvato player data', group='anvp'),
+            video_id)
         return self._get_anvato_videos(
             anvplayer_data['accessKey'], anvplayer_data['video'])
+
+    def _real_extract(self, url):
+        url, smuggled_data = unsmuggle_url(url, {})
+        self._initialize_geo_bypass(smuggled_data.get('geo_countries'))
+
+        mobj = re.match(self._VALID_URL, url)
+        access_key, video_id = mobj.group('access_key_or_mcp', 'id')
+        if access_key not in self._ANVACK_TABLE:
+            access_key = self._MCP_TO_ACCESS_KEY_TABLE[access_key]
+        return self._get_anvato_videos(access_key, video_id)

youtube_dl/extractor/aparat.py

@@ -3,13 +3,13 @@ from __future__ import unicode_literals
 
 from .common import InfoExtractor
 from ..utils import (
-    ExtractorError,
-    HEADRequest,
+    int_or_none,
+    mimetype2ext,
 )
 
 
 class AparatIE(InfoExtractor):
-    _VALID_URL = r'^https?://(?:www\.)?aparat\.com/(?:v/|video/video/embed/videohash/)(?P<id>[a-zA-Z0-9]+)'
+    _VALID_URL = r'https?://(?:www\.)?aparat\.com/(?:v/|video/video/embed/videohash/)(?P<id>[a-zA-Z0-9]+)'
 
     _TEST = {
         'url': 'http://www.aparat.com/v/wP8On',
@@ -29,30 +29,41 @@ class AparatIE(InfoExtractor):
         # Note: There is an easier-to-parse configuration at
         # http://www.aparat.com/video/video/config/videohash/%video_id
         # but the URL in there does not work
-        embed_url = 'http://www.aparat.com/video/video/embed/vt/frame/showvideo/yes/videohash/' + video_id
-        webpage = self._download_webpage(embed_url, video_id)
-
-        file_list = self._parse_json(self._search_regex(
-            r'fileList\s*=\s*JSON\.parse\(\'([^\']+)\'\)', webpage, 'file list'), video_id)
-        for i, item in enumerate(file_list[0]):
-            video_url = item['file']
-            req = HEADRequest(video_url)
-            res = self._request_webpage(
-                req, video_id, note='Testing video URL %d' % i, errnote=False)
-            if res:
-                break
-        else:
-            raise ExtractorError('No working video URLs found')
+        webpage = self._download_webpage(
+            'http://www.aparat.com/video/video/embed/vt/frame/showvideo/yes/videohash/' + video_id,
+            video_id)
 
         title = self._search_regex(r'\s+title:\s*"([^"]+)"', webpage, 'title')
+
+        file_list = self._parse_json(
+            self._search_regex(
+                r'fileList\s*=\s*JSON\.parse\(\'([^\']+)\'\)', webpage,
+                'file list'),
+            video_id)
+
+        formats = []
+        for item in file_list[0]:
+            file_url = item.get('file')
+            if not file_url:
+                continue
+            ext = mimetype2ext(item.get('type'))
+            label = item.get('label')
+            formats.append({
+                'url': file_url,
+                'ext': ext,
+                'format_id': label or ext,
+                'height': int_or_none(self._search_regex(
+                    r'(\d+)[pP]', label or '', 'height', default=None)),
+            })
+        self._sort_formats(formats)
+
         thumbnail = self._search_regex(
             r'image:\s*"([^"]+)"', webpage, 'thumbnail', fatal=False)
 
         return {
             'id': video_id,
             'title': title,
-            'url': video_url,
-            'ext': 'mp4',
             'thumbnail': thumbnail,
             'age_limit': self._family_friendly_search(webpage),
+            'formats': formats,
         }

youtube_dl/extractor/appleconnect.py

@@ -12,13 +12,13 @@ class AppleConnectIE(InfoExtractor):
     _VALID_URL = r'https?://itunes\.apple\.com/\w{0,2}/?post/idsa\.(?P<id>[\w-]+)'
     _TEST = {
         'url': 'https://itunes.apple.com/us/post/idsa.4ab17a39-2720-11e5-96c5-a5b38f6c42d3',
-        'md5': '10d0f2799111df4cb1c924520ca78f98',
+        'md5': 'e7c38568a01ea45402570e6029206723',
         'info_dict': {
             'id': '4ab17a39-2720-11e5-96c5-a5b38f6c42d3',
             'ext': 'm4v',
             'title': 'Energy',
             'uploader': 'Drake',
-            'thumbnail': 'http://is5.mzstatic.com/image/thumb/Video5/v4/78/61/c5/7861c5fa-ad6d-294b-1464-cf7605b911d6/source/1920x1080sr.jpg',
+            'thumbnail': r're:^https?://.*\.jpg$',
             'upload_date': '20150710',
             'timestamp': 1436545535,
         },

youtube_dl/extractor/appletrailers.py

@@ -70,7 +70,8 @@ class AppleTrailersIE(InfoExtractor):
     }, {
         'url': 'http://trailers.apple.com/trailers/magnolia/blackthorn/',
         'info_dict': {
-            'id': 'blackthorn',
+            'id': '4489',
+            'title': 'Blackthorn',
         },
         'playlist_mincount': 2,
         'expected_warnings': ['Unable to download JSON metadata'],
@@ -116,7 +117,7 @@ class AppleTrailersIE(InfoExtractor):
                             continue
                         formats.append({
                             'format_id': '%s-%s' % (version, size),
-                            'url': re.sub(r'_(\d+p.mov)', r'_h\1', src),
+                            'url': re.sub(r'_(\d+p\.mov)', r'_h\1', src),
                             'width': int_or_none(size_data.get('width')),
                             'height': int_or_none(size_data.get('height')),
                             'language': version[:2],
@@ -178,7 +179,7 @@ class AppleTrailersIE(InfoExtractor):
             formats = []
             for format in settings['metadata']['sizes']:
                 # The src is a file pointing to the real video file
-                format_url = re.sub(r'_(\d*p.mov)', r'_h\1', format['src'])
+                format_url = re.sub(r'_(\d*p\.mov)', r'_h\1', format['src'])
                 formats.append({
                     'url': format_url,
                     'format': format['type'],
@@ -261,7 +262,7 @@ class AppleTrailersSectionIE(InfoExtractor):
             'title': 'Most Popular',
             'id': 'mostpopular',
         },
-        'playlist_mincount': 80,
+        'playlist_mincount': 30,
     }, {
         'url': 'http://trailers.apple.com/#section=moviestudios',
         'info_dict': {

youtube_dl/extractor/archiveorg.py

@@ -1,13 +1,13 @@
 from __future__ import unicode_literals
 
-from .jwplatform import JWPlatformBaseIE
+from .common import InfoExtractor
 from ..utils import (
     unified_strdate,
     clean_html,
 )
 
 
-class ArchiveOrgIE(JWPlatformBaseIE):
+class ArchiveOrgIE(InfoExtractor):
     IE_NAME = 'archive.org'
     IE_DESC = 'archive.org videos'
     _VALID_URL = r'https?://(?:www\.)?archive\.org/(?:details|embed)/(?P<id>[^/?#]+)(?:[?].*)?$'
@@ -24,12 +24,12 @@ class ArchiveOrgIE(JWPlatformBaseIE):
         }
     }, {
         'url': 'https://archive.org/details/Cops1922',
-        'md5': 'bc73c8ab3838b5a8fc6c6651fa7b58ba',
+        'md5': '0869000b4ce265e8ca62738b336b268a',
         'info_dict': {
             'id': 'Cops1922',
             'ext': 'mp4',
             'title': 'Buster Keaton\'s "Cops" (1922)',
-            'description': 'md5:b4544662605877edd99df22f9620d858',
+            'description': 'md5:89e7c77bf5d965dd5c0372cfb49470f6',
         }
     }, {
         'url': 'http://archive.org/embed/XD300-23_68HighlightsAResearchCntAugHumanIntellect',

youtube_dl/extractor/ard.py

@@ -5,6 +5,7 @@ import re
 
 from .common import InfoExtractor
 from .generic import GenericIE
+from ..compat import compat_str
 from ..utils import (
     determine_ext,
     ExtractorError,
@@ -93,6 +94,7 @@ class ARDMediathekIE(InfoExtractor):
 
         duration = int_or_none(media_info.get('_duration'))
         thumbnail = media_info.get('_previewImage')
+        is_live = media_info.get('_isLive') is True
 
         subtitles = {}
         subtitle_url = media_info.get('_subtitleUrl')
@@ -106,6 +108,7 @@ class ARDMediathekIE(InfoExtractor):
             'id': video_id,
             'duration': duration,
             'thumbnail': thumbnail,
+            'is_live': is_live,
             'formats': formats,
             'subtitles': subtitles,
         }
@@ -124,6 +127,8 @@ class ARDMediathekIE(InfoExtractor):
                 quality = stream.get('_quality')
                 server = stream.get('_server')
                 for stream_url in stream_urls:
+                    if not isinstance(stream_url, compat_str) or '//' not in stream_url:
+                        continue
                     ext = determine_ext(stream_url)
                     if quality != 'auto' and ext in ('f4m', 'm3u8'):
                         continue
@@ -144,13 +149,11 @@ class ARDMediathekIE(InfoExtractor):
                                 'play_path': stream_url,
                                 'format_id': 'a%s-rtmp-%s' % (num, quality),
                             }
-                        elif stream_url.startswith('http'):
+                        else:
                             f = {
                                 'url': stream_url,
                                 'format_id': 'a%s-%s-%s' % (num, ext, quality)
                             }
-                        else:
-                            continue
                         m = re.search(r'_(?P<width>\d+)x(?P<height>\d+)\.mp4$', stream_url)
                         if m:
                             f.update({
@@ -166,9 +169,11 @@ class ARDMediathekIE(InfoExtractor):
         # determine video id from url
         m = re.match(self._VALID_URL, url)
 
+        document_id = None
+
         numid = re.search(r'documentId=([0-9]+)', url)
         if numid:
-            video_id = numid.group(1)
+            document_id = video_id = numid.group(1)
         else:
             video_id = m.group('video_id')
 
@@ -191,7 +196,7 @@ class ARDMediathekIE(InfoExtractor):
 
         title = self._html_search_regex(
             [r'<h1(?:\s+class="boxTopHeadline")?>(.*?)</h1>',
-             r'<meta name="dcterms.title" content="(.*?)"/>',
+             r'<meta name="dcterms\.title" content="(.*?)"/>',
              r'<h4 class="headline">(.*?)</h4>'],
             webpage, 'title')
         description = self._html_search_meta(
@@ -228,12 +233,16 @@ class ARDMediathekIE(InfoExtractor):
                 'formats': formats,
             }
         else:  # request JSON file
+            if not document_id:
+                video_id = self._search_regex(
+                    r'/play/(?:config|media)/(\d+)', webpage, 'media id')
             info = self._extract_media_info(
-                'http://www.ardmediathek.de/play/media/%s' % video_id, webpage, video_id)
+                'http://www.ardmediathek.de/play/media/%s' % video_id,
+                webpage, video_id)
 
         info.update({
             'id': video_id,
-            'title': title,
+            'title': self._live_title(title) if info.get('is_live') else title,
             'description': description,
             'thumbnail': thumbnail,
         })

youtube_dl/extractor/arkena.py

@@ -93,8 +93,7 @@ class ArkenaIE(InfoExtractor):
                 exts = (mimetype2ext(f.get('Type')), determine_ext(f_url, None))
                 if kind == 'm3u8' or 'm3u8' in exts:
                     formats.extend(self._extract_m3u8_formats(
-                        f_url, video_id, 'mp4',
-                        entry_protocol='m3u8' if is_live else 'm3u8_native',
+                        f_url, video_id, 'mp4', 'm3u8_native',
                         m3u8_id=kind, fatal=False, live=is_live))
                 elif kind == 'flash' or 'f4m' in exts:
                     formats.extend(self._extract_f4m_formats(

youtube_dl/extractor/arte.py

@@ -6,15 +6,18 @@ import re
 from .common import InfoExtractor
 from ..compat import (
     compat_parse_qs,
+    compat_str,
     compat_urllib_parse_urlparse,
 )
 from ..utils import (
+    ExtractorError,
     find_xpath_attr,
-    unified_strdate,
     get_element_by_attribute,
     int_or_none,
     NO_DEFAULT,
     qualities,
+    try_get,
+    unified_strdate,
 )
 
 # There are different sources of video in arte.tv, the extraction process
@@ -79,6 +82,16 @@ class ArteTVBaseIE(InfoExtractor):
         info = self._download_json(json_url, video_id)
         player_info = info['videoJsonPlayer']
 
+        vsr = try_get(player_info, lambda x: x['VSR'], dict)
+        if not vsr:
+            error = None
+            if try_get(player_info, lambda x: x['custom_msg']['type']) == 'error':
+                error = try_get(
+                    player_info, lambda x: x['custom_msg']['msg'], compat_str)
+            if not error:
+                error = 'Video %s is not available' % player_info.get('VID') or video_id
+            raise ExtractorError(error, expected=True)
+
         upload_date_str = player_info.get('shootingDate')
         if not upload_date_str:
             upload_date_str = (player_info.get('VRA') or player_info.get('VDA') or '').split(' ')[0]
@@ -107,7 +120,7 @@ class ArteTVBaseIE(InfoExtractor):
         langcode = LANGS.get(lang, lang)
 
         formats = []
-        for format_id, format_dict in player_info['VSR'].items():
+        for format_id, format_dict in vsr.items():
             f = dict(format_dict)
             versionCode = f.get('versionCode')
             l = re.escape(langcode)
@@ -180,7 +193,7 @@ class ArteTVBaseIE(InfoExtractor):
 
 class ArteTVPlus7IE(ArteTVBaseIE):
     IE_NAME = 'arte.tv:+7'
-    _VALID_URL = r'https?://(?:(?:www|sites)\.)?arte\.tv/[^/]+/(?P<lang>fr|de|en|es)/(?:[^/]+/)*(?P<id>[^/?#&]+)'
+    _VALID_URL = r'https?://(?:(?:www|sites)\.)?arte\.tv/(?:[^/]+/)?(?P<lang>fr|de|en|es)/(?:videos/)?(?:[^/]+/)*(?P<id>[^/?#&]+)'
 
     _TESTS = [{
         'url': 'http://www.arte.tv/guide/de/sendungen/XEN/xenius/?vid=055918-015_PLUS7-D',
@@ -188,6 +201,9 @@ class ArteTVPlus7IE(ArteTVBaseIE):
     }, {
         'url': 'http://sites.arte.tv/karambolage/de/video/karambolage-22',
         'only_matching': True,
+    }, {
+        'url': 'http://www.arte.tv/de/videos/048696-000-A/der-kluge-bauch-unser-zweites-gehirn',
+        'only_matching': True,
     }]
 
     @classmethod

youtube_dl/extractor/asiancrush.py

@@ -0,0 +1,93 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from .kaltura import KalturaIE
+from ..utils import (
+    extract_attributes,
+    remove_end,
+    urlencode_postdata,
+)
+
+
+class AsianCrushIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?asiancrush\.com/video/(?:[^/]+/)?0+(?P<id>\d+)v\b'
+    _TESTS = [{
+        'url': 'https://www.asiancrush.com/video/012869v/women-who-flirt/',
+        'md5': 'c3b740e48d0ba002a42c0b72857beae6',
+        'info_dict': {
+            'id': '1_y4tmjm5r',
+            'ext': 'mp4',
+            'title': 'Women Who Flirt',
+            'description': 'md5:3db14e9186197857e7063522cb89a805',
+            'timestamp': 1496936429,
+            'upload_date': '20170608',
+            'uploader_id': 'craig@crifkin.com',
+        },
+    }, {
+        'url': 'https://www.asiancrush.com/video/she-was-pretty/011886v-pretty-episode-3/',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        data = self._download_json(
+            'https://www.asiancrush.com/wp-admin/admin-ajax.php', video_id,
+            data=urlencode_postdata({
+                'postid': video_id,
+                'action': 'get_channel_kaltura_vars',
+            }))
+
+        entry_id = data['entry_id']
+
+        return self.url_result(
+            'kaltura:%s:%s' % (data['partner_id'], entry_id),
+            ie=KalturaIE.ie_key(), video_id=entry_id,
+            video_title=data.get('vid_label'))
+
+
+class AsianCrushPlaylistIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?asiancrush\.com/series/0+(?P<id>\d+)s\b'
+    _TEST = {
+        'url': 'https://www.asiancrush.com/series/012481s/scholar-walks-night/',
+        'info_dict': {
+            'id': '12481',
+            'title': 'Scholar Who Walks the Night',
+            'description': 'md5:7addd7c5132a09fd4741152d96cce886',
+        },
+        'playlist_count': 20,
+    }
+
+    def _real_extract(self, url):
+        playlist_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, playlist_id)
+
+        entries = []
+
+        for mobj in re.finditer(
+                r'<a[^>]+href=(["\'])(?P<url>%s.*?)\1[^>]*>' % AsianCrushIE._VALID_URL,
+                webpage):
+            attrs = extract_attributes(mobj.group(0))
+            if attrs.get('class') == 'clearfix':
+                entries.append(self.url_result(
+                    mobj.group('url'), ie=AsianCrushIE.ie_key()))
+
+        title = remove_end(
+            self._html_search_regex(
+                r'(?s)<h1\b[^>]\bid=["\']movieTitle[^>]+>(.+?)</h1>', webpage,
+                'title', default=None) or self._og_search_title(
+                webpage, default=None) or self._html_search_meta(
+                'twitter:title', webpage, 'title',
+                default=None) or self._search_regex(
+                r'<title>([^<]+)</title>', webpage, 'title', fatal=False),
+            ' | AsianCrush')
+
+        description = self._og_search_description(
+            webpage, default=None) or self._html_search_meta(
+            'twitter:description', webpage, 'description', fatal=False)
+
+        return self.playlist_result(entries, playlist_id, title, description)

youtube_dl/extractor/atresplayer.py

@@ -36,7 +36,7 @@ class AtresPlayerIE(InfoExtractor):
         },
         {
             'url': 'http://www.atresplayer.com/television/especial/videoencuentros/temporada-1/capitulo-112-david-bustamante_2014121600375.html',
-            'md5': '0d0e918533bbd4b263f2de4d197d4aac',
+            'md5': '6e52cbb513c405e403dbacb7aacf8747',
             'info_dict': {
                 'id': 'capitulo-112-david-bustamante',
                 'ext': 'flv',
@@ -87,10 +87,11 @@ class AtresPlayerIE(InfoExtractor):
             self._LOGIN_URL, urlencode_postdata(login_form))
         request.add_header('Content-Type', 'application/x-www-form-urlencoded')
         response = self._download_webpage(
-            request, None, 'Logging in as %s' % username)
+            request, None, 'Logging in')
 
         error = self._html_search_regex(
-            r'(?s)<ul class="list_error">(.+?)</ul>', response, 'error', default=None)
+            r'(?s)<ul[^>]+class="[^"]*\blist_error\b[^"]*">(.+?)</ul>',
+            response, 'error', default=None)
         if error:
             raise ExtractorError(
                 'Unable to login: %s' % error, expected=True)
@@ -155,13 +156,17 @@ class AtresPlayerIE(InfoExtractor):
             if format_id == 'token' or not video_url.startswith('http'):
                 continue
             if 'geodeswowsmpra3player' in video_url:
-                f4m_path = video_url.split('smil:', 1)[-1].split('free_', 1)[0]
-                f4m_url = 'http://drg.antena3.com/{0}hds/es/sd.f4m'.format(f4m_path)
+                # f4m_path = video_url.split('smil:', 1)[-1].split('free_', 1)[0]
+                # f4m_url = 'http://drg.antena3.com/{0}hds/es/sd.f4m'.format(f4m_path)
                 # this videos are protected by DRM, the f4m downloader doesn't support them
                 continue
-            else:
-                f4m_url = video_url[:-9] + '/manifest.f4m'
-            formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False))
+            video_url_hd = video_url.replace('free_es', 'es')
+            formats.extend(self._extract_f4m_formats(
+                video_url_hd[:-9] + '/manifest.f4m', video_id, f4m_id='hds',
+                fatal=False))
+            formats.extend(self._extract_mpd_formats(
+                video_url_hd[:-9] + '/manifest.mpd', video_id, mpd_id='dash',
+                fatal=False))
         self._sort_formats(formats)
 
         path_data = player.get('pathData')

youtube_dl/extractor/atvat.py

@@ -0,0 +1,73 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    unescapeHTML,
+)
+
+
+class ATVAtIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?atv\.at/(?:[^/]+/){2}(?P<id>[dv]\d+)'
+    _TESTS = [{
+        'url': 'http://atv.at/aktuell/di-210317-2005-uhr/v1698449/',
+        'md5': 'c3b6b975fb3150fc628572939df205f2',
+        'info_dict': {
+            'id': '1698447',
+            'ext': 'mp4',
+            'title': 'DI, 21.03.17 | 20:05 Uhr 1/1',
+        }
+    }, {
+        'url': 'http://atv.at/aktuell/meinrad-knapp/d8416/',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url)
+        webpage = self._download_webpage(url, display_id)
+        video_data = self._parse_json(unescapeHTML(self._search_regex(
+            r'class="[^"]*jsb_video/FlashPlayer[^"]*"[^>]+data-jsb="([^"]+)"',
+            webpage, 'player data')), display_id)['config']['initial_video']
+
+        video_id = video_data['id']
+        video_title = video_data['title']
+
+        parts = []
+        for part in video_data.get('parts', []):
+            part_id = part['id']
+            part_title = part['title']
+
+            formats = []
+            for source in part.get('sources', []):
+                source_url = source.get('src')
+                if not source_url:
+                    continue
+                ext = determine_ext(source_url)
+                if ext == 'm3u8':
+                    formats.extend(self._extract_m3u8_formats(
+                        source_url, part_id, 'mp4', 'm3u8_native',
+                        m3u8_id='hls', fatal=False))
+                else:
+                    formats.append({
+                        'format_id': source.get('delivery'),
+                        'url': source_url,
+                    })
+            self._sort_formats(formats)
+
+            parts.append({
+                'id': part_id,
+                'title': part_title,
+                'thumbnail': part.get('preview_image_url'),
+                'duration': int_or_none(part.get('duration')),
+                'is_live': part.get('is_livestream'),
+                'formats': formats,
+            })
+
+        return {
+            '_type': 'multi_video',
+            'id': video_id,
+            'title': video_title,
+            'entries': parts,
+        }

youtube_dl/extractor/audioboom.py

@@ -16,7 +16,7 @@ class AudioBoomIE(InfoExtractor):
             'title': '3/09/2016 Czaban Hour 3',
             'description': 'Guest:   Nate Davis - NFL free agency,   Guest:   Stan Gans',
             'duration': 2245.72,
-            'uploader': 'Steve Czaban',
+            'uploader': 'SB Nation A.M.',
             'uploader_url': r're:https?://(?:www\.)?audioboom\.com/channel/steveczabanyahoosportsradio',
         }
     }, {
@@ -43,7 +43,7 @@ class AudioBoomIE(InfoExtractor):
 
         def from_clip(field):
             if clip:
-                clip.get(field)
+                return clip.get(field)
 
         audio_url = from_clip('clipURLPriorToLoading') or self._og_search_property(
             'audio', webpage, 'audio url')

youtube_dl/extractor/aws.py

@@ -0,0 +1,78 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import datetime
+import hashlib
+import hmac
+
+from .common import InfoExtractor
+from ..compat import compat_urllib_parse_urlencode
+
+
+class AWSIE(InfoExtractor):
+    _AWS_ALGORITHM = 'AWS4-HMAC-SHA256'
+    _AWS_REGION = 'us-east-1'
+
+    def _aws_execute_api(self, aws_dict, video_id, query=None):
+        query = query or {}
+        amz_date = datetime.datetime.utcnow().strftime('%Y%m%dT%H%M%SZ')
+        date = amz_date[:8]
+        headers = {
+            'Accept': 'application/json',
+            'Host': self._AWS_PROXY_HOST,
+            'X-Amz-Date': amz_date,
+            'X-Api-Key': self._AWS_API_KEY
+        }
+        session_token = aws_dict.get('session_token')
+        if session_token:
+            headers['X-Amz-Security-Token'] = session_token
+
+        def aws_hash(s):
+            return hashlib.sha256(s.encode('utf-8')).hexdigest()
+
+        # Task 1: http://docs.aws.amazon.com/general/latest/gr/sigv4-create-canonical-request.html
+        canonical_querystring = compat_urllib_parse_urlencode(query)
+        canonical_headers = ''
+        for header_name, header_value in sorted(headers.items()):
+            canonical_headers += '%s:%s\n' % (header_name.lower(), header_value)
+        signed_headers = ';'.join([header.lower() for header in sorted(headers.keys())])
+        canonical_request = '\n'.join([
+            'GET',
+            aws_dict['uri'],
+            canonical_querystring,
+            canonical_headers,
+            signed_headers,
+            aws_hash('')
+        ])
+
+        # Task 2: http://docs.aws.amazon.com/general/latest/gr/sigv4-create-string-to-sign.html
+        credential_scope_list = [date, self._AWS_REGION, 'execute-api', 'aws4_request']
+        credential_scope = '/'.join(credential_scope_list)
+        string_to_sign = '\n'.join([self._AWS_ALGORITHM, amz_date, credential_scope, aws_hash(canonical_request)])
+
+        # Task 3: http://docs.aws.amazon.com/general/latest/gr/sigv4-calculate-signature.html
+        def aws_hmac(key, msg):
+            return hmac.new(key, msg.encode('utf-8'), hashlib.sha256)
+
+        def aws_hmac_digest(key, msg):
+            return aws_hmac(key, msg).digest()
+
+        def aws_hmac_hexdigest(key, msg):
+            return aws_hmac(key, msg).hexdigest()
+
+        k_signing = ('AWS4' + aws_dict['secret_key']).encode('utf-8')
+        for value in credential_scope_list:
+            k_signing = aws_hmac_digest(k_signing, value)
+
+        signature = aws_hmac_hexdigest(k_signing, string_to_sign)
+
+        # Task 4: http://docs.aws.amazon.com/general/latest/gr/sigv4-add-signature-to-request.html
+        headers['Authorization'] = ', '.join([
+            '%s Credential=%s/%s' % (self._AWS_ALGORITHM, aws_dict['access_key'], credential_scope),
+            'SignedHeaders=%s' % signed_headers,
+            'Signature=%s' % signature,
+        ])
+
+        return self._download_json(
+            'https://%s%s%s' % (self._AWS_PROXY_HOST, aws_dict['uri'], '?' + canonical_querystring if canonical_querystring else ''),
+            video_id, headers=headers)

youtube_dl/extractor/azmedien.py

@@ -0,0 +1,213 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from .kaltura import KalturaIE
+from ..utils import (
+    get_element_by_class,
+    get_element_by_id,
+    strip_or_none,
+    urljoin,
+)
+
+
+class AZMedienBaseIE(InfoExtractor):
+    def _kaltura_video(self, partner_id, entry_id):
+        return self.url_result(
+            'kaltura:%s:%s' % (partner_id, entry_id), ie=KalturaIE.ie_key(),
+            video_id=entry_id)
+
+
+class AZMedienIE(AZMedienBaseIE):
+    IE_DESC = 'AZ Medien videos'
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:www\.)?
+                        (?:
+                            telezueri\.ch|
+                            telebaern\.tv|
+                            telem1\.ch
+                        )/
+                        [0-9]+-show-[^/\#]+
+                        (?:
+                            /[0-9]+-episode-[^/\#]+
+                            (?:
+                                /[0-9]+-segment-(?:[^/\#]+\#)?|
+                                \#
+                            )|
+                            \#
+                        )
+                        (?P<id>[^\#]+)
+                    '''
+
+    _TESTS = [{
+        # URL with 'segment'
+        'url': 'http://www.telezueri.ch/62-show-zuerinews/13772-episode-sonntag-18-dezember-2016/32419-segment-massenabweisungen-beim-hiltl-club-wegen-pelzboom',
+        'info_dict': {
+            'id': '1_2444peh4',
+            'ext': 'mp4',
+            'title': 'Massenabweisungen beim Hiltl Club wegen Pelzboom',
+            'description': 'md5:9ea9dd1b159ad65b36ddcf7f0d7c76a8',
+            'uploader_id': 'TeleZ?ri',
+            'upload_date': '20161218',
+            'timestamp': 1482084490,
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }, {
+        # URL with 'segment' and fragment:
+        'url': 'http://www.telebaern.tv/118-show-news/14240-episode-dienstag-17-januar-2017/33666-segment-achtung-gefahr#zu-wenig-pflegerinnen-und-pfleger',
+        'only_matching': True
+    }, {
+        # URL with 'episode' and fragment:
+        'url': 'http://www.telem1.ch/47-show-sonntalk/13986-episode-soldaten-fuer-grenzschutz-energiestrategie-obama-bilanz#soldaten-fuer-grenzschutz-energiestrategie-obama-bilanz',
+        'only_matching': True
+    }, {
+        # URL with 'show' and fragment:
+        'url': 'http://www.telezueri.ch/66-show-sonntalk#burka-plakate-trump-putin-china-besuch',
+        'only_matching': True
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, video_id)
+
+        partner_id = self._search_regex(
+            r'<script[^>]+src=["\'](?:https?:)?//(?:[^/]+\.)?kaltura\.com(?:/[^/]+)*/(?:p|partner_id)/([0-9]+)',
+            webpage, 'kaltura partner id')
+        entry_id = self._html_search_regex(
+            r'<a[^>]+data-id=(["\'])(?P<id>(?:(?!\1).)+)\1[^>]+data-slug=["\']%s'
+            % re.escape(video_id), webpage, 'kaltura entry id', group='id')
+
+        return self._kaltura_video(partner_id, entry_id)
+
+
+class AZMedienPlaylistIE(AZMedienBaseIE):
+    IE_DESC = 'AZ Medien playlists'
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:www\.)?
+                        (?:
+                            telezueri\.ch|
+                            telebaern\.tv|
+                            telem1\.ch
+                        )/
+                        (?P<id>[0-9]+-
+                            (?:
+                                show|
+                                topic|
+                                themen
+                            )-[^/\#]+
+                            (?:
+                                /[0-9]+-episode-[^/\#]+
+                            )?
+                        )$
+                    '''
+
+    _TESTS = [{
+        # URL with 'episode'
+        'url': 'http://www.telebaern.tv/118-show-news/13735-episode-donnerstag-15-dezember-2016',
+        'info_dict': {
+            'id': '118-show-news/13735-episode-donnerstag-15-dezember-2016',
+            'title': 'News - Donnerstag, 15. Dezember 2016',
+        },
+        'playlist_count': 9,
+    }, {
+        # URL with 'themen'
+        'url': 'http://www.telem1.ch/258-themen-tele-m1-classics',
+        'info_dict': {
+            'id': '258-themen-tele-m1-classics',
+            'title': 'Tele M1 Classics',
+        },
+        'playlist_mincount': 15,
+    }, {
+        # URL with 'topic', contains nested playlists
+        'url': 'http://www.telezueri.ch/219-topic-aera-trump-hat-offiziell-begonnen',
+        'only_matching': True,
+    }, {
+        # URL with 'show' only
+        'url': 'http://www.telezueri.ch/86-show-talktaeglich',
+        'only_matching': True
+    }]
+
+    def _real_extract(self, url):
+        show_id = self._match_id(url)
+        webpage = self._download_webpage(url, show_id)
+
+        entries = []
+
+        partner_id = self._search_regex(
+            r'src=["\'](?:https?:)?//(?:[^/]+\.)kaltura\.com/(?:[^/]+/)*(?:p|partner_id)/(\d+)',
+            webpage, 'kaltura partner id', default=None)
+
+        if partner_id:
+            entries = [
+                self._kaltura_video(partner_id, m.group('id'))
+                for m in re.finditer(
+                    r'data-id=(["\'])(?P<id>(?:(?!\1).)+)\1', webpage)]
+
+        if not entries:
+            entries = [
+                self.url_result(m.group('url'), ie=AZMedienIE.ie_key())
+                for m in re.finditer(
+                    r'<a[^>]+data-real=(["\'])(?P<url>http.+?)\1', webpage)]
+
+        if not entries:
+            entries = [
+                # May contain nested playlists (e.g. [1]) thus no explicit
+                # ie_key
+                # 1. http://www.telezueri.ch/219-topic-aera-trump-hat-offiziell-begonnen)
+                self.url_result(urljoin(url, m.group('url')))
+                for m in re.finditer(
+                    r'<a[^>]+name=[^>]+href=(["\'])(?P<url>/.+?)\1', webpage)]
+
+        title = self._search_regex(
+            r'episodeShareTitle\s*=\s*(["\'])(?P<title>(?:(?!\1).)+)\1',
+            webpage, 'title',
+            default=strip_or_none(get_element_by_id(
+                'video-title', webpage)), group='title')
+
+        return self.playlist_result(entries, show_id, title)
+
+
+class AZMedienShowPlaylistIE(AZMedienBaseIE):
+    IE_DESC = 'AZ Medien show playlists'
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:www\.)?
+                        (?:
+                            telezueri\.ch|
+                            telebaern\.tv|
+                            telem1\.ch
+                        )/
+                        (?:
+                            all-episodes|
+                            alle-episoden
+                        )/
+                        (?P<id>[^/?#&]+)
+                    '''
+
+    _TEST = {
+        'url': 'http://www.telezueri.ch/all-episodes/astrotalk',
+        'info_dict': {
+            'id': 'astrotalk',
+            'title': 'TeleZüri: AstroTalk - alle episoden',
+            'description': 'md5:4c0f7e7d741d906004266e295ceb4a26',
+        },
+        'playlist_mincount': 13,
+    }
+
+    def _real_extract(self, url):
+        playlist_id = self._match_id(url)
+        webpage = self._download_webpage(url, playlist_id)
+        episodes = get_element_by_class('search-mobile-box', webpage)
+        entries = [self.url_result(
+            urljoin(url, m.group('url'))) for m in re.finditer(
+                r'<a[^>]+href=(["\'])(?P<url>(?:(?!\1).)+)\1', episodes)]
+        title = self._og_search_title(webpage, fatal=False)
+        description = self._og_search_description(webpage)
+        return self.playlist_result(entries, playlist_id, title, description)

youtube_dl/extractor/azubu.py

@@ -1,140 +0,0 @@
-from __future__ import unicode_literals
-
-import json
-
-from .common import InfoExtractor
-from ..utils import (
-    ExtractorError,
-    float_or_none,
-    sanitized_Request,
-)
-
-
-class AzubuIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?azubu\.(?:tv|uol.com.br)/[^/]+#!/play/(?P<id>\d+)'
-    _TESTS = [
-        {
-            'url': 'http://www.azubu.tv/GSL#!/play/15575/2014-hot6-cup-last-big-match-ro8-day-1',
-            'md5': 'a88b42fcf844f29ad6035054bd9ecaf4',
-            'info_dict': {
-                'id': '15575',
-                'ext': 'mp4',
-                'title': '2014 HOT6 CUP LAST BIG MATCH Ro8 Day 1',
-                'description': 'md5:d06bdea27b8cc4388a90ad35b5c66c01',
-                'thumbnail': r're:^https?://.*\.jpe?g',
-                'timestamp': 1417523507.334,
-                'upload_date': '20141202',
-                'duration': 9988.7,
-                'uploader': 'GSL',
-                'uploader_id': 414310,
-                'view_count': int,
-            },
-        },
-        {
-            'url': 'http://www.azubu.tv/FnaticTV#!/play/9344/-fnatic-at-worlds-2014:-toyz---%22i-love-rekkles,-he-has-amazing-mechanics%22-',
-            'md5': 'b72a871fe1d9f70bd7673769cdb3b925',
-            'info_dict': {
-                'id': '9344',
-                'ext': 'mp4',
-                'title': 'Fnatic at Worlds 2014: Toyz - "I love Rekkles, he has amazing mechanics"',
-                'description': 'md5:4a649737b5f6c8b5c5be543e88dc62af',
-                'thumbnail': r're:^https?://.*\.jpe?g',
-                'timestamp': 1410530893.320,
-                'upload_date': '20140912',
-                'duration': 172.385,
-                'uploader': 'FnaticTV',
-                'uploader_id': 272749,
-                'view_count': int,
-            },
-            'skip': 'Channel offline',
-        },
-    ]
-
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-
-        data = self._download_json(
-            'http://www.azubu.tv/api/video/%s' % video_id, video_id)['data']
-
-        title = data['title'].strip()
-        description = data.get('description')
-        thumbnail = data.get('thumbnail')
-        view_count = data.get('view_count')
-        user = data.get('user', {})
-        uploader = user.get('username')
-        uploader_id = user.get('id')
-
-        stream_params = json.loads(data['stream_params'])
-
-        timestamp = float_or_none(stream_params.get('creationDate'), 1000)
-        duration = float_or_none(stream_params.get('length'), 1000)
-
-        renditions = stream_params.get('renditions') or []
-        video = stream_params.get('FLVFullLength') or stream_params.get('videoFullLength')
-        if video:
-            renditions.append(video)
-
-        if not renditions and not user.get('channel', {}).get('is_live', True):
-            raise ExtractorError('%s said: channel is offline.' % self.IE_NAME, expected=True)
-
-        formats = [{
-            'url': fmt['url'],
-            'width': fmt['frameWidth'],
-            'height': fmt['frameHeight'],
-            'vbr': float_or_none(fmt['encodingRate'], 1000),
-            'filesize': fmt['size'],
-            'vcodec': fmt['videoCodec'],
-            'container': fmt['videoContainer'],
-        } for fmt in renditions if fmt['url']]
-        self._sort_formats(formats)
-
-        return {
-            'id': video_id,
-            'title': title,
-            'description': description,
-            'thumbnail': thumbnail,
-            'timestamp': timestamp,
-            'duration': duration,
-            'uploader': uploader,
-            'uploader_id': uploader_id,
-            'view_count': view_count,
-            'formats': formats,
-        }
-
-
-class AzubuLiveIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?azubu\.(?:tv|uol.com.br)/(?P<id>[^/]+)$'
-
-    _TESTS = [{
-        'url': 'http://www.azubu.tv/MarsTVMDLen',
-        'only_matching': True,
-    }, {
-        'url': 'http://azubu.uol.com.br/adolfz',
-        'only_matching': True,
-    }]
-
-    def _real_extract(self, url):
-        user = self._match_id(url)
-
-        info = self._download_json(
-            'http://api.azubu.tv/public/modules/last-video/{0}/info'.format(user),
-            user)['data']
-        if info['type'] != 'STREAM':
-            raise ExtractorError('{0} is not streaming live'.format(user), expected=True)
-
-        req = sanitized_Request(
-            'https://edge-elb.api.brightcove.com/playback/v1/accounts/3361910549001/videos/ref:' + info['reference_id'])
-        req.add_header('Accept', 'application/json;pk=BCpkADawqM1gvI0oGWg8dxQHlgT8HkdE2LnAlWAZkOlznO39bSZX726u4JqnDsK3MDXcO01JxXK2tZtJbgQChxgaFzEVdHRjaDoxaOu8hHOO8NYhwdxw9BzvgkvLUlpbDNUuDoc4E4wxDToV')
-        bc_info = self._download_json(req, user)
-        m3u8_url = next(source['src'] for source in bc_info['sources'] if source['container'] == 'M2TS')
-        formats = self._extract_m3u8_formats(m3u8_url, user, ext='mp4')
-        self._sort_formats(formats)
-
-        return {
-            'id': info['id'],
-            'title': self._live_title(info['title']),
-            'uploader_id': user,
-            'formats': formats,
-            'is_live': True,
-            'thumbnail': bc_info['poster'],
-        }

youtube_dl/extractor/bambuser.py

@@ -59,7 +59,7 @@ class BambuserIE(InfoExtractor):
             self._LOGIN_URL, urlencode_postdata(login_form))
         request.add_header('Referer', self._LOGIN_URL)
         response = self._download_webpage(
-            request, None, 'Logging in as %s' % username)
+            request, None, 'Logging in')
 
         login_error = self._html_search_regex(
             r'(?s)<div class="messages error">(.+?)</div>',

youtube_dl/extractor/bandcamp.py

@@ -14,14 +14,16 @@ from ..utils import (
     ExtractorError,
     float_or_none,
     int_or_none,
+    KNOWN_EXTENSIONS,
     parse_filesize,
     unescapeHTML,
     update_url_query,
+    unified_strdate,
 )
 
 
 class BandcampIE(InfoExtractor):
-    _VALID_URL = r'https?://.*?\.bandcamp\.com/track/(?P<title>.*)'
+    _VALID_URL = r'https?://.*?\.bandcamp\.com/track/(?P<title>[^/?#&]+)'
     _TESTS = [{
         'url': 'http://youtube-dl.bandcamp.com/track/youtube-dl-test-song',
         'md5': 'c557841d5e50261777a6585648adf439',
@@ -34,12 +36,12 @@ class BandcampIE(InfoExtractor):
         '_skip': 'There is a limit of 200 free downloads / month for the test song'
     }, {
         'url': 'http://benprunty.bandcamp.com/track/lanius-battle',
-        'md5': '73d0b3171568232574e45652f8720b5c',
+        'md5': '0369ace6b939f0927e62c67a1a8d9fa7',
         'info_dict': {
             'id': '2650410135',
-            'ext': 'mp3',
-            'title': 'Lanius (Battle)',
-            'uploader': 'Ben Prunty Music',
+            'ext': 'aiff',
+            'title': 'Ben Prunty - Lanius (Battle)',
+            'uploader': 'Ben Prunty',
         },
     }]
 
@@ -47,6 +49,7 @@ class BandcampIE(InfoExtractor):
         mobj = re.match(self._VALID_URL, url)
         title = mobj.group('title')
         webpage = self._download_webpage(url, title)
+        thumbnail = self._html_search_meta('og:image', webpage, default=None)
         m_download = re.search(r'freeDownloadPage: "(.*?)"', webpage)
         if not m_download:
             m_trackinfo = re.search(r'trackinfo: (.+),\s*?\n', webpage)
@@ -75,6 +78,7 @@ class BandcampIE(InfoExtractor):
                 return {
                     'id': track_id,
                     'title': data['title'],
+                    'thumbnail': thumbnail,
                     'formats': formats,
                     'duration': float_or_none(data.get('duration')),
                 }
@@ -143,7 +147,7 @@ class BandcampIE(InfoExtractor):
         return {
             'id': video_id,
             'title': title,
-            'thumbnail': info.get('thumb_url'),
+            'thumbnail': info.get('thumb_url') or thumbnail,
             'uploader': info.get('artist'),
             'artist': artist,
             'track': track,
@@ -153,7 +157,7 @@ class BandcampIE(InfoExtractor):
 
 class BandcampAlbumIE(InfoExtractor):
     IE_NAME = 'Bandcamp:album'
-    _VALID_URL = r'https?://(?:(?P<subdomain>[^.]+)\.)?bandcamp\.com(?:/album/(?P<album_id>[^?#]+)|/?(?:$|[?#]))'
+    _VALID_URL = r'https?://(?:(?P<subdomain>[^.]+)\.)?bandcamp\.com(?:/album/(?P<album_id>[^/?#&]+))?'
 
     _TESTS = [{
         'url': 'http://blazo.bandcamp.com/album/jazz-format-mixtape-vol-1',
@@ -209,20 +213,44 @@ class BandcampAlbumIE(InfoExtractor):
             'id': 'entropy-ep',
         },
         'playlist_mincount': 3,
+    }, {
+        # not all tracks have songs
+        'url': 'https://insulters.bandcamp.com/album/we-are-the-plague',
+        'info_dict': {
+            'id': 'we-are-the-plague',
+            'title': 'WE ARE THE PLAGUE',
+            'uploader_id': 'insulters',
+        },
+        'playlist_count': 2,
     }]
 
+    @classmethod
+    def suitable(cls, url):
+        return (False
+                if BandcampWeeklyIE.suitable(url) or BandcampIE.suitable(url)
+                else super(BandcampAlbumIE, cls).suitable(url))
+
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
         uploader_id = mobj.group('subdomain')
         album_id = mobj.group('album_id')
         playlist_id = album_id or uploader_id
         webpage = self._download_webpage(url, playlist_id)
-        tracks_paths = re.findall(r'<a href="(.*?)" itemprop="url">', webpage)
-        if not tracks_paths:
+        track_elements = re.findall(
+            r'(?s)<div[^>]*>(.*?<a[^>]+href="([^"]+?)"[^>]+itemprop="url"[^>]*>.*?)</div>', webpage)
+        if not track_elements:
             raise ExtractorError('The page doesn\'t contain any tracks')
+        # Only tracks with duration info have songs
         entries = [
-            self.url_result(compat_urlparse.urljoin(url, t_path), ie=BandcampIE.ie_key())
-            for t_path in tracks_paths]
+            self.url_result(
+                compat_urlparse.urljoin(url, t_path),
+                ie=BandcampIE.ie_key(),
+                video_title=self._search_regex(
+                    r'<span\b[^>]+\bitemprop=["\']name["\'][^>]*>([^<]+)',
+                    elem_content, 'track title', fatal=False))
+            for elem_content, t_path in track_elements
+            if self._html_search_meta('duration', elem_content, default=None)]
+
         title = self._html_search_regex(
             r'album_title\s*:\s*"((?:\\.|[^"\\])+?)"',
             webpage, 'title', fatal=False)
@@ -235,3 +263,92 @@ class BandcampAlbumIE(InfoExtractor):
             'title': title,
             'entries': entries,
         }
+
+
+class BandcampWeeklyIE(InfoExtractor):
+    IE_NAME = 'Bandcamp:weekly'
+    _VALID_URL = r'https?://(?:www\.)?bandcamp\.com/?\?(?:.*?&)?show=(?P<id>\d+)'
+    _TESTS = [{
+        'url': 'https://bandcamp.com/?show=224',
+        'md5': 'b00df799c733cf7e0c567ed187dea0fd',
+        'info_dict': {
+            'id': '224',
+            'ext': 'opus',
+            'title': 'BC Weekly April 4th 2017 - Magic Moments',
+            'description': 'md5:5d48150916e8e02d030623a48512c874',
+            'duration': 5829.77,
+            'release_date': '20170404',
+            'series': 'Bandcamp Weekly',
+            'episode': 'Magic Moments',
+            'episode_number': 208,
+            'episode_id': '224',
+        }
+    }, {
+        'url': 'https://bandcamp.com/?blah/blah@&show=228',
+        'only_matching': True
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        blob = self._parse_json(
+            self._search_regex(
+                r'data-blob=(["\'])(?P<blob>{.+?})\1', webpage,
+                'blob', group='blob'),
+            video_id, transform_source=unescapeHTML)
+
+        show = blob['bcw_show']
+
+        # This is desired because any invalid show id redirects to `bandcamp.com`
+        # which happens to expose the latest Bandcamp Weekly episode.
+        show_id = int_or_none(show.get('show_id')) or int_or_none(video_id)
+
+        formats = []
+        for format_id, format_url in show['audio_stream'].items():
+            if not isinstance(format_url, compat_str):
+                continue
+            for known_ext in KNOWN_EXTENSIONS:
+                if known_ext in format_id:
+                    ext = known_ext
+                    break
+            else:
+                ext = None
+            formats.append({
+                'format_id': format_id,
+                'url': format_url,
+                'ext': ext,
+                'vcodec': 'none',
+            })
+        self._sort_formats(formats)
+
+        title = show.get('audio_title') or 'Bandcamp Weekly'
+        subtitle = show.get('subtitle')
+        if subtitle:
+            title += ' - %s' % subtitle
+
+        episode_number = None
+        seq = blob.get('bcw_seq')
+
+        if seq and isinstance(seq, list):
+            try:
+                episode_number = next(
+                    int_or_none(e.get('episode_number'))
+                    for e in seq
+                    if isinstance(e, dict) and int_or_none(e.get('id')) == show_id)
+            except StopIteration:
+                pass
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': show.get('desc') or show.get('short_desc'),
+            'duration': float_or_none(show.get('audio_duration')),
+            'is_live': False,
+            'release_date': unified_strdate(show.get('published_date')),
+            'series': 'Bandcamp Weekly',
+            'episode': show.get('subtitle'),
+            'episode_number': episode_number,
+            'episode_id': compat_str(video_id),
+            'formats': formats
+        }

youtube_dl/extractor/bbc.py

@@ -6,14 +6,18 @@ import itertools
 
 from .common import InfoExtractor
 from ..utils import (
+    clean_html,
     dict_get,
     ExtractorError,
     float_or_none,
+    get_element_by_class,
     int_or_none,
     parse_duration,
     parse_iso8601,
     try_get,
     unescapeHTML,
+    urlencode_postdata,
+    urljoin,
 )
 from ..compat import (
     compat_etree_fromstring,
@@ -25,19 +29,23 @@ from ..compat import (
 class BBCCoUkIE(InfoExtractor):
     IE_NAME = 'bbc.co.uk'
     IE_DESC = 'BBC iPlayer'
-    _ID_REGEX = r'[pb][\da-z]{7}'
+    _ID_REGEX = r'[pbw][\da-z]{7}'
     _VALID_URL = r'''(?x)
                     https?://
                         (?:www\.)?bbc\.co\.uk/
                         (?:
                             programmes/(?!articles/)|
                             iplayer(?:/[^/]+)?/(?:episode/|playlist/)|
-                            music/clips[/#]|
-                            radio/player/
+                            music/(?:clips|audiovideo/popular)[/#]|
+                            radio/player/|
+                            events/[^/]+/play/[^/]+/
                         )
                         (?P<id>%s)(?!/(?:episodes|broadcasts|clips))
                     ''' % _ID_REGEX
 
+    _LOGIN_URL = 'https://account.bbc.com/signin'
+    _NETRC_MACHINE = 'bbc'
+
     _MEDIASELECTOR_URLS = [
         # Provides HQ HLS streams with even better quality that pc mediaset but fails
         # with geolocation in some cases when it's even not geo restricted at all (e.g.
@@ -222,8 +230,48 @@ class BBCCoUkIE(InfoExtractor):
         }, {
             'url': 'http://www.bbc.co.uk/radio/player/p03cchwf',
             'only_matching': True,
-        }
-    ]
+        }, {
+            'url': 'https://www.bbc.co.uk/music/audiovideo/popular#p055bc55',
+            'only_matching': True,
+        }, {
+            'url': 'http://www.bbc.co.uk/programmes/w3csv1y9',
+            'only_matching': True,
+        }]
+
+    _USP_RE = r'/([^/]+?)\.ism(?:\.hlsv2\.ism)?/[^/]+\.m3u8'
+
+    def _login(self):
+        username, password = self._get_login_info()
+        if username is None:
+            return
+
+        login_page = self._download_webpage(
+            self._LOGIN_URL, None, 'Downloading signin page')
+
+        login_form = self._hidden_inputs(login_page)
+
+        login_form.update({
+            'username': username,
+            'password': password,
+        })
+
+        post_url = urljoin(self._LOGIN_URL, self._search_regex(
+            r'<form[^>]+action=(["\'])(?P<url>.+?)\1', login_page,
+            'post url', default=self._LOGIN_URL, group='url'))
+
+        response, urlh = self._download_webpage_handle(
+            post_url, None, 'Logging in', data=urlencode_postdata(login_form),
+            headers={'Referer': self._LOGIN_URL})
+
+        if self._LOGIN_URL in urlh.geturl():
+            error = clean_html(get_element_by_class('form-message', response))
+            if error:
+                raise ExtractorError(
+                    'Unable to login: %s' % error, expected=True)
+            raise ExtractorError('Unable to log in')
+
+    def _real_initialize(self):
+        self._login()
 
     class MediaSelectionError(Exception):
         def __init__(self, id):
@@ -336,6 +384,15 @@ class BBCCoUkIE(InfoExtractor):
                         formats.extend(self._extract_m3u8_formats(
                             href, programme_id, ext='mp4', entry_protocol='m3u8_native',
                             m3u8_id=format_id, fatal=False))
+                        if re.search(self._USP_RE, href):
+                            usp_formats = self._extract_m3u8_formats(
+                                re.sub(self._USP_RE, r'/\1.ism/\1.m3u8', href),
+                                programme_id, ext='mp4', entry_protocol='m3u8_native',
+                                m3u8_id=format_id, fatal=False)
+                            for f in usp_formats:
+                                if f.get('height') and f['height'] > 720:
+                                    continue
+                                formats.append(f)
                     elif transfer_format == 'hds':
                         formats.extend(self._extract_f4m_formats(
                             href, programme_id, f4m_id=format_id, fatal=False))
@@ -350,7 +407,7 @@ class BBCCoUkIE(InfoExtractor):
                             fmt.update({
                                 'width': width,
                                 'height': height,
-                                'vbr': bitrate,
+                                'tbr': bitrate,
                                 'vcodec': encoding,
                             })
                         else:
@@ -359,7 +416,7 @@ class BBCCoUkIE(InfoExtractor):
                                 'acodec': encoding,
                                 'vcodec': 'none',
                             })
-                        if protocol == 'http':
+                        if protocol in ('http', 'https'):
                             # Direct link
                             fmt.update({
                                 'url': href,
@@ -378,6 +435,8 @@ class BBCCoUkIE(InfoExtractor):
                                 'rtmp_live': False,
                                 'ext': 'flv',
                             })
+                        else:
+                            continue
                         formats.append(fmt)
             elif kind == 'captions':
                 subtitles = self.extract_subtitles(media, programme_id)
@@ -396,7 +455,7 @@ class BBCCoUkIE(InfoExtractor):
                 description = smp_config['summary']
                 for item in smp_config['items']:
                     kind = item['kind']
-                    if kind != 'programme' and kind != 'radioProgramme':
+                    if kind not in ('programme', 'radioProgramme'):
                         continue
                     programme_id = item.get('vpid')
                     duration = int_or_none(item.get('duration'))
@@ -437,7 +496,7 @@ class BBCCoUkIE(InfoExtractor):
 
         for item in self._extract_items(playlist):
             kind = item.get('kind')
-            if kind != 'programme' and kind != 'radioProgramme':
+            if kind not in ('programme', 'radioProgramme'):
                 continue
             title = playlist.find('./{%s}title' % self._EMP_PLAYLIST_NS).text
             description_el = playlist.find('./{%s}summary' % self._EMP_PLAYLIST_NS)
@@ -470,6 +529,12 @@ class BBCCoUkIE(InfoExtractor):
 
         webpage = self._download_webpage(url, group_id, 'Downloading video page')
 
+        error = self._search_regex(
+            r'<div\b[^>]+\bclass=["\']smp__message delta["\'][^>]*>([^<]+)<',
+            webpage, 'error', default=None)
+        if error:
+            raise ExtractorError(error, expected=True)
+
         programme_id = None
         duration = None
 

youtube_dl/extractor/beampro.py

@@ -0,0 +1,188 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    ExtractorError,
+    clean_html,
+    compat_str,
+    float_or_none,
+    int_or_none,
+    parse_iso8601,
+    try_get,
+    urljoin,
+)
+
+
+class BeamProBaseIE(InfoExtractor):
+    _API_BASE = 'https://mixer.com/api/v1'
+    _RATINGS = {'family': 0, 'teen': 13, '18+': 18}
+
+    def _extract_channel_info(self, chan):
+        user_id = chan.get('userId') or try_get(chan, lambda x: x['user']['id'])
+        return {
+            'uploader': chan.get('token') or try_get(
+                chan, lambda x: x['user']['username'], compat_str),
+            'uploader_id': compat_str(user_id) if user_id else None,
+            'age_limit': self._RATINGS.get(chan.get('audience')),
+        }
+
+
+class BeamProLiveIE(BeamProBaseIE):
+    IE_NAME = 'Mixer:live'
+    _VALID_URL = r'https?://(?:\w+\.)?(?:beam\.pro|mixer\.com)/(?P<id>[^/?#&]+)'
+    _TEST = {
+        'url': 'http://mixer.com/niterhayven',
+        'info_dict': {
+            'id': '261562',
+            'ext': 'mp4',
+            'title': 'Introducing The Witcher 3 //  The Grind Starts Now!',
+            'description': 'md5:0b161ac080f15fe05d18a07adb44a74d',
+            'thumbnail': r're:https://.*\.jpg$',
+            'timestamp': 1483477281,
+            'upload_date': '20170103',
+            'uploader': 'niterhayven',
+            'uploader_id': '373396',
+            'age_limit': 18,
+            'is_live': True,
+            'view_count': int,
+        },
+        'skip': 'niterhayven is offline',
+        'params': {
+            'skip_download': True,
+        },
+    }
+
+    _MANIFEST_URL_TEMPLATE = '%s/channels/%%s/manifest.%%s' % BeamProBaseIE._API_BASE
+
+    @classmethod
+    def suitable(cls, url):
+        return False if BeamProVodIE.suitable(url) else super(BeamProLiveIE, cls).suitable(url)
+
+    def _real_extract(self, url):
+        channel_name = self._match_id(url)
+
+        chan = self._download_json(
+            '%s/channels/%s' % (self._API_BASE, channel_name), channel_name)
+
+        if chan.get('online') is False:
+            raise ExtractorError(
+                '{0} is offline'.format(channel_name), expected=True)
+
+        channel_id = chan['id']
+
+        def manifest_url(kind):
+            return self._MANIFEST_URL_TEMPLATE % (channel_id, kind)
+
+        formats = self._extract_m3u8_formats(
+            manifest_url('m3u8'), channel_name, ext='mp4', m3u8_id='hls',
+            fatal=False)
+        formats.extend(self._extract_smil_formats(
+            manifest_url('smil'), channel_name, fatal=False))
+        self._sort_formats(formats)
+
+        info = {
+            'id': compat_str(chan.get('id') or channel_name),
+            'title': self._live_title(chan.get('name') or channel_name),
+            'description': clean_html(chan.get('description')),
+            'thumbnail': try_get(
+                chan, lambda x: x['thumbnail']['url'], compat_str),
+            'timestamp': parse_iso8601(chan.get('updatedAt')),
+            'is_live': True,
+            'view_count': int_or_none(chan.get('viewersTotal')),
+            'formats': formats,
+        }
+        info.update(self._extract_channel_info(chan))
+
+        return info
+
+
+class BeamProVodIE(BeamProBaseIE):
+    IE_NAME = 'Mixer:vod'
+    _VALID_URL = r'https?://(?:\w+\.)?(?:beam\.pro|mixer\.com)/[^/?#&]+\?.*?\bvod=(?P<id>\d+)'
+    _TEST = {
+        'url': 'https://mixer.com/willow8714?vod=2259830',
+        'md5': 'b2431e6e8347dc92ebafb565d368b76b',
+        'info_dict': {
+            'id': '2259830',
+            'ext': 'mp4',
+            'title': 'willow8714\'s Channel',
+            'duration': 6828.15,
+            'thumbnail': r're:https://.*source\.png$',
+            'timestamp': 1494046474,
+            'upload_date': '20170506',
+            'uploader': 'willow8714',
+            'uploader_id': '6085379',
+            'age_limit': 13,
+            'view_count': int,
+        },
+        'params': {
+            'skip_download': True,
+        },
+    }
+
+    @staticmethod
+    def _extract_format(vod, vod_type):
+        if not vod.get('baseUrl'):
+            return []
+
+        if vod_type == 'hls':
+            filename, protocol = 'manifest.m3u8', 'm3u8_native'
+        elif vod_type == 'raw':
+            filename, protocol = 'source.mp4', 'https'
+        else:
+            assert False
+
+        data = vod.get('data') if isinstance(vod.get('data'), dict) else {}
+
+        format_id = [vod_type]
+        if isinstance(data.get('Height'), compat_str):
+            format_id.append('%sp' % data['Height'])
+
+        return [{
+            'url': urljoin(vod['baseUrl'], filename),
+            'format_id': '-'.join(format_id),
+            'ext': 'mp4',
+            'protocol': protocol,
+            'width': int_or_none(data.get('Width')),
+            'height': int_or_none(data.get('Height')),
+            'fps': int_or_none(data.get('Fps')),
+            'tbr': int_or_none(data.get('Bitrate'), 1000),
+        }]
+
+    def _real_extract(self, url):
+        vod_id = self._match_id(url)
+
+        vod_info = self._download_json(
+            '%s/recordings/%s' % (self._API_BASE, vod_id), vod_id)
+
+        state = vod_info.get('state')
+        if state != 'AVAILABLE':
+            raise ExtractorError(
+                'VOD %s is not available (state: %s)' % (vod_id, state),
+                expected=True)
+
+        formats = []
+        thumbnail_url = None
+
+        for vod in vod_info['vods']:
+            vod_type = vod.get('format')
+            if vod_type in ('hls', 'raw'):
+                formats.extend(self._extract_format(vod, vod_type))
+            elif vod_type == 'thumbnail':
+                thumbnail_url = urljoin(vod.get('baseUrl'), 'source.png')
+
+        self._sort_formats(formats)
+
+        info = {
+            'id': vod_id,
+            'title': vod_info.get('name') or vod_id,
+            'duration': float_or_none(vod_info.get('duration')),
+            'thumbnail': thumbnail_url,
+            'timestamp': parse_iso8601(vod_info.get('createdAt')),
+            'view_count': int_or_none(vod_info.get('viewsTotal')),
+            'formats': formats,
+        }
+        info.update(self._extract_channel_info(vod_info.get('channel') or {}))
+
+        return info

youtube_dl/extractor/beeg.py

@@ -9,6 +9,7 @@ from ..compat import (
 from ..utils import (
     int_or_none,
     parse_iso8601,
+    urljoin,
 )
 
 
@@ -16,7 +17,7 @@ class BeegIE(InfoExtractor):
     _VALID_URL = r'https?://(?:www\.)?beeg\.com/(?P<id>\d+)'
     _TEST = {
         'url': 'http://beeg.com/5416503',
-        'md5': '46c384def73b33dbc581262e5ee67cef',
+        'md5': 'a1a1b1a8bc70a89e49ccfd113aed0820',
         'info_dict': {
             'id': '5416503',
             'ext': 'mp4',
@@ -36,9 +37,11 @@ class BeegIE(InfoExtractor):
         webpage = self._download_webpage(url, video_id)
 
         cpl_url = self._search_regex(
-            r'<script[^>]+src=(["\'])(?P<url>(?:https?:)?//static\.beeg\.com/cpl/\d+\.js.*?)\1',
+            r'<script[^>]+src=(["\'])(?P<url>(?:/static|(?:https?:)?//static\.beeg\.com)/cpl/\d+\.js.*?)\1',
             webpage, 'cpl', default=None, group='url')
 
+        cpl_url = urljoin(url, cpl_url)
+
         beeg_version, beeg_salt = [None] * 2
 
         if cpl_url:
@@ -54,12 +57,16 @@ class BeegIE(InfoExtractor):
                     r'beeg_salt\s*=\s*(["\'])(?P<beeg_salt>.+?)\1', cpl, 'beeg salt',
                     default=None, group='beeg_salt')
 
-        beeg_version = beeg_version or '2000'
+        beeg_version = beeg_version or '2185'
         beeg_salt = beeg_salt or 'pmweAkq8lAYKdfWcFCUj0yoVgoPlinamH5UE1CB3H'
 
-        video = self._download_json(
-            'https://api.beeg.com/api/v6/%s/video/%s' % (beeg_version, video_id),
-            video_id)
+        for api_path in ('', 'api.'):
+            video = self._download_json(
+                'https://%sbeeg.com/api/v6/%s/video/%s'
+                % (api_path, beeg_version, video_id), video_id,
+                fatal=api_path == 'api.')
+            if video:
+                break
 
         def split(o, e):
             def cut(s, x):

youtube_dl/extractor/bellmedia.py

@@ -21,10 +21,11 @@ class BellMediaIE(InfoExtractor):
                 animalplanet|
                 bravo|
                 mtv|
-                space
+                space|
+                etalk
             )\.ca|
             much\.com
-        )/.*?(?:\bvid=|-vid|~|%7E|/(?:episode)?)(?P<id>[0-9]{6})'''
+        )/.*?(?:\bvid(?:eoid)?=|-vid|~|%7E|/(?:episode)?)(?P<id>[0-9]{6,})'''
     _TESTS = [{
         'url': 'http://www.ctv.ca/video/player?vid=706966',
         'md5': 'ff2ebbeae0aa2dcc32a830c3fd69b7b0',
@@ -55,6 +56,12 @@ class BellMediaIE(InfoExtractor):
     }, {
         'url': 'http://www.much.com/shows/the-almost-impossible-gameshow/928979/episode-6',
         'only_matching': True,
+    }, {
+        'url': 'http://www.ctv.ca/DCs-Legends-of-Tomorrow/Video/S2E11-Turncoat-vid1051430',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.etalk.ca/video?videoid=663455',
+        'only_matching': True,
     }]
     _DOMAINS = {
         'thecomedynetwork': 'comedy',
@@ -62,6 +69,7 @@ class BellMediaIE(InfoExtractor):
         'sciencechannel': 'discsci',
         'investigationdiscovery': 'invdisc',
         'animalplanet': 'aniplan',
+        'etalk': 'ctv',
     }
 
     def _real_extract(self, url):

youtube_dl/extractor/bilibili.py

@@ -5,19 +5,27 @@ import hashlib
 import re
 
 from .common import InfoExtractor
-from ..compat import compat_parse_qs
+from ..compat import (
+    compat_parse_qs,
+    compat_urlparse,
+)
 from ..utils import (
+    ExtractorError,
     int_or_none,
     float_or_none,
+    parse_iso8601,
+    smuggle_url,
+    strip_jsonp,
     unified_timestamp,
+    unsmuggle_url,
     urlencode_postdata,
 )
 
 
 class BiliBiliIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.|bangumi\.|)bilibili\.(?:tv|com)/(?:video/av|anime/v/)(?P<id>\d+)'
+    _VALID_URL = r'https?://(?:www\.|bangumi\.|)bilibili\.(?:tv|com)/(?:video/av|anime/(?P<anime_id>\d+)/play#)(?P<id>\d+)'
 
-    _TEST = {
+    _TESTS = [{
         'url': 'http://www.bilibili.tv/video/av1074402/',
         'md5': '9fa226fe2b8a9a4d5a69b4c6a183417e',
         'info_dict': {
@@ -32,25 +40,77 @@ class BiliBiliIE(InfoExtractor):
             'uploader': '菊子桑',
             'uploader_id': '156160',
         },
-    }
+    }, {
+        # Tested in BiliBiliBangumiIE
+        'url': 'http://bangumi.bilibili.com/anime/1869/play#40062',
+        'only_matching': True,
+    }, {
+        'url': 'http://bangumi.bilibili.com/anime/5802/play#100643',
+        'md5': '3f721ad1e75030cc06faf73587cfec57',
+        'info_dict': {
+            'id': '100643',
+            'ext': 'mp4',
+            'title': 'CHAOS;CHILD',
+            'description': '如果你是神明，并且能够让妄想成为现实。那你会进行怎么样的妄想？是淫靡的世界？独裁社会？毁灭性的制裁？还是……2015年，涩谷。从6年前发生的大灾害“涩谷地震”之后复兴了的这个街区里新设立的私立高中...',
+        },
+        'skip': 'Geo-restricted to China',
+    }, {
+        # Title with double quotes
+        'url': 'http://www.bilibili.com/video/av8903802/',
+        'info_dict': {
+            'id': '8903802',
+            'ext': 'mp4',
+            'title': '阿滴英文｜英文歌分享#6 "Closer',
+            'description': '滴妹今天唱Closer給你聽! 有史以来，被推最多次也是最久的歌曲，其实歌词跟我原本想像差蛮多的，不过还是好听！ 微博@阿滴英文',
+            'uploader': '阿滴英文',
+            'uploader_id': '65880958',
+            'timestamp': 1488382620,
+            'upload_date': '20170301',
+        },
+        'params': {
+            'skip_download': True,  # Test metadata only
+        },
+    }]
 
-    _APP_KEY = '6f90a59ac58a4123'
-    _BILIBILI_KEY = '0bfd84cc3940035173f35e6777508326'
+    _APP_KEY = '84956560bc028eb7'
+    _BILIBILI_KEY = '94aba54af9065f71de72f5508f1cd42e'
+
+    def _report_error(self, result):
+        if 'message' in result:
+            raise ExtractorError('%s said: %s' % (self.IE_NAME, result['message']), expected=True)
+        elif 'code' in result:
+            raise ExtractorError('%s returns error %d' % (self.IE_NAME, result['code']), expected=True)
+        else:
+            raise ExtractorError('Can\'t extract Bangumi episode ID')
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
+        url, smuggled_data = unsmuggle_url(url, {})
+
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
+        anime_id = mobj.group('anime_id')
         webpage = self._download_webpage(url, video_id)
 
-        if 'anime/v' not in url:
+        if 'anime/' not in url:
             cid = compat_parse_qs(self._search_regex(
                 [r'EmbedPlayer\([^)]+,\s*"([^"]+)"\)',
                  r'<iframe[^>]+src="https://secure\.bilibili\.com/secure,([^"]+)"'],
                 webpage, 'player parameters'))['cid'][0]
         else:
+            if 'no_bangumi_tip' not in smuggled_data:
+                self.to_screen('Downloading episode %s. To download all videos in anime %s, re-run youtube-dl with %s' % (
+                    video_id, anime_id, compat_urlparse.urljoin(url, '//bangumi.bilibili.com/anime/%s' % anime_id)))
+            headers = {
+                'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8',
+            }
+            headers.update(self.geo_verification_headers())
+
             js = self._download_json(
                 'http://bangumi.bilibili.com/web_api/get_source', video_id,
                 data=urlencode_postdata({'episode_id': video_id}),
-                headers={'Content-Type': 'application/x-www-form-urlencoded; charset=UTF-8'})
+                headers=headers)
+            if 'result' not in js:
+                self._report_error(js)
             cid = js['result']['cid']
 
         payload = 'appkey=%s&cid=%s&otype=json&quality=2&type=mp4' % (self._APP_KEY, cid)
@@ -58,7 +118,11 @@ class BiliBiliIE(InfoExtractor):
 
         video_info = self._download_json(
             'http://interface.bilibili.com/playurl?%s&sign=%s' % (payload, sign),
-            video_id, note='Downloading video info page')
+            video_id, note='Downloading video info page',
+            headers=self.geo_verification_headers())
+
+        if 'durl' not in video_info:
+            self._report_error(video_info)
 
         entries = []
 
@@ -74,6 +138,11 @@ class BiliBiliIE(InfoExtractor):
                     'preference': -2 if 'hd.mp4' in backup_url else -3,
                 })
 
+            for a_format in formats:
+                a_format.setdefault('http_headers', {}).update({
+                    'Referer': url,
+                })
+
             self._sort_formats(formats)
 
             entries.append({
@@ -82,10 +151,10 @@ class BiliBiliIE(InfoExtractor):
                 'formats': formats,
             })
 
-        title = self._html_search_regex('<h1[^>]+title="([^"]+)">', webpage, 'title')
+        title = self._html_search_regex('<h1[^>]*>([^<]+)</h1>', webpage, 'title')
         description = self._html_search_meta('description', webpage)
         timestamp = unified_timestamp(self._html_search_regex(
-            r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time', fatal=False))
+            r'<time[^>]+datetime="([^"]+)"', webpage, 'upload time', default=None))
         thumbnail = self._html_search_meta(['og:image', 'thumbnailUrl'], webpage)
 
         # TODO 'view_count' requires deobfuscating Javascript
@@ -99,7 +168,7 @@ class BiliBiliIE(InfoExtractor):
         }
 
         uploader_mobj = re.search(
-            r'<a[^>]+href="https?://space\.bilibili\.com/(?P<id>\d+)"[^>]+title="(?P<name>[^"]+)"',
+            r'<a[^>]+href="(?:https?:)?//space\.bilibili\.com/(?P<id>\d+)"[^>]+title="(?P<name>[^"]+)"',
             webpage)
         if uploader_mobj:
             info.update({
@@ -123,3 +192,70 @@ class BiliBiliIE(InfoExtractor):
                 'description': description,
                 'entries': entries,
             }
+
+
+class BiliBiliBangumiIE(InfoExtractor):
+    _VALID_URL = r'https?://bangumi\.bilibili\.com/anime/(?P<id>\d+)'
+
+    IE_NAME = 'bangumi.bilibili.com'
+    IE_DESC = 'BiliBili番剧'
+
+    _TESTS = [{
+        'url': 'http://bangumi.bilibili.com/anime/1869',
+        'info_dict': {
+            'id': '1869',
+            'title': '混沌武士',
+            'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
+        },
+        'playlist_count': 26,
+    }, {
+        'url': 'http://bangumi.bilibili.com/anime/1869',
+        'info_dict': {
+            'id': '1869',
+            'title': '混沌武士',
+            'description': 'md5:6a9622b911565794c11f25f81d6a97d2',
+        },
+        'playlist': [{
+            'md5': '91da8621454dd58316851c27c68b0c13',
+            'info_dict': {
+                'id': '40062',
+                'ext': 'mp4',
+                'title': '混沌武士',
+                'description': '故事发生在日本的江户时代。风是一个小酒馆的打工女。一日，酒馆里来了一群恶霸，虽然他们的举动令风十分不满，但是毕竟风只是一届女流，无法对他们采取什么行动，只能在心里嘟哝。这时，酒家里又进来了个“不良份子...',
+                'timestamp': 1414538739,
+                'upload_date': '20141028',
+                'episode': '疾风怒涛 Tempestuous Temperaments',
+                'episode_number': 1,
+            },
+        }],
+        'params': {
+            'playlist_items': '1',
+        },
+    }]
+
+    @classmethod
+    def suitable(cls, url):
+        return False if BiliBiliIE.suitable(url) else super(BiliBiliBangumiIE, cls).suitable(url)
+
+    def _real_extract(self, url):
+        bangumi_id = self._match_id(url)
+
+        # Sometimes this API returns a JSONP response
+        season_info = self._download_json(
+            'http://bangumi.bilibili.com/jsonp/seasoninfo/%s.ver' % bangumi_id,
+            bangumi_id, transform_source=strip_jsonp)['result']
+
+        entries = [{
+            '_type': 'url_transparent',
+            'url': smuggle_url(episode['webplay_url'], {'no_bangumi_tip': 1}),
+            'ie_key': BiliBiliIE.ie_key(),
+            'timestamp': parse_iso8601(episode.get('update_time'), delimiter=' '),
+            'episode': episode.get('index_title'),
+            'episode_number': int_or_none(episode.get('index')),
+        } for episode in season_info['episodes']]
+
+        entries = sorted(entries, key=lambda entry: entry.get('episode_number'))
+
+        return self.playlist_result(
+            entries, bangumi_id,
+            season_info.get('bangumi_title'), season_info.get('evaluate'))

youtube_dl/extractor/bleacherreport.py

@@ -35,7 +35,7 @@ class BleacherReportIE(InfoExtractor):
             'title': 'Aussie Golfers Get Fright of Their Lives After Being Chased by Angry Kangaroo',
             'timestamp': 1446839961,
             'uploader': 'Sean Fay',
-            'description': 'md5:825e94e0f3521df52fa83b2ed198fa20',
+            'description': 'md5:b1601e2314c4d8eec23b6eafe086a757',
             'uploader_id': 6466954,
             'upload_date': '20151011',
         },
@@ -90,17 +90,13 @@ class BleacherReportCMSIE(AMPIE):
     _VALID_URL = r'https?://(?:www\.)?bleacherreport\.com/video_embed\?id=(?P<id>[0-9a-f-]{36})'
     _TESTS = [{
         'url': 'http://bleacherreport.com/video_embed?id=8fd44c2f-3dc5-4821-9118-2c825a98c0e1',
-        'md5': '8c2c12e3af7805152675446c905d159b',
+        'md5': '2e4b0a997f9228ffa31fada5c53d1ed1',
         'info_dict': {
             'id': '8fd44c2f-3dc5-4821-9118-2c825a98c0e1',
-            'ext': 'mp4',
+            'ext': 'flv',
             'title': 'Cena vs. Rollins Would Expose the Heavyweight Division',
             'description': 'md5:984afb4ade2f9c0db35f3267ed88b36e',
         },
-        'params': {
-            # m3u8 download
-            'skip_download': True,
-        },
     }]
 
     def _real_extract(self, url):

youtube_dl/extractor/bloomberg.py

@@ -33,6 +33,10 @@ class BloombergIE(InfoExtractor):
         'params': {
             'format': 'best[format_id^=hds]',
         },
+    }, {
+        # data-bmmrid=
+        'url': 'https://www.bloomberg.com/politics/articles/2017-02-08/le-pen-aide-briefed-french-central-banker-on-plan-to-print-money',
+        'only_matching': True,
     }, {
         'url': 'http://www.bloomberg.com/news/articles/2015-11-12/five-strange-things-that-have-been-happening-in-financial-markets',
         'only_matching': True,
@@ -45,9 +49,10 @@ class BloombergIE(InfoExtractor):
         name = self._match_id(url)
         webpage = self._download_webpage(url, name)
         video_id = self._search_regex(
-            (r'["\']bmmrId["\']\s*:\s*(["\'])(?P<url>(?:(?!\1).)+)\1',
-             r'videoId\s*:\s*(["\'])(?P<url>(?:(?!\1).)+)\1'),
-            webpage, 'id', group='url', default=None)
+            (r'["\']bmmrId["\']\s*:\s*(["\'])(?P<id>(?:(?!\1).)+)\1',
+             r'videoId\s*:\s*(["\'])(?P<id>(?:(?!\1).)+)\1',
+             r'data-bmmrid=(["\'])(?P<id>(?:(?!\1).)+)\1'),
+            webpage, 'id', group='id', default=None)
         if not video_id:
             bplayer_data = self._parse_json(self._search_regex(
                 r'BPlayer\(null,\s*({[^;]+})\);', webpage, 'id'), name)

youtube_dl/extractor/bostonglobe.py

@@ -0,0 +1,72 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+
+from ..utils import (
+    extract_attributes,
+)
+
+
+class BostonGlobeIE(InfoExtractor):
+    _VALID_URL = r'(?i)https?://(?:www\.)?bostonglobe\.com/.*/(?P<id>[^/]+)/\w+(?:\.html)?'
+    _TESTS = [
+        {
+            'url': 'http://www.bostonglobe.com/metro/2017/02/11/tree-finally-succumbs-disease-leaving-hole-neighborhood/h1b4lviqzMTIn9sVy8F3gP/story.html',
+            'md5': '0a62181079c85c2d2b618c9a738aedaf',
+            'info_dict': {
+                'title': 'A tree finally succumbs to disease, leaving a hole in a neighborhood',
+                'id': '5320421710001',
+                'ext': 'mp4',
+                'description': 'It arrived as a sapling when the Back Bay was in its infancy, a spindly American elm tamped down into a square of dirt cut into the brick sidewalk of 1880s Marlborough Street, no higher than the first bay window of the new brownstone behind it.',
+                'timestamp': 1486877593,
+                'upload_date': '20170212',
+                'uploader_id': '245991542',
+            },
+        },
+        {
+            # Embedded youtube video; we hand it off to the Generic extractor.
+            'url': 'https://www.bostonglobe.com/lifestyle/names/2017/02/17/does-ben-affleck-play-matt-damon-favorite-version-batman/ruqkc9VxKBYmh5txn1XhSI/story.html',
+            'md5': '582b40327089d5c0c949b3c54b13c24b',
+            'info_dict': {
+                'title': "Who Is Matt Damon's Favorite Batman?",
+                'id': 'ZW1QCnlA6Qc',
+                'ext': 'mp4',
+                'upload_date': '20170217',
+                'description': 'md5:3b3dccb9375867e0b4d527ed87d307cb',
+                'uploader': 'The Late Late Show with James Corden',
+                'uploader_id': 'TheLateLateShow',
+            },
+            'expected_warnings': ['404'],
+        },
+    ]
+
+    def _real_extract(self, url):
+        page_id = self._match_id(url)
+        webpage = self._download_webpage(url, page_id)
+
+        page_title = self._og_search_title(webpage, default=None)
+
+        # <video data-brightcove-video-id="5320421710001" data-account="245991542" data-player="SJWAiyYWg" data-embed="default" class="video-js" controls itemscope itemtype="http://schema.org/VideoObject">
+        entries = []
+        for video in re.findall(r'(?i)(<video[^>]+>)', webpage):
+            attrs = extract_attributes(video)
+
+            video_id = attrs.get('data-brightcove-video-id')
+            account_id = attrs.get('data-account')
+            player_id = attrs.get('data-player')
+            embed = attrs.get('data-embed')
+
+            if video_id and account_id and player_id and embed:
+                entries.append(
+                    'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s'
+                    % (account_id, player_id, embed, video_id))
+
+        if len(entries) == 0:
+            return self.url_result(url, 'Generic')
+        elif len(entries) == 1:
+            return self.url_result(entries[0], 'BrightcoveNew')
+        else:
+            return self.playlist_from_matches(entries, page_id, page_title, ie='BrightcoveNew')

youtube_dl/extractor/bpb.py

@@ -33,13 +33,18 @@ class BpbIE(InfoExtractor):
         title = self._html_search_regex(
             r'<h2 class="white">(.*?)</h2>', webpage, 'title')
         video_info_dicts = re.findall(
-            r"({\s*src:\s*'http://film\.bpb\.de/[^}]+})", webpage)
+            r"({\s*src\s*:\s*'https?://film\.bpb\.de/[^}]+})", webpage)
 
         formats = []
         for video_info in video_info_dicts:
-            video_info = self._parse_json(video_info, video_id, transform_source=js_to_json)
-            quality = video_info['quality']
-            video_url = video_info['src']
+            video_info = self._parse_json(
+                video_info, video_id, transform_source=js_to_json, fatal=False)
+            if not video_info:
+                continue
+            video_url = video_info.get('src')
+            if not video_url:
+                continue
+            quality = 'high' if '_high' in video_url else 'low'
             formats.append({
                 'url': video_url,
                 'preference': 10 if quality == 'high' else 0,

youtube_dl/extractor/br.py

@@ -1,20 +1,23 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import json
 import re
 
 from .common import InfoExtractor
 from ..utils import (
+    determine_ext,
     ExtractorError,
     int_or_none,
     parse_duration,
+    parse_iso8601,
     xpath_element,
     xpath_text,
 )
 
 
 class BRIE(InfoExtractor):
-    IE_DESC = 'Bayerischer Rundfunk Mediathek'
+    IE_DESC = 'Bayerischer Rundfunk'
     _VALID_URL = r'(?P<base_url>https?://(?:www\.)?br(?:-klassik)?\.de)/(?:[a-z0-9\-_]+/)+(?P<id>[a-z0-9\-_]+)\.html'
 
     _TESTS = [
@@ -77,7 +80,7 @@ class BRIE(InfoExtractor):
                 'description': 'md5:bb659990e9e59905c3d41e369db1fbe3',
                 'duration': 893,
                 'uploader': 'Eva Maria Steimle',
-                'upload_date': '20140117',
+                'upload_date': '20170208',
             }
         },
     ]
@@ -123,10 +126,10 @@ class BRIE(InfoExtractor):
         for asset in assets.findall('asset'):
             format_url = xpath_text(asset, ['downloadUrl', 'url'])
             asset_type = asset.get('type')
-            if asset_type == 'HDS':
+            if asset_type.startswith('HDS'):
                 formats.extend(self._extract_f4m_formats(
                     format_url + '?hdcore=3.2.0', media_id, f4m_id='hds', fatal=False))
-            elif asset_type == 'HLS':
+            elif asset_type.startswith('HLS'):
                 formats.extend(self._extract_m3u8_formats(
                     format_url, media_id, 'mp4', 'm3u8_native', m3u8_id='hds', fatal=False))
             else:
@@ -169,3 +172,140 @@ class BRIE(InfoExtractor):
         } for variant in variants.findall('variant') if xpath_text(variant, 'url')]
         thumbnails.sort(key=lambda x: x['width'] * x['height'], reverse=True)
         return thumbnails
+
+
+class BRMediathekIE(InfoExtractor):
+    IE_DESC = 'Bayerischer Rundfunk Mediathek'
+    _VALID_URL = r'https?://(?:www\.)?br\.de/mediathek/video/[^/?&#]*?-(?P<id>av:[0-9a-f]{24})'
+
+    _TESTS = [{
+        'url': 'https://www.br.de/mediathek/video/gesundheit-die-sendung-vom-28112017-av:5a1e6a6e8fce6d001871cc8e',
+        'md5': 'fdc3d485835966d1622587d08ba632ec',
+        'info_dict': {
+            'id': 'av:5a1e6a6e8fce6d001871cc8e',
+            'ext': 'mp4',
+            'title': 'Die Sendung vom 28.11.2017',
+            'description': 'md5:6000cdca5912ab2277e5b7339f201ccc',
+            'timestamp': 1511942766,
+            'upload_date': '20171129',
+        }
+    }]
+
+    def _real_extract(self, url):
+        clip_id = self._match_id(url)
+
+        clip = self._download_json(
+            'https://proxy-base.master.mango.express/graphql',
+            clip_id, data=json.dumps({
+                "query": """{
+  viewer {
+    clip(id: "%s") {
+      title
+      description
+      duration
+      createdAt
+      ageRestriction
+      videoFiles {
+        edges {
+          node {
+            publicLocation
+            fileSize
+            videoProfile {
+              width
+              height
+              bitrate
+              encoding
+            }
+          }
+        }
+      }
+      captionFiles {
+        edges {
+          node {
+            publicLocation
+          }
+        }
+      }
+      teaserImages {
+        edges {
+          node {
+            imageFiles {
+              edges {
+                node {
+                  publicLocation
+                  width
+                  height
+                }
+              }
+            }
+          }
+        }
+      }
+    }
+  }
+}""" % clip_id}).encode(), headers={
+                'Content-Type': 'application/json',
+            })['data']['viewer']['clip']
+        title = clip['title']
+
+        formats = []
+        for edge in clip.get('videoFiles', {}).get('edges', []):
+            node = edge.get('node', {})
+            n_url = node.get('publicLocation')
+            if not n_url:
+                continue
+            ext = determine_ext(n_url)
+            if ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    n_url, clip_id, 'mp4', 'm3u8_native',
+                    m3u8_id='hls', fatal=False))
+            else:
+                video_profile = node.get('videoProfile', {})
+                tbr = int_or_none(video_profile.get('bitrate'))
+                format_id = 'http'
+                if tbr:
+                    format_id += '-%d' % tbr
+                formats.append({
+                    'format_id': format_id,
+                    'url': n_url,
+                    'width': int_or_none(video_profile.get('width')),
+                    'height': int_or_none(video_profile.get('height')),
+                    'tbr': tbr,
+                    'filesize': int_or_none(node.get('fileSize')),
+                })
+        self._sort_formats(formats)
+
+        subtitles = {}
+        for edge in clip.get('captionFiles', {}).get('edges', []):
+            node = edge.get('node', {})
+            n_url = node.get('publicLocation')
+            if not n_url:
+                continue
+            subtitles.setdefault('de', []).append({
+                'url': n_url,
+            })
+
+        thumbnails = []
+        for edge in clip.get('teaserImages', {}).get('edges', []):
+            for image_edge in edge.get('node', {}).get('imageFiles', {}).get('edges', []):
+                node = image_edge.get('node', {})
+                n_url = node.get('publicLocation')
+                if not n_url:
+                    continue
+                thumbnails.append({
+                    'url': n_url,
+                    'width': int_or_none(node.get('width')),
+                    'height': int_or_none(node.get('height')),
+                })
+
+        return {
+            'id': clip_id,
+            'title': title,
+            'description': clip.get('description'),
+            'duration': int_or_none(clip.get('duration')),
+            'timestamp': parse_iso8601(clip.get('createdAt')),
+            'age_limit': int_or_none(clip.get('ageRestriction')),
+            'formats': formats,
+            'subtitles': subtitles,
+            'thumbnails': thumbnails,
+        }

youtube_dl/extractor/brightcove.py

@@ -5,6 +5,7 @@ import re
 import json
 
 from .common import InfoExtractor
+from .adobepass import AdobePassIE
 from ..compat import (
     compat_etree_fromstring,
     compat_parse_qs,
@@ -17,6 +18,7 @@ from ..compat import (
 from ..utils import (
     determine_ext,
     ExtractorError,
+    extract_attributes,
     find_xpath_attr,
     fix_xml_ampersands,
     float_or_none,
@@ -109,6 +111,7 @@ class BrightcoveLegacyIE(InfoExtractor):
                 'upload_date': '20140827',
                 'uploader_id': '710858724001',
             },
+            'skip': 'Video gone',
         },
         {
             # playlist with 'videoList'
@@ -129,6 +132,12 @@ class BrightcoveLegacyIE(InfoExtractor):
             },
             'playlist_mincount': 10,
         },
+        {
+            # playerID inferred from bcpid
+            # from http://www.un.org/chinese/News/story.asp?NewsID=27724
+            'url': 'https://link.brightcove.com/services/player/bcpid1722935254001/?bctid=5360463607001&autoStart=false&secureConnections=true&width=650&height=350',
+            'only_matching': True,  # Tested in GenericIE
+        }
     ]
     FLV_VCODECS = {
         1: 'SORENSON',
@@ -179,7 +188,7 @@ class BrightcoveLegacyIE(InfoExtractor):
 
         params = {}
 
-        playerID = find_param('playerID')
+        playerID = find_param('playerID') or find_param('playerId')
         if playerID is None:
             raise ExtractorError('Cannot find player ID')
         params['playerID'] = playerID
@@ -191,6 +200,16 @@ class BrightcoveLegacyIE(InfoExtractor):
         # These fields hold the id of the video
         videoPlayer = find_param('@videoPlayer') or find_param('videoId') or find_param('videoID') or find_param('@videoList')
         if videoPlayer is not None:
+            if isinstance(videoPlayer, list):
+                videoPlayer = videoPlayer[0]
+            videoPlayer = videoPlayer.strip()
+            # UUID is also possible for videoPlayer (e.g.
+            # http://www.popcornflix.com/hoodies-vs-hooligans/7f2d2b87-bbf2-4623-acfb-ea942b4f01dd
+            # or http://www8.hp.com/cn/zh/home.html)
+            if not (re.match(
+                    r'^(?:\d+|[\da-fA-F]{8}-?[\da-fA-F]{4}-?[\da-fA-F]{4}-?[\da-fA-F]{4}-?[\da-fA-F]{12})$',
+                    videoPlayer) or videoPlayer.startswith('ref:')):
+                return None
             params['@videoPlayer'] = videoPlayer
         linkBase = find_param('linkBaseURL')
         if linkBase is not None:
@@ -204,7 +223,7 @@ class BrightcoveLegacyIE(InfoExtractor):
         #   // build Brightcove <object /> XML
         # }
         m = re.search(
-            r'''(?x)customBC.\createVideo\(
+            r'''(?x)customBC\.createVideo\(
                 .*?                                                  # skipping width and height
                 ["\'](?P<playerID>\d+)["\']\s*,\s*                   # playerID
                 ["\'](?P<playerKey>AQ[^"\']{48})[^"\']*["\']\s*,\s*  # playerKey begins with AQ and is 50 characters
@@ -254,9 +273,13 @@ class BrightcoveLegacyIE(InfoExtractor):
         if matches:
             return list(filter(None, [cls._build_brighcove_url(m) for m in matches]))
 
-        return list(filter(None, [
-            cls._build_brighcove_url_from_js(custom_bc)
-            for custom_bc in re.findall(r'(customBC\.createVideo\(.+?\);)', webpage)]))
+        matches = re.findall(r'(customBC\.createVideo\(.+?\);)', webpage)
+        if matches:
+            return list(filter(None, [
+                cls._build_brighcove_url_from_js(custom_bc)
+                for custom_bc in matches]))
+        return [src for _, src in re.findall(
+            r'<iframe[^>]+src=([\'"])((?:https?:)?//link\.brightcove\.com/services/player/(?!\1).+)\1', webpage)]
 
     def _real_extract(self, url):
         url, smuggled_data = unsmuggle_url(url, {})
@@ -273,6 +296,10 @@ class BrightcoveLegacyIE(InfoExtractor):
         if videoPlayer:
             # We set the original url as the default 'Referer' header
             referer = smuggled_data.get('Referer', url)
+            if 'playerID' not in query:
+                mobj = re.search(r'/bcpid(\d+)', url)
+                if mobj is not None:
+                    query['playerID'] = [mobj.group(1)]
             return self._get_video_info(
                 videoPlayer[0], query, referer=referer)
         elif 'playerKey' in query:
@@ -422,7 +449,7 @@ class BrightcoveLegacyIE(InfoExtractor):
         return info
 
 
-class BrightcoveNewIE(InfoExtractor):
+class BrightcoveNewIE(AdobePassIE):
     IE_NAME = 'brightcove:new'
     _VALID_URL = r'https?://players\.brightcove\.net/(?P<account_id>\d+)/(?P<player_id>[^/]+)_(?P<embed>[^/]+)/index\.html\?.*videoId=(?P<video_id>\d+|ref:[^&]+)'
     _TESTS = [{
@@ -437,7 +464,7 @@ class BrightcoveNewIE(InfoExtractor):
             'timestamp': 1441391203,
             'upload_date': '20150904',
             'uploader_id': '929656772001',
-            'formats': 'mincount:22',
+            'formats': 'mincount:20',
         },
     }, {
         # with rtmp streams
@@ -451,7 +478,7 @@ class BrightcoveNewIE(InfoExtractor):
             'timestamp': 1433556729,
             'upload_date': '20150606',
             'uploader_id': '4036320279001',
-            'formats': 'mincount:41',
+            'formats': 'mincount:39',
         },
         'params': {
             # m3u8 download
@@ -472,17 +499,18 @@ class BrightcoveNewIE(InfoExtractor):
     }]
 
     @staticmethod
-    def _extract_url(webpage):
-        urls = BrightcoveNewIE._extract_urls(webpage)
+    def _extract_url(ie, webpage):
+        urls = BrightcoveNewIE._extract_urls(ie, webpage)
         return urls[0] if urls else None
 
     @staticmethod
-    def _extract_urls(webpage):
+    def _extract_urls(ie, webpage):
         # Reference:
         # 1. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideoiniframe
-        # 2. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideousingjavascript
-        # 3. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/embed-in-page.html
-        # 4. https://support.brightcove.com/en/video-cloud/docs/dynamically-assigning-videos-player
+        # 2. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#tag
+        # 3. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/publish-video.html#setvideousingjavascript
+        # 4. http://docs.brightcove.com/en/video-cloud/brightcove-player/guides/in-page-embed-player-implementation.html
+        # 5. https://support.brightcove.com/en/video-cloud/docs/dynamically-assigning-videos-player
 
         entries = []
 
@@ -491,59 +519,52 @@ class BrightcoveNewIE(InfoExtractor):
                 r'<iframe[^>]+src=(["\'])((?:https?:)?//players\.brightcove\.net/\d+/[^/]+/index\.html.+?)\1', webpage):
             entries.append(url if url.startswith('http') else 'http:' + url)
 
-        # Look for embed_in_page embeds [2]
-        for video_id, account_id, player_id, embed in re.findall(
-                # According to examples from [3] it's unclear whether video id
-                # may be optional and what to do when it is
-                # According to [4] data-video-id may be prefixed with ref:
-                r'''(?sx)
-                    <video[^>]+
-                        data-video-id=["\'](\d+|ref:[^"\']+)["\'][^>]*>.*?
-                    </video>.*?
-                    <script[^>]+
-                        src=["\'](?:https?:)?//players\.brightcove\.net/
-                        (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
+        # Look for <video> tags [2] and embed_in_page embeds [3]
+        # [2] looks like:
+        for video, script_tag, account_id, player_id, embed in re.findall(
+                r'''(?isx)
+                    (<video\s+[^>]*\bdata-video-id\s*=\s*['"]?[^>]+>)
+                    (?:.*?
+                        (<script[^>]+
+                            src=["\'](?:https?:)?//players\.brightcove\.net/
+                            (\d+)/([^/]+)_([^/]+)/index(?:\.min)?\.js
+                        )
+                    )?
                 ''', webpage):
-            entries.append(
-                'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s'
-                % (account_id, player_id, embed, video_id))
+            attrs = extract_attributes(video)
+
+            # According to examples from [4] it's unclear whether video id
+            # may be optional and what to do when it is
+            video_id = attrs.get('data-video-id')
+            if not video_id:
+                continue
+
+            account_id = account_id or attrs.get('data-account')
+            if not account_id:
+                continue
+
+            player_id = player_id or attrs.get('data-player') or 'default'
+            embed = embed or attrs.get('data-embed') or 'default'
+
+            bc_url = 'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s' % (
+                account_id, player_id, embed, video_id)
+
+            # Some brightcove videos may be embedded with video tag only and
+            # without script tag or any mentioning of brightcove at all. Such
+            # embeds are considered ambiguous since they are matched based only
+            # on data-video-id and data-account attributes and in the wild may
+            # not be brightcove embeds at all. Let's check reconstructed
+            # brightcove URLs in case of such embeds and only process valid
+            # ones. By this we ensure there is indeed a brightcove embed.
+            if not script_tag and not ie._is_valid_url(
+                    bc_url, video_id, 'possible brightcove video'):
+                continue
+
+            entries.append(bc_url)
 
         return entries
 
-    def _real_extract(self, url):
-        account_id, player_id, embed, video_id = re.match(self._VALID_URL, url).groups()
-
-        webpage = self._download_webpage(
-            'http://players.brightcove.net/%s/%s_%s/index.min.js'
-            % (account_id, player_id, embed), video_id)
-
-        policy_key = None
-
-        catalog = self._search_regex(
-            r'catalog\(({.+?})\);', webpage, 'catalog', default=None)
-        if catalog:
-            catalog = self._parse_json(
-                js_to_json(catalog), video_id, fatal=False)
-            if catalog:
-                policy_key = catalog.get('policyKey')
-
-        if not policy_key:
-            policy_key = self._search_regex(
-                r'policyKey\s*:\s*(["\'])(?P<pk>.+?)\1',
-                webpage, 'policy key', group='pk')
-
-        api_url = 'https://edge.api.brightcove.com/playback/v1/accounts/%s/videos/%s' % (account_id, video_id)
-        try:
-            json_data = self._download_json(api_url, video_id, headers={
-                'Accept': 'application/json;pk=%s' % policy_key
-            })
-        except ExtractorError as e:
-            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
-                json_data = self._parse_json(e.cause.read().decode(), video_id)[0]
-                raise ExtractorError(
-                    json_data.get('message') or json_data['error_code'], expected=True)
-            raise
-
+    def _parse_brightcove_metadata(self, json_data, video_id):
         title = json_data['name'].strip()
 
         formats = []
@@ -626,7 +647,7 @@ class BrightcoveNewIE(InfoExtractor):
 
         is_live = False
         duration = float_or_none(json_data.get('duration'), 1000)
-        if duration and duration < 0:
+        if duration is not None and duration <= 0:
             is_live = True
 
         return {
@@ -636,9 +657,64 @@ class BrightcoveNewIE(InfoExtractor):
             'thumbnail': json_data.get('thumbnail') or json_data.get('poster'),
             'duration': duration,
             'timestamp': parse_iso8601(json_data.get('published_at')),
-            'uploader_id': account_id,
+            'uploader_id': json_data.get('account_id'),
             'formats': formats,
             'subtitles': subtitles,
             'tags': json_data.get('tags', []),
             'is_live': is_live,
         }
+
+    def _real_extract(self, url):
+        url, smuggled_data = unsmuggle_url(url, {})
+        self._initialize_geo_bypass(smuggled_data.get('geo_countries'))
+
+        account_id, player_id, embed, video_id = re.match(self._VALID_URL, url).groups()
+
+        webpage = self._download_webpage(
+            'http://players.brightcove.net/%s/%s_%s/index.min.js'
+            % (account_id, player_id, embed), video_id)
+
+        policy_key = None
+
+        catalog = self._search_regex(
+            r'catalog\(({.+?})\);', webpage, 'catalog', default=None)
+        if catalog:
+            catalog = self._parse_json(
+                js_to_json(catalog), video_id, fatal=False)
+            if catalog:
+                policy_key = catalog.get('policyKey')
+
+        if not policy_key:
+            policy_key = self._search_regex(
+                r'policyKey\s*:\s*(["\'])(?P<pk>.+?)\1',
+                webpage, 'policy key', group='pk')
+
+        api_url = 'https://edge.api.brightcove.com/playback/v1/accounts/%s/videos/%s' % (account_id, video_id)
+        try:
+            json_data = self._download_json(api_url, video_id, headers={
+                'Accept': 'application/json;pk=%s' % policy_key
+            })
+        except ExtractorError as e:
+            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
+                json_data = self._parse_json(e.cause.read().decode(), video_id)[0]
+                message = json_data.get('message') or json_data['error_code']
+                if json_data.get('error_subcode') == 'CLIENT_GEO':
+                    self.raise_geo_restricted(msg=message)
+                raise ExtractorError(message, expected=True)
+            raise
+
+        errors = json_data.get('errors')
+        if errors and errors[0].get('error_subcode') == 'TVE_AUTH':
+            custom_fields = json_data['custom_fields']
+            tve_token = self._extract_mvpd_auth(
+                smuggled_data['source_url'], video_id,
+                custom_fields['bcadobepassrequestorid'],
+                custom_fields['bcadobepassresourceid'])
+            json_data = self._download_json(
+                api_url, video_id, headers={
+                    'Accept': 'application/json;pk=%s' % policy_key
+                }, query={
+                    'tveToken': tve_token,
+                })
+
+        return self._parse_brightcove_metadata(json_data, video_id)

youtube_dl/extractor/buzzfeed.py

@@ -84,9 +84,10 @@ class BuzzFeedIE(InfoExtractor):
                 continue
             entries.append(self.url_result(video['url']))
 
-        facebook_url = FacebookIE._extract_url(webpage)
-        if facebook_url:
-            entries.append(self.url_result(facebook_url))
+        facebook_urls = FacebookIE._extract_urls(webpage)
+        entries.extend([
+            self.url_result(facebook_url)
+            for facebook_url in facebook_urls])
 
         return {
             '_type': 'playlist',

youtube_dl/extractor/byutv.py

@@ -3,20 +3,19 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..utils import ExtractorError
 
 
 class BYUtvIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?byutv\.org/watch/(?!event/)(?P<id>[0-9a-f-]+)(?:/(?P<display_id>[^/?#&]+))?'
+    _VALID_URL = r'https?://(?:www\.)?byutv\.org/(?:watch|player)/(?!event/)(?P<id>[0-9a-f-]+)(?:/(?P<display_id>[^/?#&]+))?'
     _TESTS = [{
         'url': 'http://www.byutv.org/watch/6587b9a3-89d2-42a6-a7f7-fd2f81840a7d/studio-c-season-5-episode-5',
         'info_dict': {
-            'id': '6587b9a3-89d2-42a6-a7f7-fd2f81840a7d',
+            'id': 'ZvanRocTpW-G5_yZFeltTAMv6jxOU9KH',
             'display_id': 'studio-c-season-5-episode-5',
             'ext': 'mp4',
             'title': 'Season 5 Episode 5',
-            'description': 'md5:e07269172baff037f8e8bf9956bc9747',
-            'thumbnail': r're:^https?://.*\.jpg$',
+            'description': 'md5:1d31dc18ef4f075b28f6a65937d22c65',
+            'thumbnail': r're:^https?://.*',
             'duration': 1486.486,
         },
         'params': {
@@ -26,6 +25,9 @@ class BYUtvIE(InfoExtractor):
     }, {
         'url': 'http://www.byutv.org/watch/6587b9a3-89d2-42a6-a7f7-fd2f81840a7d',
         'only_matching': True,
+    }, {
+        'url': 'https://www.byutv.org/player/27741493-dc83-40b0-8420-e7ae38a2ae98/byu-football-toledo-vs-byu-93016?listid=4fe0fee5-0d3c-4a29-b725-e4948627f472&listindex=0&q=toledo',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
@@ -33,16 +35,16 @@ class BYUtvIE(InfoExtractor):
         video_id = mobj.group('id')
         display_id = mobj.group('display_id') or video_id
 
-        webpage = self._download_webpage(url, display_id)
-        episode_code = self._search_regex(
-            r'(?s)episode:(.*?\}),\s*\n', webpage, 'episode information')
-
-        ep = self._parse_json(
-            episode_code, display_id, transform_source=lambda s:
-            re.sub(r'(\n\s+)([a-zA-Z]+):\s+\'(.*?)\'', r'\1"\2": "\3"', s))
-
-        if ep['providerType'] != 'Ooyala':
-            raise ExtractorError('Unsupported provider %s' % ep['provider'])
+        ep = self._download_json(
+            'https://api.byutv.org/api3/catalog/getvideosforcontent', video_id,
+            query={
+                'contentid': video_id,
+                'channel': 'byutv',
+                'x-byutv-context': 'web$US',
+            }, headers={
+                'x-byutv-context': 'web$US',
+                'x-byutv-platformkey': 'xsaaw9c7y5',
+            })['ooyalaVOD']
 
         return {
             '_type': 'url_transparent',
@@ -50,44 +52,7 @@ class BYUtvIE(InfoExtractor):
             'url': 'ooyala:%s' % ep['providerId'],
             'id': video_id,
             'display_id': display_id,
-            'title': ep['title'],
+            'title': ep.get('title'),
             'description': ep.get('description'),
             'thumbnail': ep.get('imageThumbnail'),
         }
-
-
-class BYUtvEventIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?byutv\.org/watch/event/(?P<id>[0-9a-f-]+)'
-    _TEST = {
-        'url': 'http://www.byutv.org/watch/event/29941b9b-8bf6-48d2-aebf-7a87add9e34b',
-        'info_dict': {
-            'id': '29941b9b-8bf6-48d2-aebf-7a87add9e34b',
-            'ext': 'mp4',
-            'title': 'Toledo vs. BYU (9/30/16)',
-        },
-        'params': {
-            'skip_download': True,
-        },
-        'add_ie': ['Ooyala'],
-    }
-
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-
-        webpage = self._download_webpage(url, video_id)
-
-        ooyala_id = self._search_regex(
-            r'providerId\s*:\s*(["\'])(?P<id>(?:(?!\1).)+)\1',
-            webpage, 'ooyala id', group='id')
-
-        title = self._search_regex(
-            r'class=["\']description["\'][^>]*>\s*<h1>([^<]+)</h1>', webpage,
-            'title').strip()
-
-        return {
-            '_type': 'url_transparent',
-            'ie_key': 'Ooyala',
-            'url': 'ooyala:%s' % ooyala_id,
-            'id': video_id,
-            'title': title,
-        }

youtube_dl/extractor/canalc2.py

@@ -16,13 +16,10 @@ class Canalc2IE(InfoExtractor):
         'md5': '060158428b650f896c542dfbb3d6487f',
         'info_dict': {
             'id': '12163',
-            'ext': 'flv',
+            'ext': 'mp4',
             'title': 'Terrasses du Numérique',
             'duration': 122,
         },
-        'params': {
-            'skip_download': True,  # Requires rtmpdump
-        }
     }, {
         'url': 'http://archives-canalc2.u-strasbg.fr/video.asp?idVideo=11427&voir=oui',
         'only_matching': True,

youtube_dl/extractor/canalplus.py

@@ -7,8 +7,8 @@ from .common import InfoExtractor
 from ..compat import compat_urllib_parse_urlparse
 from ..utils import (
     dict_get,
-    ExtractorError,
-    HEADRequest,
+    # ExtractorError,
+    # HEADRequest,
     int_or_none,
     qualities,
     remove_end,
@@ -27,6 +27,7 @@ class CanalplusIE(InfoExtractor):
                                     (?:www\.)?d8\.tv|
                                     (?:www\.)?c8\.fr|
                                     (?:www\.)?d17\.tv|
+                                    (?:(?:football|www)\.)?cstar\.fr|
                                     (?:www\.)?itele\.fr
                                 )/(?:(?:[^/]+/)*(?P<display_id>[^/?#&]+))?(?:\?.*\bvid=(?P<vid>\d+))?|
                                 player\.canalplus\.fr/#/(?P<id>\d+)
@@ -40,9 +41,13 @@ class CanalplusIE(InfoExtractor):
         'd8': 'd8',
         'c8': 'd8',
         'd17': 'd17',
+        'cstar': 'd17',
         'itele': 'itele',
     }
 
+    # Only works for direct mp4 URLs
+    _GEO_COUNTRIES = ['FR']
+
     _TESTS = [{
         'url': 'http://www.canalplus.fr/c-emissions/pid1830-c-zapping.html?vid=1192814',
         'info_dict': {
@@ -54,6 +59,7 @@ class CanalplusIE(InfoExtractor):
             'upload_date': '20160702',
         },
     }, {
+        # geo restricted, bypassed
         'url': 'http://www.piwiplus.fr/videos-piwi/pid1405-le-labyrinthe-boing-super-ranger.html?vid=1108190',
         'info_dict': {
             'id': '1108190',
@@ -63,19 +69,20 @@ class CanalplusIE(InfoExtractor):
             'description': 'md5:4cea7a37153be42c1ba2c1d3064376ff',
             'upload_date': '20140724',
         },
-        'skip': 'Only works from France',
+        'expected_warnings': ['HTTP Error 403: Forbidden'],
     }, {
-        'url': 'http://www.c8.fr/c8-divertissement/ms-touche-pas-a-mon-poste/pid6318-videos-integrales.html',
-        'md5': '4b47b12b4ee43002626b97fad8fb1de5',
+        # geo restricted, bypassed
+        'url': 'http://www.c8.fr/c8-divertissement/ms-touche-pas-a-mon-poste/pid6318-videos-integrales.html?vid=1443684',
+        'md5': 'bb6f9f343296ab7ebd88c97b660ecf8d',
         'info_dict': {
-            'id': '1420213',
+            'id': '1443684',
             'display_id': 'pid6318-videos-integrales',
             'ext': 'mp4',
-            'title': 'TPMP ! Même le matin - Les 35H de Baba - 14/10/2016',
-            'description': 'md5:f96736c1b0ffaa96fd5b9e60ad871799',
-            'upload_date': '20161014',
+            'title': 'Guess my iep ! - TPMP - 07/04/2017',
+            'description': 'md5:6f005933f6e06760a9236d9b3b5f17fa',
+            'upload_date': '20170407',
         },
-        'skip': 'Only works from France',
+        'expected_warnings': ['HTTP Error 403: Forbidden'],
     }, {
         'url': 'http://www.itele.fr/chroniques/invite-michael-darmon/rachida-dati-nicolas-sarkozy-est-le-plus-en-phase-avec-les-inquietudes-des-francais-171510',
         'info_dict': {
@@ -86,6 +93,19 @@ class CanalplusIE(InfoExtractor):
             'description': 'Chaque matin du lundi au vendredi, Michaël Darmon reçoit un invité politique à 8h25.',
             'upload_date': '20161014',
         },
+    }, {
+        'url': 'http://football.cstar.fr/cstar-minisite-foot/pid7566-feminines-videos.html?vid=1416769',
+        'info_dict': {
+            'id': '1416769',
+            'display_id': 'pid7566-feminines-videos',
+            'ext': 'mp4',
+            'title': 'France - Albanie : les temps forts de la soirée - 20/09/2016',
+            'description': 'md5:c3f30f2aaac294c1c969b3294de6904e',
+            'upload_date': '20160921',
+        },
+        'params': {
+            'skip_download': True,
+        },
     }, {
         'url': 'http://m.canalplus.fr/?vid=1398231',
         'only_matching': True,
@@ -107,7 +127,7 @@ class CanalplusIE(InfoExtractor):
             [r'<canal:player[^>]+?videoId=(["\'])(?P<id>\d+)',
              r'id=["\']canal_video_player(?P<id>\d+)',
              r'data-video=["\'](?P<id>\d+)'],
-            webpage, 'video id', group='id')
+            webpage, 'video id', default=mobj.group('vid'), group='id')
 
         info_url = self._VIDEO_INFO_TEMPLATE % (site_id, video_id)
         video_data = self._download_json(info_url, video_id, 'Downloading video JSON')
@@ -119,15 +139,15 @@ class CanalplusIE(InfoExtractor):
 
         preference = qualities(['MOBILE', 'BAS_DEBIT', 'HAUT_DEBIT', 'HD'])
 
-        fmt_url = next(iter(media.get('VIDEOS')))
-        if '/geo' in fmt_url.lower():
-            response = self._request_webpage(
-                HEADRequest(fmt_url), video_id,
-                'Checking if the video is georestricted')
-            if '/blocage' in response.geturl():
-                raise ExtractorError(
-                    'The video is not available in your country',
-                    expected=True)
+        # _, fmt_url = next(iter(media['VIDEOS'].items()))
+        # if '/geo' in fmt_url.lower():
+        #     response = self._request_webpage(
+        #         HEADRequest(fmt_url), video_id,
+        #         'Checking if the video is georestricted')
+        #     if '/blocage' in response.geturl():
+        #         raise ExtractorError(
+        #             'The video is not available in your country',
+        #             expected=True)
 
         formats = []
         for format_id, format_url in media['VIDEOS'].items():

youtube_dl/extractor/canvas.py

@@ -1,25 +1,112 @@
 from __future__ import unicode_literals
 
 import re
+import json
 
 from .common import InfoExtractor
-from ..utils import float_or_none
+from .gigya import GigyaBaseIE
+from ..compat import compat_HTTPError
+from ..utils import (
+    ExtractorError,
+    strip_or_none,
+    float_or_none,
+    int_or_none,
+    parse_iso8601,
+)
 
 
 class CanvasIE(InfoExtractor):
+    _VALID_URL = r'https?://mediazone\.vrt\.be/api/v1/(?P<site_id>canvas|een|ketnet|vrtvideo)/assets/(?P<id>[^/?#&]+)'
+    _TESTS = [{
+        'url': 'https://mediazone.vrt.be/api/v1/ketnet/assets/md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
+        'md5': '90139b746a0a9bd7bb631283f6e2a64e',
+        'info_dict': {
+            'id': 'md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
+            'display_id': 'md-ast-4ac54990-ce66-4d00-a8ca-9eac86f4c475',
+            'ext': 'flv',
+            'title': 'Nachtwacht: De Greystook',
+            'description': 'md5:1db3f5dc4c7109c821261e7512975be7',
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'duration': 1468.03,
+        },
+        'expected_warnings': ['is not a supported codec', 'Unknown MIME type'],
+    }, {
+        'url': 'https://mediazone.vrt.be/api/v1/canvas/assets/mz-ast-5e5f90b6-2d72-4c40-82c2-e134f884e93e',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        site_id, video_id = mobj.group('site_id'), mobj.group('id')
+
+        data = self._download_json(
+            'https://mediazone.vrt.be/api/v1/%s/assets/%s'
+            % (site_id, video_id), video_id)
+
+        title = data['title']
+        description = data.get('description')
+
+        formats = []
+        for target in data['targetUrls']:
+            format_url, format_type = target.get('url'), target.get('type')
+            if not format_url or not format_type:
+                continue
+            if format_type == 'HLS':
+                formats.extend(self._extract_m3u8_formats(
+                    format_url, video_id, 'mp4', entry_protocol='m3u8_native',
+                    m3u8_id=format_type, fatal=False))
+            elif format_type == 'HDS':
+                formats.extend(self._extract_f4m_formats(
+                    format_url, video_id, f4m_id=format_type, fatal=False))
+            elif format_type == 'MPEG_DASH':
+                formats.extend(self._extract_mpd_formats(
+                    format_url, video_id, mpd_id=format_type, fatal=False))
+            elif format_type == 'HSS':
+                formats.extend(self._extract_ism_formats(
+                    format_url, video_id, ism_id='mss', fatal=False))
+            else:
+                formats.append({
+                    'format_id': format_type,
+                    'url': format_url,
+                })
+        self._sort_formats(formats)
+
+        subtitles = {}
+        subtitle_urls = data.get('subtitleUrls')
+        if isinstance(subtitle_urls, list):
+            for subtitle in subtitle_urls:
+                subtitle_url = subtitle.get('url')
+                if subtitle_url and subtitle.get('type') == 'CLOSED':
+                    subtitles.setdefault('nl', []).append({'url': subtitle_url})
+
+        return {
+            'id': video_id,
+            'display_id': video_id,
+            'title': title,
+            'description': description,
+            'formats': formats,
+            'duration': float_or_none(data.get('duration'), 1000),
+            'thumbnail': data.get('posterImageUrl'),
+            'subtitles': subtitles,
+        }
+
+
+class CanvasEenIE(InfoExtractor):
+    IE_DESC = 'canvas.be and een.be'
     _VALID_URL = r'https?://(?:www\.)?(?P<site_id>canvas|een)\.be/(?:[^/]+/)*(?P<id>[^/?#&]+)'
     _TESTS = [{
         'url': 'http://www.canvas.be/video/de-afspraak/najaar-2015/de-afspraak-veilt-voor-de-warmste-week',
-        'md5': 'ea838375a547ac787d4064d8c7860a6c',
+        'md5': 'ed66976748d12350b118455979cca293',
         'info_dict': {
             'id': 'mz-ast-5e5f90b6-2d72-4c40-82c2-e134f884e93e',
             'display_id': 'de-afspraak-veilt-voor-de-warmste-week',
-            'ext': 'mp4',
+            'ext': 'flv',
             'title': 'De afspraak veilt voor de Warmste Week',
             'description': 'md5:24cb860c320dc2be7358e0e5aa317ba6',
             'thumbnail': r're:^https?://.*\.jpg$',
             'duration': 49.02,
-        }
+        },
+        'expected_warnings': ['is not a supported codec'],
     }, {
         # with subtitles
         'url': 'http://www.canvas.be/video/panorama/2016/pieter-0167',
@@ -39,7 +126,8 @@ class CanvasIE(InfoExtractor):
         },
         'params': {
             'skip_download': True,
-        }
+        },
+        'skip': 'Pagina niet gevonden',
     }, {
         'url': 'https://www.een.be/sorry-voor-alles/herbekijk-sorry-voor-alles',
         'info_dict': {
@@ -53,7 +141,8 @@ class CanvasIE(InfoExtractor):
         },
         'params': {
             'skip_download': True,
-        }
+        },
+        'skip': 'Episode no longer available',
     }, {
         'url': 'https://www.canvas.be/check-point/najaar-2016/de-politie-uw-vriend',
         'only_matching': True,
@@ -65,55 +154,157 @@ class CanvasIE(InfoExtractor):
 
         webpage = self._download_webpage(url, display_id)
 
-        title = (self._search_regex(
+        title = strip_or_none(self._search_regex(
             r'<h1[^>]+class="video__body__header__title"[^>]*>(.+?)</h1>',
             webpage, 'title', default=None) or self._og_search_title(
-            webpage)).strip()
+            webpage, default=None))
 
         video_id = self._html_search_regex(
-            r'data-video=(["\'])(?P<id>(?:(?!\1).)+)\1', webpage, 'video id', group='id')
-
-        data = self._download_json(
-            'https://mediazone.vrt.be/api/v1/%s/assets/%s'
-            % (site_id, video_id), display_id)
-
-        formats = []
-        for target in data['targetUrls']:
-            format_url, format_type = target.get('url'), target.get('type')
-            if not format_url or not format_type:
-                continue
-            if format_type == 'HLS':
-                formats.extend(self._extract_m3u8_formats(
-                    format_url, display_id, entry_protocol='m3u8_native',
-                    ext='mp4', preference=0, fatal=False, m3u8_id=format_type))
-            elif format_type == 'HDS':
-                formats.extend(self._extract_f4m_formats(
-                    format_url, display_id, f4m_id=format_type, fatal=False))
-            elif format_type == 'MPEG_DASH':
-                formats.extend(self._extract_mpd_formats(
-                    format_url, display_id, mpd_id=format_type, fatal=False))
-            else:
-                formats.append({
-                    'format_id': format_type,
-                    'url': format_url,
-                })
-        self._sort_formats(formats)
-
-        subtitles = {}
-        subtitle_urls = data.get('subtitleUrls')
-        if isinstance(subtitle_urls, list):
-            for subtitle in subtitle_urls:
-                subtitle_url = subtitle.get('url')
-                if subtitle_url and subtitle.get('type') == 'CLOSED':
-                    subtitles.setdefault('nl', []).append({'url': subtitle_url})
+            r'data-video=(["\'])(?P<id>(?:(?!\1).)+)\1', webpage, 'video id',
+            group='id')
 
         return {
+            '_type': 'url_transparent',
+            'url': 'https://mediazone.vrt.be/api/v1/%s/assets/%s' % (site_id, video_id),
+            'ie_key': CanvasIE.ie_key(),
             'id': video_id,
             'display_id': display_id,
             'title': title,
             'description': self._og_search_description(webpage),
-            'formats': formats,
-            'duration': float_or_none(data.get('duration'), 1000),
-            'thumbnail': data.get('posterImageUrl'),
-            'subtitles': subtitles,
+        }
+
+
+class VrtNUIE(GigyaBaseIE):
+    IE_DESC = 'VrtNU.be'
+    _VALID_URL = r'https?://(?:www\.)?vrt\.be/(?P<site_id>vrtnu)/(?:[^/]+/)*(?P<id>[^/?#&]+)'
+    _TESTS = [{
+        'url': 'https://www.vrt.be/vrtnu/a-z/postbus-x/1/postbus-x-s1a1/',
+        'info_dict': {
+            'id': 'pbs-pub-2e2d8c27-df26-45c9-9dc6-90c78153044d$vid-90c932b1-e21d-4fb8-99b1-db7b49cf74de',
+            'ext': 'flv',
+            'title': 'De zwarte weduwe',
+            'description': 'md5:d90c21dced7db869a85db89a623998d4',
+            'duration': 1457.04,
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'season': '1',
+            'season_number': 1,
+            'episode_number': 1,
+        },
+        'skip': 'This video is only available for registered users'
+    }]
+    _NETRC_MACHINE = 'vrtnu'
+    _APIKEY = '3_0Z2HujMtiWq_pkAjgnS2Md2E11a1AwZjYiBETtwNE-EoEHDINgtnvcAOpNgmrVGy'
+    _CONTEXT_ID = 'R3595707040'
+
+    def _real_initialize(self):
+        self._login()
+
+    def _login(self):
+        username, password = self._get_login_info()
+        if username is None:
+            return
+
+        auth_data = {
+            'APIKey': self._APIKEY,
+            'targetEnv': 'jssdk',
+            'loginID': username,
+            'password': password,
+            'authMode': 'cookie',
+        }
+
+        auth_info = self._gigya_login(auth_data)
+
+        # Sometimes authentication fails for no good reason, retry
+        login_attempt = 1
+        while login_attempt <= 3:
+            try:
+                # When requesting a token, no actual token is returned, but the
+                # necessary cookies are set.
+                self._request_webpage(
+                    'https://token.vrt.be',
+                    None, note='Requesting a token', errnote='Could not get a token',
+                    headers={
+                        'Content-Type': 'application/json',
+                        'Referer': 'https://www.vrt.be/vrtnu/',
+                    },
+                    data=json.dumps({
+                        'uid': auth_info['UID'],
+                        'uidsig': auth_info['UIDSignature'],
+                        'ts': auth_info['signatureTimestamp'],
+                        'email': auth_info['profile']['email'],
+                    }).encode('utf-8'))
+            except ExtractorError as e:
+                if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
+                    login_attempt += 1
+                    self.report_warning('Authentication failed')
+                    self._sleep(1, None, msg_template='Waiting for %(timeout)s seconds before trying again')
+                else:
+                    raise e
+            else:
+                break
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, display_id)
+
+        title = self._html_search_regex(
+            r'(?ms)<h1 class="content__heading">(.+?)</h1>',
+            webpage, 'title').strip()
+
+        description = self._html_search_regex(
+            r'(?ms)<div class="content__description">(.+?)</div>',
+            webpage, 'description', default=None)
+
+        season = self._html_search_regex(
+            [r'''(?xms)<div\ class="tabs__tab\ tabs__tab--active">\s*
+                    <span>seizoen\ (.+?)</span>\s*
+                </div>''',
+             r'<option value="seizoen (\d{1,3})" data-href="[^"]+?" selected>'],
+            webpage, 'season', default=None)
+
+        season_number = int_or_none(season)
+
+        episode_number = int_or_none(self._html_search_regex(
+            r'''(?xms)<div\ class="content__episode">\s*
+                    <abbr\ title="aflevering">afl</abbr>\s*<span>(\d+)</span>
+                </div>''',
+            webpage, 'episode_number', default=None))
+
+        release_date = parse_iso8601(self._html_search_regex(
+            r'(?ms)<div class="content__broadcastdate">\s*<time\ datetime="(.+?)"',
+            webpage, 'release_date', default=None))
+
+        # If there's a ? or a # in the URL, remove them and everything after
+        clean_url = url.split('?')[0].split('#')[0].strip('/')
+        securevideo_url = clean_url + '.mssecurevideo.json'
+
+        try:
+            video = self._download_json(securevideo_url, display_id)
+        except ExtractorError as e:
+            if isinstance(e.cause, compat_HTTPError) and e.cause.code == 401:
+                self.raise_login_required()
+            raise
+
+        # We are dealing with a '../<show>.relevant' URL
+        redirect_url = video.get('url')
+        if redirect_url:
+            return self.url_result(self._proto_relative_url(redirect_url, 'https:'))
+
+        # There is only one entry, but with an unknown key, so just get
+        # the first one
+        video_id = list(video.values())[0].get('videoid')
+
+        return {
+            '_type': 'url_transparent',
+            'url': 'https://mediazone.vrt.be/api/v1/vrtvideo/assets/%s' % video_id,
+            'ie_key': CanvasIE.ie_key(),
+            'id': video_id,
+            'display_id': display_id,
+            'title': title,
+            'description': description,
+            'season': season,
+            'season_number': season_number,
+            'episode_number': episode_number,
+            'release_date': release_date,
         }

youtube_dl/extractor/cartoonnetwork.py

@@ -31,7 +31,7 @@ class CartoonNetworkIE(TurnerBaseIE):
             'http://www.cartoonnetwork.com/video-seo-svc/episodeservices/getCvpPlaylist?networkName=CN2&' + query, video_id, {
                 'secure': {
                     'media_src': 'http://androidhls-secure.cdn.turner.com/toon/big',
-                    'tokenizer_src': 'http://www.cartoonnetwork.com/cntv/mvpd/processors/services/token_ipadAdobe.do',
+                    'tokenizer_src': 'https://token.vgtf.net/token/token_mobile',
                 },
             }, {
                 'url': url,

youtube_dl/extractor/cbc.py

@@ -90,36 +90,50 @@ class CBCIE(InfoExtractor):
             },
         }],
         'skip': 'Geo-restricted to Canada',
+    }, {
+        # multiple CBC.APP.Caffeine.initInstance(...)
+        'url': 'http://www.cbc.ca/news/canada/calgary/dog-indoor-exercise-winter-1.3928238',
+        'info_dict': {
+            'title': 'Keep Rover active during the deep freeze with doggie pushups and other fun indoor tasks',
+            'id': 'dog-indoor-exercise-winter-1.3928238',
+            'description': 'md5:c18552e41726ee95bd75210d1ca9194c',
+        },
+        'playlist_mincount': 6,
     }]
 
     @classmethod
     def suitable(cls, url):
         return False if CBCPlayerIE.suitable(url) else super(CBCIE, cls).suitable(url)
 
+    def _extract_player_init(self, player_init, display_id):
+        player_info = self._parse_json(player_init, display_id, js_to_json)
+        media_id = player_info.get('mediaId')
+        if not media_id:
+            clip_id = player_info['clipId']
+            feed = self._download_json(
+                'http://tpfeed.cbc.ca/f/ExhSPC/vms_5akSXx4Ng_Zn?byCustomValue={:mpsReleases}{%s}' % clip_id,
+                clip_id, fatal=False)
+            if feed:
+                media_id = try_get(feed, lambda x: x['entries'][0]['guid'], compat_str)
+            if not media_id:
+                media_id = self._download_json(
+                    'http://feed.theplatform.com/f/h9dtGB/punlNGjMlc1F?fields=id&byContent=byReleases%3DbyId%253D' + clip_id,
+                    clip_id)['entries'][0]['id'].split('/')[-1]
+        return self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
+
     def _real_extract(self, url):
         display_id = self._match_id(url)
         webpage = self._download_webpage(url, display_id)
-        player_init = self._search_regex(
-            r'CBC\.APP\.Caffeine\.initInstance\(({.+?})\);', webpage, 'player init',
-            default=None)
-        if player_init:
-            player_info = self._parse_json(player_init, display_id, js_to_json)
-            media_id = player_info.get('mediaId')
-            if not media_id:
-                clip_id = player_info['clipId']
-                feed = self._download_json(
-                    'http://tpfeed.cbc.ca/f/ExhSPC/vms_5akSXx4Ng_Zn?byCustomValue={:mpsReleases}{%s}' % clip_id,
-                    clip_id, fatal=False)
-                if feed:
-                    media_id = try_get(feed, lambda x: x['entries'][0]['guid'], compat_str)
-                if not media_id:
-                    media_id = self._download_json(
-                        'http://feed.theplatform.com/f/h9dtGB/punlNGjMlc1F?fields=id&byContent=byReleases%3DbyId%253D' + clip_id,
-                        clip_id)['entries'][0]['id'].split('/')[-1]
-            return self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
-        else:
-            entries = [self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id) for media_id in re.findall(r'<iframe[^>]+src="[^"]+?mediaId=(\d+)"', webpage)]
-            return self.playlist_result(entries)
+        entries = [
+            self._extract_player_init(player_init, display_id)
+            for player_init in re.findall(r'CBC\.APP\.Caffeine\.initInstance\(({.+?})\);', webpage)]
+        entries.extend([
+            self.url_result('cbcplayer:%s' % media_id, 'CBCPlayer', media_id)
+            for media_id in re.findall(r'<iframe[^>]+src="[^"]+?mediaId=(\d+)"', webpage)])
+        return self.playlist_result(
+            entries, display_id,
+            self._og_search_title(webpage, fatal=False),
+            self._og_search_description(webpage))
 
 
 class CBCPlayerIE(InfoExtractor):
@@ -152,12 +166,11 @@ class CBCPlayerIE(InfoExtractor):
             'uploader': 'CBCC-NEW',
         },
     }, {
-        # available only when we add `formats=MPEG4,FLV,MP3` to theplatform url
         'url': 'http://www.cbc.ca/player/play/2164402062',
-        'md5': '17a61eb813539abea40618d6323a7f82',
+        'md5': '33fcd8f6719b9dd60a5e73adcb83b9f6',
         'info_dict': {
             'id': '2164402062',
-            'ext': 'flv',
+            'ext': 'mp4',
             'title': 'Cancer survivor four times over',
             'description': 'Tim Mayer has beaten three different forms of cancer four times in five years.',
             'timestamp': 1320410746,
@@ -187,6 +200,7 @@ class CBCWatchBaseIE(InfoExtractor):
         'media': 'http://search.yahoo.com/mrss/',
         'clearleap': 'http://www.clearleap.com/namespace/clearleap/1.0/',
     }
+    _GEO_COUNTRIES = ['CA']
 
     def _call_api(self, path, video_id):
         url = path if path.startswith('http') else self._API_BASE_URL + path
@@ -274,6 +288,11 @@ class CBCWatchBaseIE(InfoExtractor):
 class CBCWatchVideoIE(CBCWatchBaseIE):
     IE_NAME = 'cbc.ca:watch:video'
     _VALID_URL = r'https?://api-cbc\.cloud\.clearleap\.com/cloffice/client/web/play/?\?.*?\bcontentId=(?P<id>[\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})'
+    _TEST = {
+        # geo-restricted to Canada, bypassable
+        'url': 'https://api-cbc.cloud.clearleap.com/cloffice/client/web/play/?contentId=3c84472a-1eea-4dee-9267-2655d5055dcf&categoryId=ebc258f5-ee40-4cca-b66b-ba6bd55b7235',
+        'only_matching': True,
+    }
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
@@ -283,6 +302,12 @@ class CBCWatchVideoIE(CBCWatchBaseIE):
         formats = self._extract_m3u8_formats(re.sub(r'/([^/]+)/[^/?]+\.m3u8', r'/\1/\1.m3u8', m3u8_url), video_id, 'mp4', fatal=False)
         if len(formats) < 2:
             formats = self._extract_m3u8_formats(m3u8_url, video_id, 'mp4')
+        for f in formats:
+            format_id = f.get('format_id')
+            if format_id.startswith('AAC'):
+                f['acodec'] = 'aac'
+            elif format_id.startswith('AC3'):
+                f['acodec'] = 'ac-3'
         self._sort_formats(formats)
 
         info = {
@@ -304,9 +329,10 @@ class CBCWatchIE(CBCWatchBaseIE):
     IE_NAME = 'cbc.ca:watch'
     _VALID_URL = r'https?://watch\.cbc\.ca/(?:[^/]+/)+(?P<id>[0-9a-f-]+)'
     _TESTS = [{
+        # geo-restricted to Canada, bypassable
         'url': 'http://watch.cbc.ca/doc-zone/season-6/customer-disservice/38e815a-009e3ab12e4',
         'info_dict': {
-            'id': '38e815a-009e3ab12e4',
+            'id': '9673749a-5e77-484c-8b62-a1092a6b5168',
             'ext': 'mp4',
             'title': 'Customer (Dis)Service',
             'description': 'md5:8bdd6913a0fe03d4b2a17ebe169c7c87',
@@ -318,8 +344,8 @@ class CBCWatchIE(CBCWatchBaseIE):
             'skip_download': True,
             'format': 'bestvideo',
         },
-        'skip': 'Geo-restricted to Canada',
     }, {
+        # geo-restricted to Canada, bypassable
         'url': 'http://watch.cbc.ca/arthur/all/1ed4b385-cd84-49cf-95f0-80f004680057',
         'info_dict': {
             'id': '1ed4b385-cd84-49cf-95f0-80f004680057',
@@ -327,7 +353,6 @@ class CBCWatchIE(CBCWatchBaseIE):
             'description': 'Arthur, the sweetest 8-year-old aardvark, and his pals solve all kinds of problems with humour, kindness and teamwork.',
         },
         'playlist_mincount': 30,
-        'skip': 'Geo-restricted to Canada',
     }]
 
     def _real_extract(self, url):

youtube_dl/extractor/cbs.py

@@ -49,13 +49,13 @@ class CBSIE(CBSBaseIE):
         'only_matching': True,
     }]
 
-    def _extract_video_info(self, content_id):
+    def _extract_video_info(self, content_id, site='cbs', mpx_acc=2198311517):
         items_data = self._download_xml(
             'http://can.cbs.com/thunder/player/videoPlayerService.php',
-            content_id, query={'partner': 'cbs', 'contentId': content_id})
+            content_id, query={'partner': site, 'contentId': content_id})
         video_data = xpath_element(items_data, './/item')
         title = xpath_text(video_data, 'videoTitle', 'title', True)
-        tp_path = 'dJ5BDC/media/guid/2198311517/%s' % content_id
+        tp_path = 'dJ5BDC/media/guid/%d/%s' % (mpx_acc, content_id)
         tp_release_url = 'http://link.theplatform.com/s/' + tp_path
 
         asset_types = []

youtube_dl/extractor/cbsinteractive.py

@@ -3,17 +3,18 @@ from __future__ import unicode_literals
 
 import re
 
-from .theplatform import ThePlatformIE
+from .cbs import CBSIE
 from ..utils import int_or_none
 
 
-class CBSInteractiveIE(ThePlatformIE):
-    _VALID_URL = r'https?://(?:www\.)?(?P<site>cnet|zdnet)\.com/(?:videos|video/share)/(?P<id>[^/?]+)'
+class CBSInteractiveIE(CBSIE):
+    _VALID_URL = r'https?://(?:www\.)?(?P<site>cnet|zdnet)\.com/(?:videos|video(?:/share)?)/(?P<id>[^/?]+)'
     _TESTS = [{
         'url': 'http://www.cnet.com/videos/hands-on-with-microsofts-windows-8-1-update/',
         'info_dict': {
-            'id': '56f4ea68-bd21-4852-b08c-4de5b8354c60',
-            'ext': 'flv',
+            'id': 'R49SYt__yAfmlXR85z4f7gNmCBDcN_00',
+            'display_id': 'hands-on-with-microsofts-windows-8-1-update',
+            'ext': 'mp4',
             'title': 'Hands-on with Microsoft Windows 8.1 Update',
             'description': 'The new update to the Windows 8 OS brings improved performance for mouse and keyboard users.',
             'uploader_id': '6085384d-619e-11e3-b231-14feb5ca9861',
@@ -22,13 +23,19 @@ class CBSInteractiveIE(ThePlatformIE):
             'timestamp': 1396479627,
             'upload_date': '20140402',
         },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
     }, {
         'url': 'http://www.cnet.com/videos/whiny-pothole-tweets-at-local-government-when-hit-by-cars-tomorrow-daily-187/',
+        'md5': 'f11d27b2fa18597fbf92444d2a9ed386',
         'info_dict': {
-            'id': '56527b93-d25d-44e3-b738-f989ce2e49ba',
-            'ext': 'flv',
+            'id': 'kjOJd_OoVJqbg_ZD8MZCOk8Wekb9QccK',
+            'display_id': 'whiny-pothole-tweets-at-local-government-when-hit-by-cars-tomorrow-daily-187',
+            'ext': 'mp4',
             'title': 'Whiny potholes tweet at local government when hit by cars (Tomorrow Daily 187)',
-            'description': 'Khail and Ashley wonder what other civic woes can be solved by self-tweeting objects, investigate a new kind of VR camera and watch an origami robot self-assemble, walk, climb, dig and dissolve. #TDPothole',
+            'description': 'md5:d2b9a95a5ffe978ae6fbd4cf944d618f',
             'uploader_id': 'b163284d-6b73-44fc-b3e6-3da66c392d40',
             'uploader': 'Ashley Esqueda',
             'duration': 1482,
@@ -38,23 +45,28 @@ class CBSInteractiveIE(ThePlatformIE):
     }, {
         'url': 'http://www.zdnet.com/video/share/video-keeping-android-smartphones-and-tablets-secure/',
         'info_dict': {
-            'id': 'bc1af9f0-a2b5-4e54-880d-0d95525781c0',
+            'id': 'k0r4T_ehht4xW_hAOqiVQPuBDPZ8SRjt',
+            'display_id': 'video-keeping-android-smartphones-and-tablets-secure',
             'ext': 'mp4',
             'title': 'Video: Keeping Android smartphones and tablets secure',
             'description': 'Here\'s the best way to keep Android devices secure, and what you do when they\'ve come to the end of their lives.',
             'uploader_id': 'f2d97ea2-8175-11e2-9d12-0018fe8a00b0',
             'uploader': 'Adrian Kingsley-Hughes',
-            'timestamp': 1448961720,
-            'upload_date': '20151201',
+            'duration': 731,
+            'timestamp': 1449129925,
+            'upload_date': '20151203',
         },
         'params': {
             # m3u8 download
             'skip_download': True,
-        }
+        },
+    }, {
+        'url': 'http://www.zdnet.com/video/huawei-matebook-x-video/',
+        'only_matching': True,
     }]
-    TP_RELEASE_URL_TEMPLATE = 'http://link.theplatform.com/s/kYEXFC/%s?mbr=true'
+
     MPX_ACCOUNTS = {
-        'cnet': 2288573011,
+        'cnet': 2198311517,
         'zdnet': 2387448114,
     }
 
@@ -68,7 +80,8 @@ class CBSInteractiveIE(ThePlatformIE):
         data = self._parse_json(data_json, display_id)
         vdata = data.get('video') or data['videos'][0]
 
-        video_id = vdata['id']
+        video_id = vdata['mpxRefId']
+
         title = vdata['title']
         author = vdata.get('author')
         if author:
@@ -78,20 +91,7 @@ class CBSInteractiveIE(ThePlatformIE):
             uploader = None
             uploader_id = None
 
-        media_guid_path = 'media/guid/%d/%s' % (self.MPX_ACCOUNTS[site], vdata['mpxRefId'])
-        formats, subtitles = [], {}
-        for (fkey, vid) in vdata['files'].items():
-            if fkey == 'hls_phone' and 'hls_tablet' in vdata['files']:
-                continue
-            release_url = self.TP_RELEASE_URL_TEMPLATE % vid
-            if fkey == 'hds':
-                release_url += '&manifest=f4m'
-            tp_formats, tp_subtitles = self._extract_theplatform_smil(release_url, video_id, 'Downloading %s SMIL data' % fkey)
-            formats.extend(tp_formats)
-            subtitles = self._merge_subtitles(subtitles, tp_subtitles)
-        self._sort_formats(formats)
-
-        info = self._extract_theplatform_metadata('kYEXFC/%s' % media_guid_path, video_id)
+        info = self._extract_video_info(video_id, site, self.MPX_ACCOUNTS[site])
         info.update({
             'id': video_id,
             'display_id': display_id,
@@ -99,7 +99,5 @@ class CBSInteractiveIE(ThePlatformIE):
             'duration': int_or_none(vdata.get('duration')),
             'uploader': uploader,
             'uploader_id': uploader_id,
-            'subtitles': subtitles,
-            'formats': formats,
         })
         return info

youtube_dl/extractor/cbslocal.py

@@ -60,8 +60,8 @@ class CBSLocalIE(AnvatoIE):
             'title': 'A Very Blue Anniversary',
             'description': 'CBS2’s Cindy Hsu has more.',
             'thumbnail': 're:^https?://.*',
-            'timestamp': 1479962220,
-            'upload_date': '20161124',
+            'timestamp': int,
+            'upload_date': r're:^\d{8}$',
             'uploader': 'CBS',
             'subtitles': {
                 'en': 'mincount:5',
@@ -91,12 +91,10 @@ class CBSLocalIE(AnvatoIE):
 
         info_dict = self._extract_anvato_videos(webpage, display_id)
 
-        time_str = self._html_search_regex(
-            r'class="entry-date">([^<]+)<', webpage, 'released date', default=None)
-        if time_str:
-            timestamp = unified_timestamp(time_str)
-        else:
-            timestamp = parse_iso8601(self._html_search_meta('uploadDate', webpage))
+        timestamp = unified_timestamp(self._html_search_regex(
+            r'class="(?:entry|post)-date"[^>]*>([^<]+)', webpage,
+            'released date', default=None)) or parse_iso8601(
+            self._html_search_meta('uploadDate', webpage))
 
         info_dict.update({
             'display_id': display_id,

youtube_dl/extractor/cbsnews.py

@@ -15,19 +15,23 @@ class CBSNewsIE(CBSIE):
 
     _TESTS = [
         {
-            'url': 'http://www.cbsnews.com/news/tesla-and-spacex-elon-musks-industrial-empire/',
+            # 60 minutes
+            'url': 'http://www.cbsnews.com/news/artificial-intelligence-positioned-to-be-a-game-changer/',
             'info_dict': {
-                'id': 'tesla-and-spacex-elon-musks-industrial-empire',
-                'ext': 'flv',
-                'title': 'Tesla and SpaceX: Elon Musk\'s industrial empire',
-                'thumbnail': 'http://beta.img.cbsnews.com/i/2014/03/30/60147937-2f53-4565-ad64-1bdd6eb64679/60-0330-pelley-640x360.jpg',
-                'duration': 791,
+                'id': '_B6Ga3VJrI4iQNKsir_cdFo9Re_YJHE_',
+                'ext': 'mp4',
+                'title': 'Artificial Intelligence',
+                'description': 'md5:8818145f9974431e0fb58a1b8d69613c',
+                'thumbnail': r're:^https?://.*\.jpg$',
+                'duration': 1606,
+                'uploader': 'CBSI-NEW',
+                'timestamp': 1498431900,
+                'upload_date': '20170625',
             },
             'params': {
-                # rtmp download
+                # m3u8 download
                 'skip_download': True,
             },
-            'skip': 'Subscribers only',
         },
         {
             'url': 'http://www.cbsnews.com/videos/fort-hood-shooting-army-downplays-mental-illness-as-cause-of-attack/',
@@ -52,6 +56,22 @@ class CBSNewsIE(CBSIE):
                 'skip_download': True,
             },
         },
+        {
+            # 48 hours
+            'url': 'http://www.cbsnews.com/news/maria-ridulph-murder-will-the-nations-oldest-cold-case-to-go-to-trial-ever-get-solved/',
+            'info_dict': {
+                'id': 'QpM5BJjBVEAUFi7ydR9LusS69DPLqPJ1',
+                'ext': 'mp4',
+                'title': 'Cold as Ice',
+                'description': 'Can a childhood memory of a friend\'s murder solve a 1957 cold case? "48 Hours" correspondent Erin Moriarty has the latest.',
+                'upload_date': '20170604',
+                'timestamp': 1496538000,
+                'uploader': 'CBSI-NEW',
+            },
+            'params': {
+                'skip_download': True,
+            },
+        },
     ]
 
     def _real_extract(self, url):
@@ -60,12 +80,18 @@ class CBSNewsIE(CBSIE):
         webpage = self._download_webpage(url, video_id)
 
         video_info = self._parse_json(self._html_search_regex(
-            r'(?:<ul class="media-list items" id="media-related-items"><li data-video-info|<div id="cbsNewsVideoPlayer" data-video-player-options)=\'({.+?})\'',
-            webpage, 'video JSON info'), video_id)
+            r'(?:<ul class="media-list items" id="media-related-items"[^>]*><li data-video-info|<div id="cbsNewsVideoPlayer" data-video-player-options)=\'({.+?})\'',
+            webpage, 'video JSON info', default='{}'), video_id, fatal=False)
 
-        item = video_info['item'] if 'item' in video_info else video_info
-        guid = item['mpxRefId']
-        return self._extract_video_info(guid)
+        if video_info:
+            item = video_info['item'] if 'item' in video_info else video_info
+        else:
+            state = self._parse_json(self._search_regex(
+                r'data-cbsvideoui-options=(["\'])(?P<json>{.+?})\1', webpage,
+                'playlist JSON info', group='json'), video_id)['state']
+            item = state['playlist'][state['pid']]
+
+        return self._extract_video_info(item['mpxRefId'], 'cbsnews')
 
 
 class CBSNewsLiveVideoIE(InfoExtractor):

youtube_dl/extractor/ccma.py

@@ -93,7 +93,7 @@ class CCMAIE(InfoExtractor):
             'description': clean_html(informacio.get('descripcio')),
             'duration': duration,
             'timestamp': timestamp,
-            'thumnails': thumbnails,
+            'thumbnails': thumbnails,
             'subtitles': subtitles,
             'formats': formats,
         }

youtube_dl/extractor/cda.py

@@ -1,6 +1,7 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import codecs
 import re
 
 from .common import InfoExtractor
@@ -8,7 +9,10 @@ from ..utils import (
     ExtractorError,
     float_or_none,
     int_or_none,
+    multipart_encode,
     parse_duration,
+    random_birthday,
+    urljoin,
 )
 
 
@@ -26,7 +30,8 @@ class CDAIE(InfoExtractor):
             'description': 'md5:269ccd135d550da90d1662651fcb9772',
             'thumbnail': r're:^https?://.*\.jpg$',
             'average_rating': float,
-            'duration': 39
+            'duration': 39,
+            'age_limit': 0,
         }
     }, {
         'url': 'http://www.cda.pl/video/57413289',
@@ -40,13 +45,41 @@ class CDAIE(InfoExtractor):
             'uploader': 'crash404',
             'view_count': int,
             'average_rating': float,
-            'duration': 137
+            'duration': 137,
+            'age_limit': 0,
         }
+    }, {
+        # Age-restricted
+        'url': 'http://www.cda.pl/video/1273454c4',
+        'info_dict': {
+            'id': '1273454c4',
+            'ext': 'mp4',
+            'title': 'Bronson (2008) napisy HD 1080p',
+            'description': 'md5:1b6cb18508daf2dc4e0fa4db77fec24c',
+            'height': 1080,
+            'uploader': 'boniek61',
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'duration': 5554,
+            'age_limit': 18,
+            'view_count': int,
+            'average_rating': float,
+        },
     }, {
         'url': 'http://ebd.cda.pl/0x0/5749950c',
         'only_matching': True,
     }]
 
+    def _download_age_confirm_page(self, url, video_id, *args, **kwargs):
+        form_data = random_birthday('rok', 'miesiac', 'dzien')
+        form_data.update({'return': url, 'module': 'video', 'module_id': video_id})
+        data, content_type = multipart_encode(form_data)
+        return self._download_webpage(
+            urljoin(url, '/a/validatebirth'), video_id, *args,
+            data=data, headers={
+                'Referer': url,
+                'Content-Type': content_type,
+            }, **kwargs)
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
         self._set_cookie('cda.pl', 'cda.player', 'html5')
@@ -56,6 +89,13 @@ class CDAIE(InfoExtractor):
         if 'Ten film jest dostępny dla użytkowników premium' in webpage:
             raise ExtractorError('This video is only available for premium users.', expected=True)
 
+        need_confirm_age = False
+        if self._html_search_regex(r'(<form[^>]+action="/a/validatebirth")',
+                                   webpage, 'birthday validate form', default=None):
+            webpage = self._download_age_confirm_page(
+                url, video_id, note='Confirming age')
+            need_confirm_age = True
+
         formats = []
 
         uploader = self._search_regex(r'''(?x)
@@ -80,10 +120,11 @@ class CDAIE(InfoExtractor):
             'thumbnail': self._og_search_thumbnail(webpage),
             'formats': formats,
             'duration': None,
+            'age_limit': 18 if need_confirm_age else 0,
         }
 
         def extract_format(page, version):
-            json_str = self._search_regex(
+            json_str = self._html_search_regex(
                 r'player_data=(\\?["\'])(?P<player_data>.+?)\1', page,
                 '%s player_json' % version, fatal=False, group='player_data')
             if not json_str:
@@ -96,6 +137,10 @@ class CDAIE(InfoExtractor):
             if not video or 'file' not in video:
                 self.report_warning('Unable to extract %s version information' % version)
                 return
+            if video['file'].startswith('uggc'):
+                video['file'] = codecs.decode(video['file'], 'rot_13')
+                if video['file'].endswith('adc.mp4'):
+                    video['file'] = video['file'].replace('adc.mp4', '.mp4')
             f = {
                 'url': video['file'],
             }
@@ -116,7 +161,12 @@ class CDAIE(InfoExtractor):
         for href, resolution in re.findall(
                 r'<a[^>]+data-quality="[^"]+"[^>]+href="([^"]+)"[^>]+class="quality-btn"[^>]*>([0-9]+p)',
                 webpage):
-            webpage = self._download_webpage(
+            if need_confirm_age:
+                handler = self._download_age_confirm_page
+            else:
+                handler = self._download_webpage
+
+            webpage = handler(
                 self._BASE_URL + href, video_id,
                 'Downloading %s version information' % resolution, fatal=False)
             if not webpage:
@@ -124,6 +174,7 @@ class CDAIE(InfoExtractor):
                 # invalid version is requested.
                 self.report_warning('Unable to download %s version information' % resolution)
                 continue
+
             extract_format(webpage, resolution)
 
         self._sort_formats(formats)

youtube_dl/extractor/ceskatelevize.py

@@ -12,19 +12,21 @@ from ..utils import (
     ExtractorError,
     float_or_none,
     sanitized_Request,
+    unescapeHTML,
     urlencode_postdata,
+    USER_AGENTS,
 )
 
 
 class CeskaTelevizeIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?ceskatelevize\.cz/(porady|ivysilani)/(?:[^/]+/)*(?P<id>[^/#?]+)/*(?:[#?].*)?$'
+    _VALID_URL = r'https?://(?:www\.)?ceskatelevize\.cz/ivysilani/(?:[^/?#&]+/)*(?P<id>[^/#?]+)'
     _TESTS = [{
         'url': 'http://www.ceskatelevize.cz/ivysilani/ivysilani/10441294653-hyde-park-civilizace/214411058091220',
         'info_dict': {
-            'id': '61924494876951776',
+            'id': '61924494877246241',
             'ext': 'mp4',
-            'title': 'Hyde Park Civilizace',
-            'description': 'md5:fe93f6eda372d150759d11644ebbfb4a',
+            'title': 'Hyde Park Civilizace: Život v Grónsku',
+            'description': 'md5:3fec8f6bb497be5cdb0c9e8781076626',
             'thumbnail': r're:^https?://.*\.jpg',
             'duration': 3350,
         },
@@ -61,40 +63,12 @@ class CeskaTelevizeIE(InfoExtractor):
         },
         'skip': 'Georestricted to Czech Republic',
     }, {
-        # video with 18+ caution trailer
-        'url': 'http://www.ceskatelevize.cz/porady/10520528904-queer/215562210900007-bogotart/',
-        'info_dict': {
-            'id': '215562210900007-bogotart',
-            'title': 'Queer: Bogotart',
-            'description': 'Alternativní průvodce současným queer světem',
-        },
-        'playlist': [{
-            'info_dict': {
-                'id': '61924494876844842',
-                'ext': 'mp4',
-                'title': 'Queer: Bogotart (Varování 18+)',
-                'duration': 10.2,
-            },
-        }, {
-            'info_dict': {
-                'id': '61924494877068022',
-                'ext': 'mp4',
-                'title': 'Queer: Bogotart (Queer)',
-                'thumbnail': r're:^https?://.*\.jpg',
-                'duration': 1558.3,
-            },
-        }],
-        'params': {
-            # m3u8 download
-            'skip_download': True,
-        },
+        'url': 'http://www.ceskatelevize.cz/ivysilani/embed/iFramePlayer.php?hash=d6a3e1370d2e4fa76296b90bad4dfc19673b641e&IDEC=217 562 22150/0004&channelID=1&width=100%25',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
-        url = url.replace('/porady/', '/ivysilani/').replace('/video/', '')
-
-        mobj = re.match(self._VALID_URL, url)
-        playlist_id = mobj.group('id')
+        playlist_id = self._match_id(url)
 
         webpage = self._download_webpage(url, playlist_id)
 
@@ -102,82 +76,126 @@ class CeskaTelevizeIE(InfoExtractor):
         if '%s</p>' % NOT_AVAILABLE_STRING in webpage:
             raise ExtractorError(NOT_AVAILABLE_STRING, expected=True)
 
-        typ = self._html_search_regex(
-            r'getPlaylistUrl\(\[\{"type":"(.+?)","id":".+?"\}\],', webpage, 'type')
-        episode_id = self._html_search_regex(
-            r'getPlaylistUrl\(\[\{"type":".+?","id":"(.+?)"\}\],', webpage, 'episode_id')
+        type_ = None
+        episode_id = None
+
+        playlist = self._parse_json(
+            self._search_regex(
+                r'getPlaylistUrl\(\[({.+?})\]', webpage, 'playlist',
+                default='{}'), playlist_id)
+        if playlist:
+            type_ = playlist.get('type')
+            episode_id = playlist.get('id')
+
+        if not type_:
+            type_ = self._html_search_regex(
+                r'getPlaylistUrl\(\[\{"type":"(.+?)","id":".+?"\}\],',
+                webpage, 'type')
+        if not episode_id:
+            episode_id = self._html_search_regex(
+                r'getPlaylistUrl\(\[\{"type":".+?","id":"(.+?)"\}\],',
+                webpage, 'episode_id')
 
         data = {
-            'playlist[0][type]': typ,
+            'playlist[0][type]': type_,
             'playlist[0][id]': episode_id,
             'requestUrl': compat_urllib_parse_urlparse(url).path,
             'requestSource': 'iVysilani',
         }
 
-        req = sanitized_Request(
-            'http://www.ceskatelevize.cz/ivysilani/ajax/get-client-playlist',
-            data=urlencode_postdata(data))
-
-        req.add_header('Content-type', 'application/x-www-form-urlencoded')
-        req.add_header('x-addr', '127.0.0.1')
-        req.add_header('X-Requested-With', 'XMLHttpRequest')
-        req.add_header('Referer', url)
-
-        playlistpage = self._download_json(req, playlist_id)
-
-        playlist_url = playlistpage['url']
-        if playlist_url == 'error_region':
-            raise ExtractorError(NOT_AVAILABLE_STRING, expected=True)
-
-        req = sanitized_Request(compat_urllib_parse_unquote(playlist_url))
-        req.add_header('Referer', url)
-
-        playlist_title = self._og_search_title(webpage, default=None)
-        playlist_description = self._og_search_description(webpage, default=None)
-
-        playlist = self._download_json(req, playlist_id)['playlist']
-        playlist_len = len(playlist)
-
         entries = []
-        for item in playlist:
-            is_live = item.get('type') == 'LIVE'
-            formats = []
-            for format_id, stream_url in item['streamUrls'].items():
-                formats.extend(self._extract_m3u8_formats(
-                    stream_url, playlist_id, 'mp4',
-                    entry_protocol='m3u8' if is_live else 'm3u8_native',
-                    fatal=False))
-            self._sort_formats(formats)
 
-            item_id = item.get('id') or item['assetId']
-            title = item['title']
+        for user_agent in (None, USER_AGENTS['Safari']):
+            req = sanitized_Request(
+                'http://www.ceskatelevize.cz/ivysilani/ajax/get-client-playlist',
+                data=urlencode_postdata(data))
 
-            duration = float_or_none(item.get('duration'))
-            thumbnail = item.get('previewImageUrl')
+            req.add_header('Content-type', 'application/x-www-form-urlencoded')
+            req.add_header('x-addr', '127.0.0.1')
+            req.add_header('X-Requested-With', 'XMLHttpRequest')
+            if user_agent:
+                req.add_header('User-Agent', user_agent)
+            req.add_header('Referer', url)
 
-            subtitles = {}
-            if item.get('type') == 'VOD':
-                subs = item.get('subtitles')
-                if subs:
-                    subtitles = self.extract_subtitles(episode_id, subs)
+            playlistpage = self._download_json(req, playlist_id, fatal=False)
 
-            if playlist_len == 1:
-                final_title = playlist_title or title
-                if is_live:
-                    final_title = self._live_title(final_title)
-            else:
-                final_title = '%s (%s)' % (playlist_title, title)
+            if not playlistpage:
+                continue
 
-            entries.append({
-                'id': item_id,
-                'title': final_title,
-                'description': playlist_description if playlist_len == 1 else None,
-                'thumbnail': thumbnail,
-                'duration': duration,
-                'formats': formats,
-                'subtitles': subtitles,
-                'is_live': is_live,
-            })
+            playlist_url = playlistpage['url']
+            if playlist_url == 'error_region':
+                raise ExtractorError(NOT_AVAILABLE_STRING, expected=True)
+
+            req = sanitized_Request(compat_urllib_parse_unquote(playlist_url))
+            req.add_header('Referer', url)
+
+            playlist_title = self._og_search_title(webpage, default=None)
+            playlist_description = self._og_search_description(webpage, default=None)
+
+            playlist = self._download_json(req, playlist_id, fatal=False)
+            if not playlist:
+                continue
+
+            playlist = playlist.get('playlist')
+            if not isinstance(playlist, list):
+                continue
+
+            playlist_len = len(playlist)
+
+            for num, item in enumerate(playlist):
+                is_live = item.get('type') == 'LIVE'
+                formats = []
+                for format_id, stream_url in item.get('streamUrls', {}).items():
+                    if 'playerType=flash' in stream_url:
+                        stream_formats = self._extract_m3u8_formats(
+                            stream_url, playlist_id, 'mp4', 'm3u8_native',
+                            m3u8_id='hls-%s' % format_id, fatal=False)
+                    else:
+                        stream_formats = self._extract_mpd_formats(
+                            stream_url, playlist_id,
+                            mpd_id='dash-%s' % format_id, fatal=False)
+                    # See https://github.com/rg3/youtube-dl/issues/12119#issuecomment-280037031
+                    if format_id == 'audioDescription':
+                        for f in stream_formats:
+                            f['source_preference'] = -10
+                    formats.extend(stream_formats)
+
+                if user_agent and len(entries) == playlist_len:
+                    entries[num]['formats'].extend(formats)
+                    continue
+
+                item_id = item.get('id') or item['assetId']
+                title = item['title']
+
+                duration = float_or_none(item.get('duration'))
+                thumbnail = item.get('previewImageUrl')
+
+                subtitles = {}
+                if item.get('type') == 'VOD':
+                    subs = item.get('subtitles')
+                    if subs:
+                        subtitles = self.extract_subtitles(episode_id, subs)
+
+                if playlist_len == 1:
+                    final_title = playlist_title or title
+                    if is_live:
+                        final_title = self._live_title(final_title)
+                else:
+                    final_title = '%s (%s)' % (playlist_title, title)
+
+                entries.append({
+                    'id': item_id,
+                    'title': final_title,
+                    'description': playlist_description if playlist_len == 1 else None,
+                    'thumbnail': thumbnail,
+                    'duration': duration,
+                    'formats': formats,
+                    'subtitles': subtitles,
+                    'is_live': is_live,
+                })
+
+        for e in entries:
+            self._sort_formats(e['formats'])
 
         return self.playlist_result(entries, playlist_id, playlist_title, playlist_description)
 
@@ -215,3 +233,47 @@ class CeskaTelevizeIE(InfoExtractor):
                     yield line
 
         return '\r\n'.join(_fix_subtitle(subtitles))
+
+
+class CeskaTelevizePoradyIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?ceskatelevize\.cz/porady/(?:[^/?#&]+/)*(?P<id>[^/#?]+)'
+    _TESTS = [{
+        # video with 18+ caution trailer
+        'url': 'http://www.ceskatelevize.cz/porady/10520528904-queer/215562210900007-bogotart/',
+        'info_dict': {
+            'id': '215562210900007-bogotart',
+            'title': 'Queer: Bogotart',
+            'description': 'Alternativní průvodce současným queer světem',
+        },
+        'playlist': [{
+            'info_dict': {
+                'id': '61924494876844842',
+                'ext': 'mp4',
+                'title': 'Queer: Bogotart (Varování 18+)',
+                'duration': 10.2,
+            },
+        }, {
+            'info_dict': {
+                'id': '61924494877068022',
+                'ext': 'mp4',
+                'title': 'Queer: Bogotart (Queer)',
+                'thumbnail': r're:^https?://.*\.jpg',
+                'duration': 1558.3,
+            },
+        }],
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, video_id)
+
+        data_url = unescapeHTML(self._search_regex(
+            r'<span[^>]*\bdata-url=(["\'])(?P<url>(?:(?!\1).)+)\1',
+            webpage, 'iframe player url', group='url'))
+
+        return self.url_result(data_url, ie=CeskaTelevizeIE.ie_key())

youtube_dl/extractor/channel9.py

@@ -4,62 +4,62 @@ import re
 
 from .common import InfoExtractor
 from ..utils import (
+    clean_html,
     ExtractorError,
-    parse_filesize,
+    int_or_none,
+    parse_iso8601,
     qualities,
+    unescapeHTML,
 )
 
 
 class Channel9IE(InfoExtractor):
-    '''
-    Common extractor for channel9.msdn.com.
-
-    The type of provided URL (video or playlist) is determined according to
-    meta Search.PageType from web page HTML rather than URL itself, as it is
-    not always possible to do.
-    '''
     IE_DESC = 'Channel 9'
     IE_NAME = 'channel9'
-    _VALID_URL = r'https?://(?:www\.)?channel9\.msdn\.com/(?P<contentpath>.+?)(?P<rss>/RSS)?/?(?:[?#&]|$)'
+    _VALID_URL = r'https?://(?:www\.)?(?:channel9\.msdn\.com|s\.ch9\.ms)/(?P<contentpath>.+?)(?P<rss>/RSS)?/?(?:[?#&]|$)'
 
     _TESTS = [{
         'url': 'http://channel9.msdn.com/Events/TechEd/Australia/2013/KOS002',
-        'md5': 'bbd75296ba47916b754e73c3a4bbdf10',
+        'md5': '32083d4eaf1946db6d454313f44510ca',
         'info_dict': {
-            'id': 'Events/TechEd/Australia/2013/KOS002',
-            'ext': 'mp4',
+            'id': '6c413323-383a-49dc-88f9-a22800cab024',
+            'ext': 'wmv',
             'title': 'Developer Kick-Off Session: Stuff We Love',
-            'description': 'md5:c08d72240b7c87fcecafe2692f80e35f',
+            'description': 'md5:b80bf9355a503c193aff7ec6cd5a7731',
             'duration': 4576,
-            'thumbnail': r're:http://.*\.jpg',
+            'thumbnail': r're:https?://.*\.jpg',
+            'timestamp': 1377717420,
+            'upload_date': '20130828',
             'session_code': 'KOS002',
-            'session_day': 'Day 1',
             'session_room': 'Arena 1A',
-            'session_speakers': ['Ed Blankenship', 'Andrew Coates', 'Brady Gaster', 'Patrick Klug',
-                                 'Mads Kristensen'],
+            'session_speakers': ['Andrew Coates', 'Brady Gaster', 'Mads Kristensen', 'Ed Blankenship', 'Patrick Klug'],
         },
     }, {
         'url': 'http://channel9.msdn.com/posts/Self-service-BI-with-Power-BI-nuclear-testing',
-        'md5': 'b43ee4529d111bc37ba7ee4f34813e68',
+        'md5': 'dcf983ee6acd2088e7188c3cf79b46bc',
         'info_dict': {
-            'id': 'posts/Self-service-BI-with-Power-BI-nuclear-testing',
-            'ext': 'mp4',
+            'id': 'fe8e435f-bb93-4e01-8e97-a28c01887024',
+            'ext': 'wmv',
             'title': 'Self-service BI with Power BI - nuclear testing',
-            'description': 'md5:d1e6ecaafa7fb52a2cacdf9599829f5b',
+            'description': 'md5:2d17fec927fc91e9e17783b3ecc88f54',
             'duration': 1540,
-            'thumbnail': r're:http://.*\.jpg',
+            'thumbnail': r're:https?://.*\.jpg',
+            'timestamp': 1386381991,
+            'upload_date': '20131207',
             'authors': ['Mike Wilmot'],
         },
     }, {
         # low quality mp4 is best
         'url': 'https://channel9.msdn.com/Events/CPP/CppCon-2015/Ranges-for-the-Standard-Library',
         'info_dict': {
-            'id': 'Events/CPP/CppCon-2015/Ranges-for-the-Standard-Library',
+            'id': '33ad69d2-6a4e-4172-83a1-a523013dec76',
             'ext': 'mp4',
             'title': 'Ranges for the Standard Library',
-            'description': 'md5:2e6b4917677af3728c5f6d63784c4c5d',
+            'description': 'md5:9895e0a9fd80822d2f01c454b8f4a372',
             'duration': 5646,
-            'thumbnail': r're:http://.*\.jpg',
+            'thumbnail': r're:https?://.*\.jpg',
+            'upload_date': '20150930',
+            'timestamp': 1443640735,
         },
         'params': {
             'skip_download': True,
@@ -70,7 +70,7 @@ class Channel9IE(InfoExtractor):
             'id': 'Niners/Splendid22/Queue/76acff796e8f411184b008028e0d492b',
             'title': 'Channel 9',
         },
-        'playlist_count': 2,
+        'playlist_mincount': 100,
     }, {
         'url': 'https://channel9.msdn.com/Events/DEVintersection/DEVintersection-2016/RSS',
         'only_matching': True,
@@ -81,188 +81,11 @@ class Channel9IE(InfoExtractor):
 
     _RSS_URL = 'http://channel9.msdn.com/%s/RSS'
 
-    def _formats_from_html(self, html):
-        FORMAT_REGEX = r'''
-            (?x)
-            <a\s+href="(?P<url>[^"]+)">(?P<quality>[^<]+)</a>\s*
-            <span\s+class="usage">\((?P<note>[^\)]+)\)</span>\s*
-            (?:<div\s+class="popup\s+rounded">\s*
-            <h3>File\s+size</h3>\s*(?P<filesize>.*?)\s*
-            </div>)?                                                # File size part may be missing
-        '''
-        quality = qualities((
-            'MP3', 'MP4',
-            'Low Quality WMV', 'Low Quality MP4',
-            'Mid Quality WMV', 'Mid Quality MP4',
-            'High Quality WMV', 'High Quality MP4'))
-        formats = [{
-            'url': x.group('url'),
-            'format_id': x.group('quality'),
-            'format_note': x.group('note'),
-            'format': '%s (%s)' % (x.group('quality'), x.group('note')),
-            'filesize_approx': parse_filesize(x.group('filesize')),
-            'quality': quality(x.group('quality')),
-            'vcodec': 'none' if x.group('note') == 'Audio only' else None,
-        } for x in list(re.finditer(FORMAT_REGEX, html))]
-
-        self._sort_formats(formats)
-
-        return formats
-
-    def _extract_title(self, html):
-        title = self._html_search_meta('title', html, 'title')
-        if title is None:
-            title = self._og_search_title(html)
-            TITLE_SUFFIX = ' (Channel 9)'
-            if title is not None and title.endswith(TITLE_SUFFIX):
-                title = title[:-len(TITLE_SUFFIX)]
-        return title
-
-    def _extract_description(self, html):
-        DESCRIPTION_REGEX = r'''(?sx)
-            <div\s+class="entry-content">\s*
-            <div\s+id="entry-body">\s*
-            (?P<description>.+?)\s*
-            </div>\s*
-            </div>
-        '''
-        m = re.search(DESCRIPTION_REGEX, html)
-        if m is not None:
-            return m.group('description')
-        return self._html_search_meta('description', html, 'description')
-
-    def _extract_duration(self, html):
-        m = re.search(r'"length": *"(?P<hours>\d{2}):(?P<minutes>\d{2}):(?P<seconds>\d{2})"', html)
-        return ((int(m.group('hours')) * 60 * 60) + (int(m.group('minutes')) * 60) + int(m.group('seconds'))) if m else None
-
-    def _extract_slides(self, html):
-        m = re.search(r'<a href="(?P<slidesurl>[^"]+)" class="slides">Slides</a>', html)
-        return m.group('slidesurl') if m is not None else None
-
-    def _extract_zip(self, html):
-        m = re.search(r'<a href="(?P<zipurl>[^"]+)" class="zip">Zip</a>', html)
-        return m.group('zipurl') if m is not None else None
-
-    def _extract_avg_rating(self, html):
-        m = re.search(r'<p class="avg-rating">Avg Rating: <span>(?P<avgrating>[^<]+)</span></p>', html)
-        return float(m.group('avgrating')) if m is not None else 0
-
-    def _extract_rating_count(self, html):
-        m = re.search(r'<div class="rating-count">\((?P<ratingcount>[^<]+)\)</div>', html)
-        return int(self._fix_count(m.group('ratingcount'))) if m is not None else 0
-
-    def _extract_view_count(self, html):
-        m = re.search(r'<li class="views">\s*<span class="count">(?P<viewcount>[^<]+)</span> Views\s*</li>', html)
-        return int(self._fix_count(m.group('viewcount'))) if m is not None else 0
-
-    def _extract_comment_count(self, html):
-        m = re.search(r'<li class="comments">\s*<a href="#comments">\s*<span class="count">(?P<commentcount>[^<]+)</span> Comments\s*</a>\s*</li>', html)
-        return int(self._fix_count(m.group('commentcount'))) if m is not None else 0
-
-    def _fix_count(self, count):
-        return int(str(count).replace(',', '')) if count is not None else None
-
-    def _extract_authors(self, html):
-        m = re.search(r'(?s)<li class="author">(.*?)</li>', html)
-        if m is None:
-            return None
-        return re.findall(r'<a href="/Niners/[^"]+">([^<]+)</a>', m.group(1))
-
-    def _extract_session_code(self, html):
-        m = re.search(r'<li class="code">\s*(?P<code>.+?)\s*</li>', html)
-        return m.group('code') if m is not None else None
-
-    def _extract_session_day(self, html):
-        m = re.search(r'<li class="day">\s*<a href="/Events/[^"]+">(?P<day>[^<]+)</a>\s*</li>', html)
-        return m.group('day').strip() if m is not None else None
-
-    def _extract_session_room(self, html):
-        m = re.search(r'<li class="room">\s*(?P<room>.+?)\s*</li>', html)
-        return m.group('room') if m is not None else None
-
-    def _extract_session_speakers(self, html):
-        return re.findall(r'<a href="/Events/Speakers/[^"]+">([^<]+)</a>', html)
-
-    def _extract_content(self, html, content_path):
-        # Look for downloadable content
-        formats = self._formats_from_html(html)
-        slides = self._extract_slides(html)
-        zip_ = self._extract_zip(html)
-
-        # Nothing to download
-        if len(formats) == 0 and slides is None and zip_ is None:
-            self._downloader.report_warning('None of recording, slides or zip are available for %s' % content_path)
-            return
-
-        # Extract meta
-        title = self._extract_title(html)
-        description = self._extract_description(html)
-        thumbnail = self._og_search_thumbnail(html)
-        duration = self._extract_duration(html)
-        avg_rating = self._extract_avg_rating(html)
-        rating_count = self._extract_rating_count(html)
-        view_count = self._extract_view_count(html)
-        comment_count = self._extract_comment_count(html)
-
-        common = {
-            '_type': 'video',
-            'id': content_path,
-            'description': description,
-            'thumbnail': thumbnail,
-            'duration': duration,
-            'avg_rating': avg_rating,
-            'rating_count': rating_count,
-            'view_count': view_count,
-            'comment_count': comment_count,
-        }
-
-        result = []
-
-        if slides is not None:
-            d = common.copy()
-            d.update({'title': title + '-Slides', 'url': slides})
-            result.append(d)
-
-        if zip_ is not None:
-            d = common.copy()
-            d.update({'title': title + '-Zip', 'url': zip_})
-            result.append(d)
-
-        if len(formats) > 0:
-            d = common.copy()
-            d.update({'title': title, 'formats': formats})
-            result.append(d)
-
-        return result
-
-    def _extract_entry_item(self, html, content_path):
-        contents = self._extract_content(html, content_path)
-        if contents is None:
-            return contents
-
-        if len(contents) > 1:
-            raise ExtractorError('Got more than one entry')
-        result = contents[0]
-        result['authors'] = self._extract_authors(html)
-
-        return result
-
-    def _extract_session(self, html, content_path):
-        contents = self._extract_content(html, content_path)
-        if contents is None:
-            return contents
-
-        session_meta = {
-            'session_code': self._extract_session_code(html),
-            'session_day': self._extract_session_day(html),
-            'session_room': self._extract_session_room(html),
-            'session_speakers': self._extract_session_speakers(html),
-        }
-
-        for content in contents:
-            content.update(session_meta)
-
-        return self.playlist_result(contents)
+    @staticmethod
+    def _extract_urls(webpage):
+        return re.findall(
+            r'<iframe[^>]+src=["\'](https?://channel9\.msdn\.com/(?:[^/]+/)+)player\b',
+            webpage)
 
     def _extract_list(self, video_id, rss_url=None):
         if not rss_url:
@@ -274,9 +97,7 @@ class Channel9IE(InfoExtractor):
         return self.playlist_result(entries, video_id, title_text)
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        content_path = mobj.group('contentpath')
-        rss = mobj.group('rss')
+        content_path, rss = re.match(self._VALID_URL, url).groups()
 
         if rss:
             return self._extract_list(content_path, url)
@@ -284,17 +105,158 @@ class Channel9IE(InfoExtractor):
         webpage = self._download_webpage(
             url, content_path, 'Downloading web page')
 
-        page_type = self._search_regex(
-            r'<meta[^>]+name=(["\'])WT\.entryid\1[^>]+content=(["\'])(?P<pagetype>[^:]+).+?\2',
-            webpage, 'page type', default=None, group='pagetype')
-        if page_type:
-            if page_type == 'Entry':      # Any 'item'-like page, may contain downloadable content
-                return self._extract_entry_item(webpage, content_path)
-            elif page_type == 'Session':  # Event session page, may contain downloadable content
-                return self._extract_session(webpage, content_path)
-            elif page_type == 'Event':
-                return self._extract_list(content_path)
+        episode_data = self._search_regex(
+            r"data-episode='([^']+)'", webpage, 'episode data', default=None)
+        if episode_data:
+            episode_data = self._parse_json(unescapeHTML(
+                episode_data), content_path)
+            content_id = episode_data['contentId']
+            is_session = '/Sessions(' in episode_data['api']
+            content_url = 'https://channel9.msdn.com/odata' + episode_data['api']
+            if is_session:
+                content_url += '?$expand=Speakers'
             else:
-                raise ExtractorError('Unexpected WT.entryid %s' % page_type, expected=True)
-        else:  # Assuming list
+                content_url += '?$expand=Authors'
+            content_data = self._download_json(content_url, content_id)
+            title = content_data['Title']
+
+            QUALITIES = (
+                'mp3',
+                'wmv', 'mp4',
+                'wmv-low', 'mp4-low',
+                'wmv-mid', 'mp4-mid',
+                'wmv-high', 'mp4-high',
+            )
+
+            quality_key = qualities(QUALITIES)
+
+            def quality(quality_id, format_url):
+                return (len(QUALITIES) if '_Source.' in format_url
+                        else quality_key(quality_id))
+
+            formats = []
+            urls = set()
+
+            SITE_QUALITIES = {
+                'MP3': 'mp3',
+                'MP4': 'mp4',
+                'Low Quality WMV': 'wmv-low',
+                'Low Quality MP4': 'mp4-low',
+                'Mid Quality WMV': 'wmv-mid',
+                'Mid Quality MP4': 'mp4-mid',
+                'High Quality WMV': 'wmv-high',
+                'High Quality MP4': 'mp4-high',
+            }
+
+            formats_select = self._search_regex(
+                r'(?s)<select[^>]+name=["\']format[^>]+>(.+?)</select', webpage,
+                'formats select', default=None)
+            if formats_select:
+                for mobj in re.finditer(
+                        r'<option\b[^>]+\bvalue=(["\'])(?P<url>(?:(?!\1).)+)\1[^>]*>\s*(?P<format>[^<]+?)\s*<',
+                        formats_select):
+                    format_url = mobj.group('url')
+                    if format_url in urls:
+                        continue
+                    urls.add(format_url)
+                    format_id = mobj.group('format')
+                    quality_id = SITE_QUALITIES.get(format_id, format_id)
+                    formats.append({
+                        'url': format_url,
+                        'format_id': quality_id,
+                        'quality': quality(quality_id, format_url),
+                        'vcodec': 'none' if quality_id == 'mp3' else None,
+                    })
+
+            API_QUALITIES = {
+                'VideoMP4Low': 'mp4-low',
+                'VideoWMV': 'wmv-mid',
+                'VideoMP4Medium': 'mp4-mid',
+                'VideoMP4High': 'mp4-high',
+                'VideoWMVHQ': 'wmv-hq',
+            }
+
+            for format_id, q in API_QUALITIES.items():
+                q_url = content_data.get(format_id)
+                if not q_url or q_url in urls:
+                    continue
+                urls.add(q_url)
+                formats.append({
+                    'url': q_url,
+                    'format_id': q,
+                    'quality': quality(q, q_url),
+                })
+
+            self._sort_formats(formats)
+
+            slides = content_data.get('Slides')
+            zip_file = content_data.get('ZipFile')
+
+            if not formats and not slides and not zip_file:
+                raise ExtractorError(
+                    'None of recording, slides or zip are available for %s' % content_path)
+
+            subtitles = {}
+            for caption in content_data.get('Captions', []):
+                caption_url = caption.get('Url')
+                if not caption_url:
+                    continue
+                subtitles.setdefault(caption.get('Language', 'en'), []).append({
+                    'url': caption_url,
+                    'ext': 'vtt',
+                })
+
+            common = {
+                'id': content_id,
+                'title': title,
+                'description': clean_html(content_data.get('Description') or content_data.get('Body')),
+                'thumbnail': content_data.get('Thumbnail') or content_data.get('VideoPlayerPreviewImage'),
+                'duration': int_or_none(content_data.get('MediaLengthInSeconds')),
+                'timestamp': parse_iso8601(content_data.get('PublishedDate')),
+                'avg_rating': int_or_none(content_data.get('Rating')),
+                'rating_count': int_or_none(content_data.get('RatingCount')),
+                'view_count': int_or_none(content_data.get('Views')),
+                'comment_count': int_or_none(content_data.get('CommentCount')),
+                'subtitles': subtitles,
+            }
+            if is_session:
+                speakers = []
+                for s in content_data.get('Speakers', []):
+                    speaker_name = s.get('FullName')
+                    if not speaker_name:
+                        continue
+                    speakers.append(speaker_name)
+
+                common.update({
+                    'session_code': content_data.get('Code'),
+                    'session_room': content_data.get('Room'),
+                    'session_speakers': speakers,
+                })
+            else:
+                authors = []
+                for a in content_data.get('Authors', []):
+                    author_name = a.get('DisplayName')
+                    if not author_name:
+                        continue
+                    authors.append(author_name)
+                common['authors'] = authors
+
+            contents = []
+
+            if slides:
+                d = common.copy()
+                d.update({'title': title + '-Slides', 'url': slides})
+                contents.append(d)
+
+            if zip_file:
+                d = common.copy()
+                d.update({'title': title + '-Zip', 'url': zip_file})
+                contents.append(d)
+
+            if formats:
+                d = common.copy()
+                d.update({'title': title, 'formats': formats})
+                contents.append(d)
+            return self.playlist_result(contents)
+        else:
             return self._extract_list(content_path)

youtube_dl/extractor/charlierose.py

@@ -5,7 +5,7 @@ from ..utils import remove_end
 
 
 class CharlieRoseIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?charlierose\.com/video(?:s|/player)/(?P<id>\d+)'
+    _VALID_URL = r'https?://(?:www\.)?charlierose\.com/(?:video|episode)(?:s|/player)/(?P<id>\d+)'
     _TESTS = [{
         'url': 'https://charlierose.com/videos/27996',
         'md5': 'fda41d49e67d4ce7c2411fd2c4702e09',
@@ -24,6 +24,9 @@ class CharlieRoseIE(InfoExtractor):
     }, {
         'url': 'https://charlierose.com/videos/27996',
         'only_matching': True,
+    }, {
+        'url': 'https://charlierose.com/episodes/30887?autoplay=true',
+        'only_matching': True,
     }]
 
     _PLAYER_BASE = 'https://charlierose.com/video/player/%s'

youtube_dl/extractor/chaturbate.py

@@ -1,5 +1,7 @@
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
 from ..utils import ExtractorError
 
@@ -31,30 +33,43 @@ class ChaturbateIE(InfoExtractor):
 
         webpage = self._download_webpage(url, video_id)
 
-        m3u8_url = self._search_regex(
-            r'src=(["\'])(?P<url>http.+?\.m3u8.*?)\1', webpage,
-            'playlist', default=None, group='url')
+        m3u8_urls = []
 
-        if not m3u8_url:
+        for m in re.finditer(
+                r'(["\'])(?P<url>http.+?\.m3u8.*?)\1', webpage):
+            m3u8_fast_url, m3u8_no_fast_url = m.group('url'), m.group(
+                'url').replace('_fast', '')
+            for m3u8_url in (m3u8_fast_url, m3u8_no_fast_url):
+                if m3u8_url not in m3u8_urls:
+                    m3u8_urls.append(m3u8_url)
+
+        if not m3u8_urls:
             error = self._search_regex(
                 [r'<span[^>]+class=(["\'])desc_span\1[^>]*>(?P<error>[^<]+)</span>',
                  r'<div[^>]+id=(["\'])defchat\1[^>]*>\s*<p><strong>(?P<error>[^<]+)<'],
                 webpage, 'error', group='error', default=None)
             if not error:
-                if any(p not in webpage for p in (
+                if any(p in webpage for p in (
                         self._ROOM_OFFLINE, 'offline_tipping', 'tip_offline')):
                     error = self._ROOM_OFFLINE
             if error:
                 raise ExtractorError(error, expected=True)
             raise ExtractorError('Unable to find stream URL')
 
-        formats = self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4')
+        formats = []
+        for m3u8_url in m3u8_urls:
+            m3u8_id = 'fast' if '_fast' in m3u8_url else 'slow'
+            formats.extend(self._extract_m3u8_formats(
+                m3u8_url, video_id, ext='mp4',
+                # ffmpeg skips segments for fast m3u8
+                preference=-10 if m3u8_id == 'fast' else None,
+                m3u8_id=m3u8_id, fatal=False, live=True))
         self._sort_formats(formats)
 
         return {
             'id': video_id,
             'title': self._live_title(video_id),
-            'thumbnail': 'https://cdn-s.highwebmedia.com/uHK3McUtGCG3SMFcd4ZJsRv8/roomimage/%s.jpg' % video_id,
+            'thumbnail': 'https://roomimg.stream.highwebmedia.com/ri/%s.jpg' % video_id,
             'age_limit': self._rta_search(webpage),
             'is_live': True,
             'formats': formats,

youtube_dl/extractor/chilloutzone.py

@@ -5,6 +5,7 @@ import base64
 import json
 
 from .common import InfoExtractor
+from .youtube import YoutubeIE
 from ..utils import (
     clean_html,
     ExtractorError
@@ -70,11 +71,9 @@ class ChilloutzoneIE(InfoExtractor):
 
         # If nativePlatform is None a fallback mechanism is used (i.e. youtube embed)
         if native_platform is None:
-            youtube_url = self._html_search_regex(
-                r'<iframe.* src="((?:https?:)?//(?:[^.]+\.)?youtube\.com/.+?)"',
-                webpage, 'fallback video URL', default=None)
-            if youtube_url is not None:
-                return self.url_result(youtube_url, ie='Youtube')
+            youtube_url = YoutubeIE._extract_url(webpage)
+            if youtube_url:
+                return self.url_result(youtube_url, ie=YoutubeIE.ie_key())
 
         # Non Fallback: Decide to use native source (e.g. youtube or vimeo) or
         # the own CDN

youtube_dl/extractor/chirbit.py

@@ -19,6 +19,7 @@ class ChirbitIE(InfoExtractor):
             'title': 'md5:f542ea253f5255240be4da375c6a5d7e',
             'description': 'md5:f24a4e22a71763e32da5fed59e47c770',
             'duration': 306,
+            'uploader': 'Gerryaudio',
         },
         'params': {
             'skip_download': True,
@@ -54,6 +55,9 @@ class ChirbitIE(InfoExtractor):
         duration = parse_duration(self._search_regex(
             r'class=["\']c-length["\'][^>]*>([^<]+)',
             webpage, 'duration', fatal=False))
+        uploader = self._search_regex(
+            r'id=["\']chirbit-username["\'][^>]*>([^<]+)',
+            webpage, 'uploader', fatal=False)
 
         return {
             'id': audio_id,
@@ -61,6 +65,7 @@ class ChirbitIE(InfoExtractor):
             'title': title,
             'description': description,
             'duration': duration,
+            'uploader': uploader,
         }
 
 

youtube_dl/extractor/cinchcast.py

@@ -9,12 +9,20 @@ from ..utils import (
 
 
 class CinchcastIE(InfoExtractor):
-    _VALID_URL = r'https?://player\.cinchcast\.com/.*?assetId=(?P<id>[0-9]+)'
-    _TEST = {
+    _VALID_URL = r'https?://player\.cinchcast\.com/.*?(?:assetId|show_id)=(?P<id>[0-9]+)'
+    _TESTS = [{
+        'url': 'http://player.cinchcast.com/?show_id=5258197&platformId=1&assetType=single',
+        'info_dict': {
+            'id': '5258197',
+            'ext': 'mp3',
+            'title': 'Train Your Brain to Up Your Game with Coach Mandy',
+            'upload_date': '20130816',
+        },
+    }, {
         # Actual test is run in generic, look for undergroundwellness
         'url': 'http://player.cinchcast.com/?platformId=1&#038;assetType=single&#038;assetId=7141703',
         'only_matching': True,
-    }
+    }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)

youtube_dl/extractor/cjsw.py

@@ -0,0 +1,72 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    determine_ext,
+    unescapeHTML,
+)
+
+
+class CJSWIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?cjsw\.com/program/(?P<program>[^/]+)/episode/(?P<id>\d+)'
+    _TESTS = [{
+        'url': 'http://cjsw.com/program/freshly-squeezed/episode/20170620',
+        'md5': 'cee14d40f1e9433632c56e3d14977120',
+        'info_dict': {
+            'id': '91d9f016-a2e7-46c5-8dcb-7cbcd7437c41',
+            'ext': 'mp3',
+            'title': 'Freshly Squeezed – Episode June 20, 2017',
+            'description': 'md5:c967d63366c3898a80d0c7b0ff337202',
+            'series': 'Freshly Squeezed',
+            'episode_id': '20170620',
+        },
+    }, {
+        # no description
+        'url': 'http://cjsw.com/program/road-pops/episode/20170707/',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        program, episode_id = mobj.group('program', 'id')
+        audio_id = '%s/%s' % (program, episode_id)
+
+        webpage = self._download_webpage(url, episode_id)
+
+        title = unescapeHTML(self._search_regex(
+            (r'<h1[^>]+class=["\']episode-header__title["\'][^>]*>(?P<title>[^<]+)',
+             r'data-audio-title=(["\'])(?P<title>(?:(?!\1).)+)\1'),
+            webpage, 'title', group='title'))
+
+        audio_url = self._search_regex(
+            r'<button[^>]+data-audio-src=(["\'])(?P<url>(?:(?!\1).)+)\1',
+            webpage, 'audio url', group='url')
+
+        audio_id = self._search_regex(
+            r'/([\da-f]{8}-[\da-f]{4}-[\da-f]{4}-[\da-f]{4}-[\da-f]{12})\.mp3',
+            audio_url, 'audio id', default=audio_id)
+
+        formats = [{
+            'url': audio_url,
+            'ext': determine_ext(audio_url, 'mp3'),
+            'vcodec': 'none',
+        }]
+
+        description = self._html_search_regex(
+            r'<p>(?P<description>.+?)</p>', webpage, 'description',
+            default=None)
+        series = self._search_regex(
+            r'data-showname=(["\'])(?P<name>(?:(?!\1).)+)\1', webpage,
+            'series', default=program, group='name')
+
+        return {
+            'id': audio_id,
+            'title': title,
+            'description': description,
+            'formats': formats,
+            'series': series,
+            'episode_id': episode_id,
+        }

youtube_dl/extractor/clipfish.py

@@ -1,67 +0,0 @@
-# coding: utf-8
-from __future__ import unicode_literals
-
-from .common import InfoExtractor
-from ..utils import (
-    int_or_none,
-    unified_strdate,
-)
-
-
-class ClipfishIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?clipfish\.de/(?:[^/]+/)+video/(?P<id>[0-9]+)'
-    _TEST = {
-        'url': 'http://www.clipfish.de/special/ugly-americans/video/4343170/s01-e01-ugly-americans-date-in-der-hoelle/',
-        'md5': '720563e467b86374c194bdead08d207d',
-        'info_dict': {
-            'id': '4343170',
-            'ext': 'mp4',
-            'title': 'S01 E01 - Ugly Americans - Date in der Hölle',
-            'description': 'Mark Lilly arbeitet im Sozialdienst der Stadt New York und soll Immigranten bei ihrer Einbürgerung in die USA zur Seite stehen.',
-            'upload_date': '20161005',
-            'duration': 1291,
-            'view_count': int,
-        }
-    }
-
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-
-        video_info = self._download_json(
-            'http://www.clipfish.de/devapi/id/%s?format=json&apikey=hbbtv' % video_id,
-            video_id)['items'][0]
-
-        formats = []
-
-        m3u8_url = video_info.get('media_videourl_hls')
-        if m3u8_url:
-            formats.append({
-                'url': m3u8_url.replace('de.hls.fra.clipfish.de', 'hls.fra.clipfish.de'),
-                'ext': 'mp4',
-                'format_id': 'hls',
-            })
-
-        mp4_url = video_info.get('media_videourl')
-        if mp4_url:
-            formats.append({
-                'url': mp4_url,
-                'format_id': 'mp4',
-                'width': int_or_none(video_info.get('width')),
-                'height': int_or_none(video_info.get('height')),
-                'tbr': int_or_none(video_info.get('bitrate')),
-            })
-
-        descr = video_info.get('descr')
-        if descr:
-            descr = descr.strip()
-
-        return {
-            'id': video_id,
-            'title': video_info['title'],
-            'description': descr,
-            'formats': formats,
-            'thumbnail': video_info.get('media_content_thumbnail_large') or video_info.get('media_thumbnail'),
-            'duration': int_or_none(video_info.get('media_length')),
-            'upload_date': unified_strdate(video_info.get('pubDate')),
-            'view_count': int_or_none(video_info.get('media_views'))
-        }

youtube_dl/extractor/clippit.py

@@ -0,0 +1,74 @@
+# coding: utf-8
+
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    parse_iso8601,
+    qualities,
+)
+
+import re
+
+
+class ClippitIE(InfoExtractor):
+
+    _VALID_URL = r'https?://(?:www\.)?clippituser\.tv/c/(?P<id>[a-z]+)'
+    _TEST = {
+        'url': 'https://www.clippituser.tv/c/evmgm',
+        'md5': '963ae7a59a2ec4572ab8bf2f2d2c5f09',
+        'info_dict': {
+            'id': 'evmgm',
+            'ext': 'mp4',
+            'title': 'Bye bye Brutus. #BattleBots  - Clippit',
+            'uploader': 'lizllove',
+            'uploader_url': 'https://www.clippituser.tv/p/lizllove',
+            'timestamp': 1472183818,
+            'upload_date': '20160826',
+            'description': 'BattleBots | ABC',
+            'thumbnail': r're:^https?://.*\.jpg$',
+        }
+    }
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        title = self._html_search_regex(r'<title.*>(.+?)</title>', webpage, 'title')
+
+        FORMATS = ('sd', 'hd')
+        quality = qualities(FORMATS)
+        formats = []
+        for format_id in FORMATS:
+            url = self._html_search_regex(r'data-%s-file="(.+?)"' % format_id,
+                                          webpage, 'url', fatal=False)
+            if not url:
+                continue
+            match = re.search(r'/(?P<height>\d+)\.mp4', url)
+            formats.append({
+                'url': url,
+                'format_id': format_id,
+                'quality': quality(format_id),
+                'height': int(match.group('height')) if match else None,
+            })
+
+        uploader = self._html_search_regex(r'class="username".*>\s+(.+?)\n',
+                                           webpage, 'uploader', fatal=False)
+        uploader_url = ('https://www.clippituser.tv/p/' + uploader
+                        if uploader else None)
+
+        timestamp = self._html_search_regex(r'datetime="(.+?)"',
+                                            webpage, 'date', fatal=False)
+        thumbnail = self._html_search_regex(r'data-image="(.+?)"',
+                                            webpage, 'thumbnail', fatal=False)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'formats': formats,
+            'uploader': uploader,
+            'uploader_url': uploader_url,
+            'timestamp': parse_iso8601(timestamp),
+            'description': self._og_search_description(webpage),
+            'thumbnail': thumbnail,
+        }

youtube_dl/extractor/cloudy.py

@@ -1,97 +1,60 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
-from ..compat import (
-    compat_parse_qs,
-    compat_HTTPError,
-)
 from ..utils import (
-    ExtractorError,
-    HEADRequest,
-    remove_end,
+    str_to_int,
+    unified_strdate,
 )
 
 
 class CloudyIE(InfoExtractor):
     _IE_DESC = 'cloudy.ec'
-    _VALID_URL = r'''(?x)
-        https?://(?:www\.)?cloudy\.ec/
-        (?:v/|embed\.php\?id=)
-        (?P<id>[A-Za-z0-9]+)
-        '''
-    _EMBED_URL = 'http://www.cloudy.ec/embed.php?id=%s'
-    _API_URL = 'http://www.cloudy.ec/api/player.api.php'
-    _MAX_TRIES = 2
-    _TEST = {
+    _VALID_URL = r'https?://(?:www\.)?cloudy\.ec/(?:v/|embed\.php\?.*?\bid=)(?P<id>[A-Za-z0-9]+)'
+    _TESTS = [{
         'url': 'https://www.cloudy.ec/v/af511e2527aac',
-        'md5': '5cb253ace826a42f35b4740539bedf07',
+        'md5': '29832b05028ead1b58be86bf319397ca',
         'info_dict': {
             'id': 'af511e2527aac',
-            'ext': 'flv',
+            'ext': 'mp4',
             'title': 'Funny Cats and Animals Compilation june 2013',
+            'upload_date': '20130913',
+            'view_count': int,
         }
-    }
-
-    def _extract_video(self, video_id, file_key, error_url=None, try_num=0):
-
-        if try_num > self._MAX_TRIES - 1:
-            raise ExtractorError('Unable to extract video URL', expected=True)
-
-        form = {
-            'file': video_id,
-            'key': file_key,
-        }
-
-        if error_url:
-            form.update({
-                'numOfErrors': try_num,
-                'errorCode': '404',
-                'errorUrl': error_url,
-            })
-
-        player_data = self._download_webpage(
-            self._API_URL, video_id, 'Downloading player data', query=form)
-        data = compat_parse_qs(player_data)
-
-        try_num += 1
-
-        if 'error' in data:
-            raise ExtractorError(
-                '%s error: %s' % (self.IE_NAME, ' '.join(data['error_msg'])),
-                expected=True)
-
-        title = data.get('title', [None])[0]
-        if title:
-            title = remove_end(title, '&asdasdas').strip()
-
-        video_url = data.get('url', [None])[0]
-
-        if video_url:
-            try:
-                self._request_webpage(HEADRequest(video_url), video_id, 'Checking video URL')
-            except ExtractorError as e:
-                if isinstance(e.cause, compat_HTTPError) and e.cause.code in [404, 410]:
-                    self.report_warning('Invalid video URL, requesting another', video_id)
-                    return self._extract_video(video_id, file_key, video_url, try_num)
-
-        return {
-            'id': video_id,
-            'url': video_url,
-            'title': title,
-        }
+    }, {
+        'url': 'http://www.cloudy.ec/embed.php?autoplay=1&id=af511e2527aac',
+        'only_matching': True,
+    }]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
+        video_id = self._match_id(url)
 
-        url = self._EMBED_URL % video_id
-        webpage = self._download_webpage(url, video_id)
+        webpage = self._download_webpage(
+            'https://www.cloudy.ec/embed.php', video_id, query={
+                'id': video_id,
+                'playerPage': 1,
+                'autoplay': 1,
+            })
 
-        file_key = self._search_regex(
-            [r'key\s*:\s*"([^"]+)"', r'filekey\s*=\s*"([^"]+)"'],
-            webpage, 'file_key')
+        info = self._parse_html5_media_entries(url, webpage, video_id)[0]
 
-        return self._extract_video(video_id, file_key)
+        webpage = self._download_webpage(
+            'https://www.cloudy.ec/v/%s' % video_id, video_id, fatal=False)
+
+        if webpage:
+            info.update({
+                'title': self._search_regex(
+                    r'<h\d[^>]*>([^<]+)<', webpage, 'title'),
+                'upload_date': unified_strdate(self._search_regex(
+                    r'>Published at (\d{4}-\d{1,2}-\d{1,2})', webpage,
+                    'upload date', fatal=False)),
+                'view_count': str_to_int(self._search_regex(
+                    r'([\d,.]+) views<', webpage, 'view count', fatal=False)),
+            })
+
+        if not info.get('title'):
+            info['title'] = video_id
+
+        info['id'] = video_id
+
+        return info

youtube_dl/extractor/cmt.py

@@ -1,13 +1,11 @@
 from __future__ import unicode_literals
 
 from .mtv import MTVIE
-from ..utils import ExtractorError
 
 
 class CMTIE(MTVIE):
     IE_NAME = 'cmt.com'
-    _VALID_URL = r'https?://(?:www\.)?cmt\.com/(?:videos|shows)/(?:[^/]+/)*(?P<videoid>\d+)'
-    _FEED_URL = 'http://www.cmt.com/sitewide/apps/player/embed/rss/'
+    _VALID_URL = r'https?://(?:www\.)?cmt\.com/(?:videos|shows|(?:full-)?episodes|video-clips)/(?P<id>[^/]+)'
 
     _TESTS = [{
         'url': 'http://www.cmt.com/videos/garth-brooks/989124/the-call-featuring-trisha-yearwood.jhtml#artist=30061',
@@ -33,17 +31,24 @@ class CMTIE(MTVIE):
     }, {
         'url': 'http://www.cmt.com/shows/party-down-south/party-down-south-ep-407-gone-girl/1738172/playlist/#id=1738172',
         'only_matching': True,
+    }, {
+        'url': 'http://www.cmt.com/full-episodes/537qb3/nashville-the-wayfaring-stranger-season-5-ep-501',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.cmt.com/video-clips/t9e4ci/nashville-juliette-in-2-minutes',
+        'only_matching': True,
     }]
 
-    @classmethod
-    def _transform_rtmp_url(cls, rtmp_video_url):
-        if 'error_not_available.swf' in rtmp_video_url:
-            raise ExtractorError(
-                '%s said: video is not available' % cls.IE_NAME, expected=True)
-
-        return super(CMTIE, cls)._transform_rtmp_url(rtmp_video_url)
-
     def _extract_mgid(self, webpage):
-        return self._search_regex(
+        mgid = self._search_regex(
             r'MTVN\.VIDEO\.contentUri\s*=\s*([\'"])(?P<mgid>.+?)\1',
-            webpage, 'mgid', group='mgid')
+            webpage, 'mgid', group='mgid', default=None)
+        if not mgid:
+            mgid = self._extract_triforce_mgid(webpage)
+        return mgid
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        mgid = self._extract_mgid(webpage)
+        return self.url_result('http://media.mtvnservices.com/embed/%s' % mgid)

youtube_dl/extractor/collegerama.py

@@ -1,92 +0,0 @@
-from __future__ import unicode_literals
-
-import json
-
-from .common import InfoExtractor
-from ..utils import (
-    float_or_none,
-    int_or_none,
-    sanitized_Request,
-)
-
-
-class CollegeRamaIE(InfoExtractor):
-    _VALID_URL = r'https?://collegerama\.tudelft\.nl/Mediasite/Play/(?P<id>[\da-f]+)'
-    _TESTS = [
-        {
-            'url': 'https://collegerama.tudelft.nl/Mediasite/Play/585a43626e544bdd97aeb71a0ec907a01d',
-            'md5': '481fda1c11f67588c0d9d8fbdced4e39',
-            'info_dict': {
-                'id': '585a43626e544bdd97aeb71a0ec907a01d',
-                'ext': 'mp4',
-                'title': 'Een nieuwe wereld: waarden, bewustzijn en techniek van de mensheid 2.0.',
-                'description': '',
-                'thumbnail': r're:^https?://.*\.jpg$',
-                'duration': 7713.088,
-                'timestamp': 1413309600,
-                'upload_date': '20141014',
-            },
-        },
-        {
-            'url': 'https://collegerama.tudelft.nl/Mediasite/Play/86a9ea9f53e149079fbdb4202b521ed21d?catalog=fd32fd35-6c99-466c-89d4-cd3c431bc8a4',
-            'md5': 'ef1fdded95bdf19b12c5999949419c92',
-            'info_dict': {
-                'id': '86a9ea9f53e149079fbdb4202b521ed21d',
-                'ext': 'wmv',
-                'title': '64ste Vakantiecursus: Afvalwater',
-                'description': 'md5:7fd774865cc69d972f542b157c328305',
-                'duration': 10853,
-                'timestamp': 1326446400,
-                'upload_date': '20120113',
-            },
-        },
-    ]
-
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-
-        player_options_request = {
-            'getPlayerOptionsRequest': {
-                'ResourceId': video_id,
-                'QueryString': '',
-            }
-        }
-
-        request = sanitized_Request(
-            'http://collegerama.tudelft.nl/Mediasite/PlayerService/PlayerService.svc/json/GetPlayerOptions',
-            json.dumps(player_options_request))
-        request.add_header('Content-Type', 'application/json')
-
-        player_options = self._download_json(request, video_id)
-
-        presentation = player_options['d']['Presentation']
-        title = presentation['Title']
-        description = presentation.get('Description')
-        thumbnail = None
-        duration = float_or_none(presentation.get('Duration'), 1000)
-        timestamp = int_or_none(presentation.get('UnixTime'), 1000)
-
-        formats = []
-        for stream in presentation['Streams']:
-            for video in stream['VideoUrls']:
-                thumbnail_url = stream.get('ThumbnailUrl')
-                if thumbnail_url:
-                    thumbnail = 'http://collegerama.tudelft.nl' + thumbnail_url
-                format_id = video['MediaType']
-                if format_id == 'SS':
-                    continue
-                formats.append({
-                    'url': video['Location'],
-                    'format_id': format_id,
-                })
-        self._sort_formats(formats)
-
-        return {
-            'id': video_id,
-            'title': title,
-            'description': description,
-            'thumbnail': thumbnail,
-            'duration': duration,
-            'timestamp': timestamp,
-            'formats': formats,
-        }

youtube_dl/extractor/comedycentral.py

@@ -48,17 +48,8 @@ class ComedyCentralFullEpisodesIE(MTVServicesInfoExtractor):
     def _real_extract(self, url):
         playlist_id = self._match_id(url)
         webpage = self._download_webpage(url, playlist_id)
-
-        feed_json = self._search_regex(r'var triforceManifestFeed\s*=\s*(\{.+?\});\n', webpage, 'triforce feeed')
-        feed = self._parse_json(feed_json, playlist_id)
-        zones = feed['manifest']['zones']
-
-        video_zone = zones['t2_lc_promo1']
-        feed = self._download_json(video_zone['feed'], playlist_id)
-        mgid = feed['result']['data']['id']
-
+        mgid = self._extract_triforce_mgid(webpage, data_zone='t2_lc_promo1')
         videos_info = self._get_videos_info(mgid)
-
         return videos_info
 
 
@@ -94,12 +85,6 @@ class ToshIE(MTVServicesInfoExtractor):
         'only_matching': True,
     }]
 
-    @classmethod
-    def _transform_rtmp_url(cls, rtmp_video_url):
-        new_urls = super(ToshIE, cls)._transform_rtmp_url(rtmp_video_url)
-        new_urls['rtmp'] = rtmp_video_url.replace('viacomccstrm', 'viacommtvstrm')
-        return new_urls
-
 
 class ComedyCentralTVIE(MTVServicesInfoExtractor):
     _VALID_URL = r'https?://(?:www\.)?comedycentral\.tv/(?:staffeln|shows)/(?P<id>[^/?#&]+)'
@@ -135,13 +120,16 @@ class ComedyCentralTVIE(MTVServicesInfoExtractor):
 
 
 class ComedyCentralShortnameIE(InfoExtractor):
-    _VALID_URL = r'^:(?P<id>tds|thedailyshow)$'
+    _VALID_URL = r'^:(?P<id>tds|thedailyshow|theopposition)$'
     _TESTS = [{
         'url': ':tds',
         'only_matching': True,
     }, {
         'url': ':thedailyshow',
         'only_matching': True,
+    }, {
+        'url': ':theopposition',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
@@ -149,5 +137,6 @@ class ComedyCentralShortnameIE(InfoExtractor):
         shortcut_map = {
             'tds': 'http://www.cc.com/shows/the-daily-show-with-trevor-noah/full-episodes',
             'thedailyshow': 'http://www.cc.com/shows/the-daily-show-with-trevor-noah/full-episodes',
+            'theopposition': 'http://www.cc.com/shows/the-opposition-with-jordan-klepper/full-episodes',
         }
         return self.url_result(shortcut_map[video_id])

youtube_dl/extractor/commonmistakes.py

@@ -1,5 +1,7 @@
 from __future__ import unicode_literals
 
+import sys
+
 from .common import InfoExtractor
 from ..utils import ExtractorError
 
@@ -7,7 +9,7 @@ from ..utils import ExtractorError
 class CommonMistakesIE(InfoExtractor):
     IE_DESC = False  # Do not list
     _VALID_URL = r'''(?x)
-        (?:url|URL)
+        (?:url|URL)$
     '''
 
     _TESTS = [{
@@ -33,7 +35,9 @@ class UnicodeBOMIE(InfoExtractor):
         IE_DESC = False
         _VALID_URL = r'(?P<bom>\ufeff)(?P<id>.*)$'
 
-        _TESTS = [{
+        # Disable test for python 3.2 since BOM is broken in re in this version
+        # (see https://github.com/rg3/youtube-dl/issues/9751)
+        _TESTS = [] if (3, 0) < sys.version_info <= (3, 3) else [{
             'url': '\ufeffhttp://www.youtube.com/watch?v=BaW_jenozKc',
             'only_matching': True,
         }]

youtube_dl/extractor/condenast.py

@@ -9,12 +9,12 @@ from ..compat import (
     compat_urlparse,
 )
 from ..utils import (
-    orderedSet,
-    remove_end,
-    extract_attributes,
-    mimetype2ext,
     determine_ext,
+    extract_attributes,
     int_or_none,
+    js_to_json,
+    mimetype2ext,
+    orderedSet,
     parse_iso8601,
 )
 
@@ -49,10 +49,17 @@ class CondeNastIE(InfoExtractor):
         'wmagazine': 'W Magazine',
     }
 
-    _VALID_URL = r'https?://(?:video|www|player)\.(?P<site>%s)\.com/(?P<type>watch|series|video|embed(?:js)?)/(?P<id>[^/?#]+)' % '|'.join(_SITES.keys())
+    _VALID_URL = r'''(?x)https?://(?:video|www|player(?:-backend)?)\.(?:%s)\.com/
+        (?:
+            (?:
+                embed(?:js)?|
+                (?:script|inline)/video
+            )/(?P<id>[0-9a-f]{24})(?:/(?P<player_id>[0-9a-f]{24}))?(?:.+?\btarget=(?P<target>[^&]+))?|
+            (?P<type>watch|series|video)/(?P<display_id>[^/?#]+)
+        )''' % '|'.join(_SITES.keys())
     IE_DESC = 'Condé Nast media group: %s' % ', '.join(sorted(_SITES.values()))
 
-    EMBED_URL = r'(?:https?:)?//player\.(?P<site>%s)\.com/(?P<type>embed(?:js)?)/.+?' % '|'.join(_SITES.keys())
+    EMBED_URL = r'(?:https?:)?//player(?:-backend)?\.(?:%s)\.com/(?:embed(?:js)?|(?:script|inline)/video)/.+?' % '|'.join(_SITES.keys())
 
     _TESTS = [{
         'url': 'http://video.wired.com/watch/3d-printed-speakers-lit-with-led',
@@ -66,6 +73,16 @@ class CondeNastIE(InfoExtractor):
             'upload_date': '20130314',
             'timestamp': 1363219200,
         }
+    }, {
+        'url': 'http://video.gq.com/watch/the-closer-with-keith-olbermann-the-only-true-surprise-trump-s-an-idiot?c=series',
+        'info_dict': {
+            'id': '58d1865bfd2e6126e2000015',
+            'ext': 'mp4',
+            'title': 'The Only True Surprise? Trump’s an Idiot',
+            'uploader': 'gq',
+            'upload_date': '20170321',
+            'timestamp': 1490126427,
+        },
     }, {
         # JS embed
         'url': 'http://player.cnevids.com/embedjs/55f9cf8b61646d1acf00000c/5511d76261646d5566020000.js',
@@ -78,6 +95,12 @@ class CondeNastIE(InfoExtractor):
             'upload_date': '20150916',
             'timestamp': 1442434955,
         }
+    }, {
+        'url': 'https://player.cnevids.com/inline/video/59138decb57ac36b83000005.js?target=js-cne-player',
+        'only_matching': True,
+    }, {
+        'url': 'http://player-backend.cnevids.com/script/video/59138decb57ac36b83000005.js',
+        'only_matching': True,
     }]
 
     def _extract_series(self, url, webpage):
@@ -93,16 +116,16 @@ class CondeNastIE(InfoExtractor):
         entries = [self.url_result(build_url(path), 'CondeNast') for path in paths]
         return self.playlist_result(entries, playlist_title=title)
 
-    def _extract_video(self, webpage, url_type):
-        query = {}
-        params = self._search_regex(
-            r'(?s)var params = {(.+?)}[;,]', webpage, 'player params', default=None)
-        if params:
-            query.update({
-                'videoId': self._search_regex(r'videoId: [\'"](.+?)[\'"]', params, 'video id'),
-                'playerId': self._search_regex(r'playerId: [\'"](.+?)[\'"]', params, 'player id'),
-                'target': self._search_regex(r'target: [\'"](.+?)[\'"]', params, 'target'),
-            })
+    def _extract_video_params(self, webpage, display_id):
+        query = self._parse_json(
+            self._search_regex(
+                r'(?s)var\s+params\s*=\s*({.+?})[;,]', webpage, 'player params',
+                default='{}'),
+            display_id, transform_source=js_to_json, fatal=False)
+        if query:
+            query['videoId'] = self._search_regex(
+                r'(?:data-video-id=|currentVideoId\s*=\s*)["\']([\da-f]+)',
+                webpage, 'video id', default=None)
         else:
             params = extract_attributes(self._search_regex(
                 r'(<[^>]+data-js="video-player"[^>]+>)',
@@ -112,28 +135,58 @@ class CondeNastIE(InfoExtractor):
                 'playerId': params['data-player'],
                 'target': params['id'],
             })
-        video_id = query['videoId']
+        return query
+
+    def _extract_video(self, params):
+        video_id = params['videoId']
+
         video_info = None
-        info_page = self._download_webpage(
-            'http://player.cnevids.com/player/video.js',
-            video_id, 'Downloading video info', query=query, fatal=False)
+
+        # New API path
+        query = params.copy()
+        query['embedType'] = 'inline'
+        info_page = self._download_json(
+            'http://player.cnevids.com/embed-api.json', video_id,
+            'Downloading embed info', fatal=False, query=query)
+
+        # Old fallbacks
+        if not info_page:
+            if params.get('playerId'):
+                info_page = self._download_json(
+                    'http://player.cnevids.com/player/video.js', video_id,
+                    'Downloading video info', fatal=False, query=params)
         if info_page:
-            video_info = self._parse_json(self._search_regex(
-                r'loadCallback\(({.+})\)', info_page, 'video info'), video_id)['video']
-        else:
+            video_info = info_page.get('video')
+        if not video_info:
             info_page = self._download_webpage(
                 'http://player.cnevids.com/player/loader.js',
-                video_id, 'Downloading loader info', query=query)
-            video_info = self._parse_json(self._search_regex(
-                r'var\s+video\s*=\s*({.+?});', info_page, 'video info'), video_id)
+                video_id, 'Downloading loader info', query=params)
+        if not video_info:
+            info_page = self._download_webpage(
+                'https://player.cnevids.com/inline/video/%s.js' % video_id,
+                video_id, 'Downloading inline info', query={
+                    'target': params.get('target', 'embedplayer')
+                })
+
+        if not video_info:
+            video_info = self._parse_json(
+                self._search_regex(
+                    r'(?s)var\s+config\s*=\s*({.+?});', info_page, 'config'),
+                video_id, transform_source=js_to_json)['video']
+
         title = video_info['title']
 
         formats = []
-        for fdata in video_info.get('sources', [{}])[0]:
+        for fdata in video_info['sources']:
             src = fdata.get('src')
             if not src:
                 continue
             ext = mimetype2ext(fdata.get('type')) or determine_ext(src)
+            if ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    src, video_id, 'mp4', entry_protocol='m3u8_native',
+                    m3u8_id='hls', fatal=False))
+                continue
             quality = fdata.get('quality')
             formats.append({
                 'format_id': ext + ('-%s' % quality if quality else ''),
@@ -143,9 +196,7 @@ class CondeNastIE(InfoExtractor):
             })
         self._sort_formats(formats)
 
-        info = self._search_json_ld(
-            webpage, video_id, fatal=False) if url_type != 'embed' else {}
-        info.update({
+        return {
             'id': video_id,
             'formats': formats,
             'title': title,
@@ -156,23 +207,26 @@ class CondeNastIE(InfoExtractor):
             'series': video_info.get('series_title'),
             'season': video_info.get('season_title'),
             'timestamp': parse_iso8601(video_info.get('premiere_date')),
-        })
-        return info
+            'categories': video_info.get('categories'),
+        }
 
     def _real_extract(self, url):
-        site, url_type, item_id = re.match(self._VALID_URL, url).groups()
+        video_id, player_id, target, url_type, display_id = re.match(self._VALID_URL, url).groups()
 
-        # Convert JS embed to regular embed
-        if url_type == 'embedjs':
-            parsed_url = compat_urlparse.urlparse(url)
-            url = compat_urlparse.urlunparse(parsed_url._replace(
-                path=remove_end(parsed_url.path, '.js').replace('/embedjs/', '/embed/')))
-            url_type = 'embed'
+        if video_id:
+            return self._extract_video({
+                'videoId': video_id,
+                'playerId': player_id,
+                'target': target,
+            })
 
-        self.to_screen('Extracting from %s with the Condé Nast extractor' % self._SITES[site])
-        webpage = self._download_webpage(url, item_id)
+        webpage = self._download_webpage(url, display_id)
 
         if url_type == 'series':
             return self._extract_series(url, webpage)
         else:
-            return self._extract_video(webpage, url_type)
+            params = self._extract_video_params(webpage, display_id)
+            info = self._search_json_ld(
+                webpage, display_id, fatal=False)
+            info.update(self._extract_video(params))
+            return info

youtube_dl/extractor/corus.py

@@ -0,0 +1,95 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .theplatform import ThePlatformFeedIE
+from ..utils import int_or_none
+
+
+class CorusIE(ThePlatformFeedIE):
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:www\.)?
+                        (?P<domain>
+                            (?:globaltv|etcanada)\.com|
+                            (?:hgtv|foodnetwork|slice|history|showcase)\.ca
+                        )
+                        /(?:video/|(?:[^/]+/)+(?:videos/[a-z0-9-]+-|video\.html\?.*?\bv=))
+                        (?P<id>\d+)
+                    '''
+    _TESTS = [{
+        'url': 'http://www.hgtv.ca/shows/bryan-inc/videos/movie-night-popcorn-with-bryan-870923331648/',
+        'md5': '05dcbca777bf1e58c2acbb57168ad3a6',
+        'info_dict': {
+            'id': '870923331648',
+            'ext': 'mp4',
+            'title': 'Movie Night Popcorn with Bryan',
+            'description': 'Bryan whips up homemade popcorn, the old fashion way for Jojo and Lincoln.',
+            'uploader': 'SHWM-NEW',
+            'upload_date': '20170206',
+            'timestamp': 1486392197,
+        },
+    }, {
+        'url': 'http://www.foodnetwork.ca/shows/chopped/video/episode/chocolate-obsession/video.html?v=872683587753',
+        'only_matching': True,
+    }, {
+        'url': 'http://etcanada.com/video/873675331955/meet-the-survivor-game-changers-castaways-part-2/',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.history.ca/the-world-without-canada/video/full-episodes/natural-resources/video.html?v=955054659646#video',
+        'only_matching': True,
+    }, {
+        'url': 'http://www.showcase.ca/eyewitness/video/eyewitness++106/video.html?v=955070531919&p=1&s=da#video',
+        'only_matching': True,
+    }]
+
+    _TP_FEEDS = {
+        'globaltv': {
+            'feed_id': 'ChQqrem0lNUp',
+            'account_id': 2269680845,
+        },
+        'etcanada': {
+            'feed_id': 'ChQqrem0lNUp',
+            'account_id': 2269680845,
+        },
+        'hgtv': {
+            'feed_id': 'L0BMHXi2no43',
+            'account_id': 2414428465,
+        },
+        'foodnetwork': {
+            'feed_id': 'ukK8o58zbRmJ',
+            'account_id': 2414429569,
+        },
+        'slice': {
+            'feed_id': '5tUJLgV2YNJ5',
+            'account_id': 2414427935,
+        },
+        'history': {
+            'feed_id': 'tQFx_TyyEq4J',
+            'account_id': 2369613659,
+        },
+        'showcase': {
+            'feed_id': '9H6qyshBZU3E',
+            'account_id': 2414426607,
+        },
+    }
+
+    def _real_extract(self, url):
+        domain, video_id = re.match(self._VALID_URL, url).groups()
+        feed_info = self._TP_FEEDS[domain.split('.')[0]]
+        return self._extract_feed_info('dtjsEC', feed_info['feed_id'], 'byId=' + video_id, video_id, lambda e: {
+            'episode_number': int_or_none(e.get('pl1$episode')),
+            'season_number': int_or_none(e.get('pl1$season')),
+            'series': e.get('pl1$show'),
+        }, {
+            'HLS': {
+                'manifest': 'm3u',
+            },
+            'DesktopHLS Default': {
+                'manifest': 'm3u',
+            },
+            'MP4 MBR': {
+                'manifest': 'm3u',
+            },
+        }, feed_info['account_id'])

youtube_dl/extractor/coub.py

@@ -24,12 +24,11 @@ class CoubIE(InfoExtractor):
             'duration': 4.6,
             'timestamp': 1428527772,
             'upload_date': '20150408',
-            'uploader': 'Артём Лоскутников',
+            'uploader': 'Artyom Loskutnikov',
             'uploader_id': 'artyom.loskutnikov',
             'view_count': int,
             'like_count': int,
             'repost_count': int,
-            'comment_count': int,
             'age_limit': 0,
         },
     }, {
@@ -118,7 +117,6 @@ class CoubIE(InfoExtractor):
         view_count = int_or_none(coub.get('views_count') or coub.get('views_increase_count'))
         like_count = int_or_none(coub.get('likes_count'))
         repost_count = int_or_none(coub.get('recoubs_count'))
-        comment_count = int_or_none(coub.get('comments_count'))
 
         age_restricted = coub.get('age_restricted', coub.get('age_restricted_by_admin'))
         if age_restricted is not None:
@@ -137,7 +135,6 @@ class CoubIE(InfoExtractor):
             'view_count': view_count,
             'like_count': like_count,
             'repost_count': repost_count,
-            'comment_count': comment_count,
             'age_limit': age_limit,
             'formats': formats,
         }

youtube_dl/extractor/cracked.py

@@ -3,6 +3,7 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
+from .youtube import YoutubeIE
 from ..utils import (
     parse_iso8601,
     str_to_int,
@@ -41,11 +42,9 @@ class CrackedIE(InfoExtractor):
 
         webpage = self._download_webpage(url, video_id)
 
-        youtube_url = self._search_regex(
-            r'<iframe[^>]+src="((?:https?:)?//www\.youtube\.com/embed/[^"]+)"',
-            webpage, 'youtube url', default=None)
+        youtube_url = YoutubeIE._extract_url(webpage)
         if youtube_url:
-            return self.url_result(youtube_url, 'Youtube')
+            return self.url_result(youtube_url, ie=YoutubeIE.ie_key())
 
         video_url = self._html_search_regex(
             [r'var\s+CK_vidSrc\s*=\s*"([^"]+)"', r'<video\s+src="([^"]+)"'],

youtube_dl/extractor/crackle.py

@@ -6,7 +6,8 @@ from ..utils import int_or_none
 
 
 class CrackleIE(InfoExtractor):
-    _VALID_URL = r'(?:crackle:|https?://(?:www\.)?crackle\.com/(?:playlist/\d+/|(?:[^/]+/)+))(?P<id>\d+)'
+    _GEO_COUNTRIES = ['US']
+    _VALID_URL = r'(?:crackle:|https?://(?:(?:www|m)\.)?crackle\.com/(?:playlist/\d+/|(?:[^/]+/)+))(?P<id>\d+)'
     _TEST = {
         'url': 'http://www.crackle.com/comedians-in-cars-getting-coffee/2498934',
         'info_dict': {
@@ -20,9 +21,10 @@ class CrackleIE(InfoExtractor):
             'season_number': 8,
             'episode_number': 4,
             'subtitles': {
-                'en-US': [{
-                    'ext': 'ttml',
-                }]
+                'en-US': [
+                    {'ext': 'vtt'},
+                    {'ext': 'tt'},
+                ]
             },
         },
         'params': {
@@ -31,8 +33,32 @@ class CrackleIE(InfoExtractor):
         }
     }
 
+    _THUMBNAIL_RES = [
+        (120, 90),
+        (208, 156),
+        (220, 124),
+        (220, 220),
+        (240, 180),
+        (250, 141),
+        (315, 236),
+        (320, 180),
+        (360, 203),
+        (400, 300),
+        (421, 316),
+        (460, 330),
+        (460, 460),
+        (462, 260),
+        (480, 270),
+        (587, 330),
+        (640, 480),
+        (700, 330),
+        (700, 394),
+        (854, 480),
+        (1024, 1024),
+        (1920, 1080),
+    ]
+
     # extracted from http://legacyweb-us.crackle.com/flash/ReferrerRedirect.ashx
-    _THUMBNAIL_TEMPLATE = 'http://images-us-am.crackle.com/%stnl_1920x1080.jpg?ts=20140107233116?c=635333335057637614'
     _MEDIA_FILE_SLOTS = {
         'c544.flv': {
             'width': 544,
@@ -61,17 +87,25 @@ class CrackleIE(InfoExtractor):
 
         item = self._download_xml(
             'http://legacyweb-us.crackle.com/app/revamp/vidwallcache.aspx?flags=-1&fm=%s' % video_id,
-            video_id).find('i')
+            video_id, headers=self.geo_verification_headers()).find('i')
         title = item.attrib['t']
 
         subtitles = {}
         formats = self._extract_m3u8_formats(
             'http://content.uplynk.com/ext/%s/%s.m3u8' % (config_doc.attrib['strUplynkOwnerId'], video_id),
             video_id, 'mp4', m3u8_id='hls', fatal=None)
-        thumbnail = None
+        thumbnails = []
         path = item.attrib.get('p')
         if path:
-            thumbnail = self._THUMBNAIL_TEMPLATE % path
+            for width, height in self._THUMBNAIL_RES:
+                res = '%dx%d' % (width, height)
+                thumbnails.append({
+                    'id': res,
+                    'url': 'http://images-us-am.crackle.com/%stnl_%s.jpg' % (path, res),
+                    'width': width,
+                    'height': height,
+                    'resolution': res,
+                })
             http_base_url = 'http://ahttp.crackle.com/' + path
             for mfs_path, mfs_info in self._MEDIA_FILE_SLOTS.items():
                 formats.append({
@@ -86,10 +120,11 @@ class CrackleIE(InfoExtractor):
                 if locale and v:
                     if locale not in subtitles:
                         subtitles[locale] = []
-                    subtitles[locale] = [{
-                        'url': '%s/%s%s_%s.xml' % (config_doc.attrib['strSubtitleServer'], path, locale, v),
-                        'ext': 'ttml',
-                    }]
+                    for url_ext, ext in (('vtt', 'vtt'), ('xml', 'tt')):
+                        subtitles.setdefault(locale, []).append({
+                            'url': '%s/%s%s_%s.%s' % (config_doc.attrib['strSubtitleServer'], path, locale, v, url_ext),
+                            'ext': ext,
+                        })
         self._sort_formats(formats, ('width', 'height', 'tbr', 'format_id'))
 
         return {
@@ -100,7 +135,7 @@ class CrackleIE(InfoExtractor):
             'series': item.attrib.get('sn'),
             'season_number': int_or_none(item.attrib.get('se')),
             'episode_number': int_or_none(item.attrib.get('ep')),
-            'thumbnail': thumbnail,
+            'thumbnails': thumbnails,
             'subtitles': subtitles,
             'formats': formats,
         }

youtube_dl/extractor/crunchyroll.py

@@ -38,11 +38,32 @@ class CrunchyrollBaseIE(InfoExtractor):
     _LOGIN_FORM = 'login_form'
     _NETRC_MACHINE = 'crunchyroll'
 
+    def _call_rpc_api(self, method, video_id, note=None, data=None):
+        data = data or {}
+        data['req'] = 'RpcApi' + method
+        data = compat_urllib_parse_urlencode(data).encode('utf-8')
+        return self._download_xml(
+            'http://www.crunchyroll.com/xml/',
+            video_id, note, fatal=False, data=data, headers={
+                'Content-Type': 'application/x-www-form-urlencoded',
+            })
+
     def _login(self):
         (username, password) = self._get_login_info()
         if username is None:
             return
 
+        self._download_webpage(
+            'https://www.crunchyroll.com/?a=formhandler',
+            None, 'Logging in', 'Wrong login info',
+            data=urlencode_postdata({
+                'formname': 'RpcApiUser_Login',
+                'next_url': 'https://www.crunchyroll.com/acct/membership',
+                'name': username,
+                'password': password,
+            }))
+
+        '''
         login_page = self._download_webpage(
             self._LOGIN_URL, None, 'Downloading login page')
 
@@ -86,6 +107,7 @@ class CrunchyrollBaseIE(InfoExtractor):
             raise ExtractorError('Unable to login: %s' % error, expected=True)
 
         raise ExtractorError('Unable to log in')
+        '''
 
     def _real_initialize(self):
         self._login()
@@ -123,7 +145,7 @@ class CrunchyrollIE(CrunchyrollBaseIE):
         'url': 'http://www.crunchyroll.com/wanna-be-the-strongest-in-the-world/episode-1-an-idol-wrestler-is-born-645513',
         'info_dict': {
             'id': '645513',
-            'ext': 'flv',
+            'ext': 'mp4',
             'title': 'Wanna be the Strongest in the World Episode 1 – An Idol-Wrestler is Born!',
             'description': 'md5:2d17137920c64f2f49981a7797d275ef',
             'thumbnail': 'http://img1.ak.crunchyroll.com/i/spire1-tmb/20c6b5e10f1a47b10516877d3c039cae1380951166_full.jpg',
@@ -166,6 +188,26 @@ class CrunchyrollIE(CrunchyrollBaseIE):
             # m3u8 download
             'skip_download': True,
         },
+    }, {
+        'url': 'http://www.crunchyroll.com/konosuba-gods-blessing-on-this-wonderful-world/episode-1-give-me-deliverance-from-this-judicial-injustice-727589',
+        'info_dict': {
+            'id': '727589',
+            'ext': 'mp4',
+            'title': "KONOSUBA -God's blessing on this wonderful world! 2 Episode 1 – Give Me Deliverance From This Judicial Injustice!",
+            'description': 'md5:cbcf05e528124b0f3a0a419fc805ea7d',
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'uploader': 'Kadokawa Pictures Inc.',
+            'upload_date': '20170118',
+            'series': "KONOSUBA -God's blessing on this wonderful world!",
+            'season': "KONOSUBA -God's blessing on this wonderful world! 2",
+            'season_number': 2,
+            'episode': 'Give Me Deliverance From This Judicial Injustice!',
+            'episode_number': 1,
+        },
+        'params': {
+            # m3u8 download
+            'skip_download': True,
+        },
     }, {
         'url': 'http://www.crunchyroll.fr/girl-friend-beta/episode-11-goodbye-la-mode-661697',
         'only_matching': True,
@@ -173,6 +215,53 @@ class CrunchyrollIE(CrunchyrollBaseIE):
         # geo-restricted (US), 18+ maturity wall, non-premium available
         'url': 'http://www.crunchyroll.com/cosplay-complex-ova/episode-1-the-birth-of-the-cosplay-club-565617',
         'only_matching': True,
+    }, {
+        # A description with double quotes
+        'url': 'http://www.crunchyroll.com/11eyes/episode-1-piros-jszaka-red-night-535080',
+        'info_dict': {
+            'id': '535080',
+            'ext': 'mp4',
+            'title': '11eyes Episode 1 – Piros éjszaka - Red Night',
+            'description': 'Kakeru and Yuka are thrown into an alternate nightmarish world they call "Red Night".',
+            'uploader': 'Marvelous AQL Inc.',
+            'upload_date': '20091021',
+        },
+        'params': {
+            # Just test metadata extraction
+            'skip_download': True,
+        },
+    }, {
+        # make sure we can extract an uploader name that's not a link
+        'url': 'http://www.crunchyroll.com/hakuoki-reimeiroku/episode-1-dawn-of-the-divine-warriors-606899',
+        'info_dict': {
+            'id': '606899',
+            'ext': 'mp4',
+            'title': 'Hakuoki Reimeiroku Episode 1 – Dawn of the Divine Warriors',
+            'description': 'Ryunosuke was left to die, but Serizawa-san asked him a simple question "Do you want to live?"',
+            'uploader': 'Geneon Entertainment',
+            'upload_date': '20120717',
+        },
+        'params': {
+            # just test metadata extraction
+            'skip_download': True,
+        },
+    }, {
+        # A video with a vastly different season name compared to the series name
+        'url': 'http://www.crunchyroll.com/nyarko-san-another-crawling-chaos/episode-1-test-590532',
+        'info_dict': {
+            'id': '590532',
+            'ext': 'mp4',
+            'title': 'Haiyoru! Nyaruani (ONA) Episode 1 – Test',
+            'description': 'Mahiro and Nyaruko talk about official certification.',
+            'uploader': 'TV TOKYO',
+            'upload_date': '20120305',
+            'series': 'Nyarko-san: Another Crawling Chaos',
+            'season': 'Haiyoru! Nyaruani (ONA)',
+        },
+        'params': {
+            # Just test metadata extraction
+            'skip_download': True,
+        },
     }]
 
     _FORMAT_IDS = {
@@ -236,8 +325,7 @@ class CrunchyrollIE(CrunchyrollBaseIE):
         output += 'WrapStyle: %s\n' % sub_root.attrib['wrap_style']
         output += 'PlayResX: %s\n' % sub_root.attrib['play_res_x']
         output += 'PlayResY: %s\n' % sub_root.attrib['play_res_y']
-        output += """ScaledBorderAndShadow: no
-
+        output += """
 [V4+ Styles]
 Format: Name, Fontname, Fontsize, PrimaryColour, SecondaryColour, OutlineColour, BackColour, Bold, Italic, Underline, StrikeOut, ScaleX, ScaleY, Spacing, Angle, BorderStyle, Outline, Shadow, Alignment, MarginL, MarginR, MarginV, Encoding
 """
@@ -299,15 +387,19 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
     def _get_subtitles(self, video_id, webpage):
         subtitles = {}
         for sub_id, sub_name in re.findall(r'\bssid=([0-9]+)"[^>]+?\btitle="([^"]+)', webpage):
-            sub_page = self._download_webpage(
-                'http://www.crunchyroll.com/xml/?req=RpcApiSubtitle_GetXml&subtitle_script_id=' + sub_id,
-                video_id, note='Downloading subtitles for ' + sub_name)
-            id = self._search_regex(r'id=\'([0-9]+)', sub_page, 'subtitle_id', fatal=False)
-            iv = self._search_regex(r'<iv>([^<]+)', sub_page, 'subtitle_iv', fatal=False)
-            data = self._search_regex(r'<data>([^<]+)', sub_page, 'subtitle_data', fatal=False)
-            if not id or not iv or not data:
+            sub_doc = self._call_rpc_api(
+                'Subtitle_GetXml', video_id,
+                'Downloading subtitles for ' + sub_name, data={
+                    'subtitle_script_id': sub_id,
+                })
+            if sub_doc is None:
                 continue
-            subtitle = self._decrypt_subtitles(data, iv, id).decode('utf-8')
+            sid = sub_doc.get('id')
+            iv = xpath_text(sub_doc, 'iv', 'subtitle iv')
+            data = xpath_text(sub_doc, 'data', 'subtitle data')
+            if not sid or not iv or not data:
+                continue
+            subtitle = self._decrypt_subtitles(data, iv, sid).decode('utf-8')
             lang_code = self._search_regex(r'lang_code=["\']([^"\']+)', subtitle, 'subtitle_lang_code', fatal=False)
             if not lang_code:
                 continue
@@ -324,7 +416,9 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
         else:
             webpage_url = 'http://www.' + mobj.group('url')
 
-        webpage = self._download_webpage(self._add_skip_wall(webpage_url), video_id, 'Downloading webpage')
+        webpage = self._download_webpage(
+            self._add_skip_wall(webpage_url), video_id,
+            headers=self.geo_verification_headers())
         note_m = self._html_search_regex(
             r'<div class="showmedia-trailer-notice">(.+?)</div>',
             webpage, 'trailer-notice', default='')
@@ -344,9 +438,9 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
             r'(?s)<h1[^>]*>((?:(?!<h1).)*?<span[^>]+itemprop=["\']title["\'][^>]*>(?:(?!<h1).)+?)</h1>',
             webpage, 'video_title')
         video_title = re.sub(r' {2,}', ' ', video_title)
-        video_description = self._html_search_regex(
-            r'<script[^>]*>\s*.+?\[media_id=%s\].+?"description"\s*:\s*"([^"]+)' % video_id,
-            webpage, 'description', default=None)
+        video_description = self._parse_json(self._html_search_regex(
+            r'<script[^>]*>\s*.+?\[media_id=%s\].+?({.+?"description"\s*:.+?})\);' % video_id,
+            webpage, 'description', default='{}'), video_id).get('description')
         if video_description:
             video_description = lowercase_escape(video_description.replace(r'\r\n', '\n'))
         video_upload_date = self._html_search_regex(
@@ -355,8 +449,9 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
         if video_upload_date:
             video_upload_date = unified_strdate(video_upload_date)
         video_uploader = self._html_search_regex(
-            r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', webpage,
-            'video_uploader', fatal=False)
+            # try looking for both an uploader that's a link and one that's not
+            [r'<a[^>]+href="/publisher/[^"]+"[^>]*>([^<]+)</a>', r'<div>\s*Publisher:\s*<span>\s*(.+?)\s*</span>\s*</div>'],
+            webpage, 'video_uploader', fatal=False)
 
         available_fmts = []
         for a, fmt in re.findall(r'(<a[^>]+token=["\']showmedia\.([0-9]{3,4})p["\'][^>]+>)', webpage):
@@ -375,70 +470,97 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
         for fmt in available_fmts:
             stream_quality, stream_format = self._FORMAT_IDS[fmt]
             video_format = fmt + 'p'
-            streamdata_req = sanitized_Request(
-                'http://www.crunchyroll.com/xml/?req=RpcApiVideoPlayer_GetStandardConfig&media_id=%s&video_format=%s&video_quality=%s'
-                % (video_id, stream_format, stream_quality),
-                compat_urllib_parse_urlencode({'current_page': url}).encode('utf-8'))
-            streamdata_req.add_header('Content-Type', 'application/x-www-form-urlencoded')
-            streamdata = self._download_xml(
-                streamdata_req, video_id,
-                note='Downloading media info for %s' % video_format)
-            stream_info = streamdata.find('./{default}preload/stream_info')
-            video_encode_id = xpath_text(stream_info, './video_encode_id')
-            if video_encode_id in video_encode_ids:
-                continue
-            video_encode_ids.append(video_encode_id)
+            stream_infos = []
+            streamdata = self._call_rpc_api(
+                'VideoPlayer_GetStandardConfig', video_id,
+                'Downloading media info for %s' % video_format, data={
+                    'media_id': video_id,
+                    'video_format': stream_format,
+                    'video_quality': stream_quality,
+                    'current_page': url,
+                })
+            if streamdata is not None:
+                stream_info = streamdata.find('./{default}preload/stream_info')
+                if stream_info is not None:
+                    stream_infos.append(stream_info)
+            stream_info = self._call_rpc_api(
+                'VideoEncode_GetStreamInfo', video_id,
+                'Downloading stream info for %s' % video_format, data={
+                    'media_id': video_id,
+                    'video_format': stream_format,
+                    'video_encode_quality': stream_quality,
+                })
+            if stream_info is not None:
+                stream_infos.append(stream_info)
+            for stream_info in stream_infos:
+                video_encode_id = xpath_text(stream_info, './video_encode_id')
+                if video_encode_id in video_encode_ids:
+                    continue
+                video_encode_ids.append(video_encode_id)
 
-            video_file = xpath_text(stream_info, './file')
-            if not video_file:
-                continue
-            if video_file.startswith('http'):
-                formats.extend(self._extract_m3u8_formats(
-                    video_file, video_id, 'mp4', entry_protocol='m3u8_native',
-                    m3u8_id='hls', fatal=False))
-                continue
-
-            video_url = xpath_text(stream_info, './host')
-            if not video_url:
-                continue
-            metadata = stream_info.find('./metadata')
-            format_info = {
-                'format': video_format,
-                'format_id': video_format,
-                'height': int_or_none(xpath_text(metadata, './height')),
-                'width': int_or_none(xpath_text(metadata, './width')),
-            }
-
-            if '.fplive.net/' in video_url:
-                video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip())
-                parsed_video_url = compat_urlparse.urlparse(video_url)
-                direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace(
-                    netloc='v.lvlt.crcdn.net',
-                    path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_file.split(':')[-1])))
-                if self._is_valid_url(direct_video_url, video_id, video_format):
-                    format_info.update({
-                        'url': direct_video_url,
-                    })
-                    formats.append(format_info)
+                video_file = xpath_text(stream_info, './file')
+                if not video_file:
+                    continue
+                if video_file.startswith('http'):
+                    formats.extend(self._extract_m3u8_formats(
+                        video_file, video_id, 'mp4', entry_protocol='m3u8_native',
+                        m3u8_id='hls', fatal=False))
                     continue
 
-            format_info.update({
-                'url': video_url,
-                'play_path': video_file,
-                'ext': 'flv',
-            })
-            formats.append(format_info)
-        self._sort_formats(formats)
+                video_url = xpath_text(stream_info, './host')
+                if not video_url:
+                    continue
+                metadata = stream_info.find('./metadata')
+                format_info = {
+                    'format': video_format,
+                    'height': int_or_none(xpath_text(metadata, './height')),
+                    'width': int_or_none(xpath_text(metadata, './width')),
+                }
 
-        metadata = self._download_xml(
-            'http://www.crunchyroll.com/xml', video_id,
-            note='Downloading media info', query={
-                'req': 'RpcApiVideoPlayer_GetMediaMetadata',
+                if '.fplive.net/' in video_url:
+                    video_url = re.sub(r'^rtmpe?://', 'http://', video_url.strip())
+                    parsed_video_url = compat_urlparse.urlparse(video_url)
+                    direct_video_url = compat_urlparse.urlunparse(parsed_video_url._replace(
+                        netloc='v.lvlt.crcdn.net',
+                        path='%s/%s' % (remove_end(parsed_video_url.path, '/'), video_file.split(':')[-1])))
+                    if self._is_valid_url(direct_video_url, video_id, video_format):
+                        format_info.update({
+                            'format_id': 'http-' + video_format,
+                            'url': direct_video_url,
+                        })
+                        formats.append(format_info)
+                        continue
+
+                format_info.update({
+                    'format_id': 'rtmp-' + video_format,
+                    'url': video_url,
+                    'play_path': video_file,
+                    'ext': 'flv',
+                })
+                formats.append(format_info)
+        self._sort_formats(formats, ('height', 'width', 'tbr', 'fps'))
+
+        metadata = self._call_rpc_api(
+            'VideoPlayer_GetMediaMetadata', video_id,
+            note='Downloading media info', data={
                 'media_id': video_id,
             })
 
         subtitles = self.extract_subtitles(video_id, webpage)
 
+        # webpage provide more accurate data than series_title from XML
+        series = self._html_search_regex(
+            r'(?s)<h\d[^>]+\bid=["\']showmedia_about_episode_num[^>]+>(.+?)</h\d',
+            webpage, 'series', fatal=False)
+        season = xpath_text(metadata, 'series_title')
+
+        episode = xpath_text(metadata, 'episode_title')
+        episode_number = int_or_none(xpath_text(metadata, 'episode_number'))
+
+        season_number = int_or_none(self._search_regex(
+            r'(?s)<h\d[^>]+id=["\']showmedia_about_episode_num[^>]+>.+?</h\d>\s*<h4>\s*Season (\d+)',
+            webpage, 'season number', default=None))
+
         return {
             'id': video_id,
             'title': video_title,
@@ -446,9 +568,11 @@ Format: Layer, Start, End, Style, Name, MarginL, MarginR, MarginV, Effect, Text
             'thumbnail': xpath_text(metadata, 'episode_image_url'),
             'uploader': video_uploader,
             'upload_date': video_upload_date,
-            'series': xpath_text(metadata, 'series_title'),
-            'episode': xpath_text(metadata, 'episode_title'),
-            'episode_number': int_or_none(xpath_text(metadata, 'episode_number')),
+            'series': series,
+            'season': season,
+            'season_number': season_number,
+            'episode': episode,
+            'episode_number': episode_number,
             'subtitles': subtitles,
             'formats': formats,
         }
@@ -483,16 +607,18 @@ class CrunchyrollShowPlaylistIE(CrunchyrollBaseIE):
     def _real_extract(self, url):
         show_id = self._match_id(url)
 
-        webpage = self._download_webpage(self._add_skip_wall(url), show_id)
+        webpage = self._download_webpage(
+            self._add_skip_wall(url), show_id,
+            headers=self.geo_verification_headers())
         title = self._html_search_regex(
             r'(?s)<h1[^>]*>\s*<span itemprop="name">(.*?)</span>',
             webpage, 'title')
         episode_paths = re.findall(
-            r'(?s)<li id="showview_videos_media_[0-9]+"[^>]+>.*?<a href="([^"]+)"',
+            r'(?s)<li id="showview_videos_media_(\d+)"[^>]+>.*?<a href="([^"]+)"',
             webpage)
         entries = [
-            self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll')
-            for ep in episode_paths
+            self.url_result('http://www.crunchyroll.com' + ep, 'Crunchyroll', ep_id)
+            for ep_id, ep in episode_paths
         ]
         entries.reverse()
 

youtube_dl/extractor/cspan.py

@@ -4,14 +4,17 @@ import re
 
 from .common import InfoExtractor
 from ..utils import (
-    int_or_none,
-    unescapeHTML,
-    find_xpath_attr,
-    smuggle_url,
     determine_ext,
     ExtractorError,
+    extract_attributes,
+    find_xpath_attr,
+    get_element_by_class,
+    int_or_none,
+    smuggle_url,
+    unescapeHTML,
 )
 from .senateisvp import SenateISVPIE
+from .ustream import UstreamIE
 
 
 class CSpanIE(InfoExtractor):
@@ -22,14 +25,13 @@ class CSpanIE(InfoExtractor):
         'md5': '94b29a4f131ff03d23471dd6f60b6a1d',
         'info_dict': {
             'id': '315139',
-            'ext': 'mp4',
             'title': 'Attorney General Eric Holder on Voting Rights Act Decision',
-            'description': 'Attorney General Eric Holder speaks to reporters following the Supreme Court decision in [Shelby County v. Holder], in which the court ruled that the preclearance provisions of the Voting Rights Act could not be enforced.',
         },
+        'playlist_mincount': 2,
         'skip': 'Regularly fails on travis, for unknown reasons',
     }, {
         'url': 'http://www.c-span.org/video/?c4486943/cspan-international-health-care-models',
-        'md5': '8e5fbfabe6ad0f89f3012a7943c1287b',
+        # md5 is unstable
         'info_dict': {
             'id': 'c4486943',
             'ext': 'mp4',
@@ -38,14 +40,11 @@ class CSpanIE(InfoExtractor):
         }
     }, {
         'url': 'http://www.c-span.org/video/?318608-1/gm-ignition-switch-recall',
-        'md5': '2ae5051559169baadba13fc35345ae74',
         'info_dict': {
             'id': '342759',
-            'ext': 'mp4',
             'title': 'General Motors Ignition Switch Recall',
-            'duration': 14848,
-            'description': 'md5:118081aedd24bf1d3b68b3803344e7f3'
         },
+        'playlist_mincount': 6,
     }, {
         # Video from senate.gov
         'url': 'http://www.c-span.org/video/?104517-1/immigration-reforms-needed-protect-skilled-american-workers',
@@ -57,12 +56,48 @@ class CSpanIE(InfoExtractor):
         'params': {
             'skip_download': True,  # m3u8 downloads
         }
+    }, {
+        # Ustream embedded video
+        'url': 'https://www.c-span.org/video/?114917-1/armed-services',
+        'info_dict': {
+            'id': '58428542',
+            'ext': 'flv',
+            'title': 'USHR07 Armed Services Committee',
+            'description': 'hsas00-2118-20150204-1000et-07\n\n\nUSHR07 Armed Services Committee',
+            'timestamp': 1423060374,
+            'upload_date': '20150204',
+            'uploader': 'HouseCommittee',
+            'uploader_id': '12987475',
+        },
+    }, {
+        # Audio Only
+        'url': 'https://www.c-span.org/video/?437336-1/judiciary-antitrust-competition-policy-consumer-rights',
+        'only_matching': True,
     }]
+    BRIGHTCOVE_URL_TEMPLATE = 'http://players.brightcove.net/%s/%s_%s/index.html?videoId=%s'
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
         video_type = None
         webpage = self._download_webpage(url, video_id)
+
+        ustream_url = UstreamIE._extract_url(webpage)
+        if ustream_url:
+            return self.url_result(ustream_url, UstreamIE.ie_key())
+
+        if '&vod' not in url:
+            bc = self._search_regex(
+                r"(<[^>]+id='brightcove-player-embed'[^>]+>)",
+                webpage, 'brightcove embed', default=None)
+            if bc:
+                bc_attr = extract_attributes(bc)
+                bc_url = self.BRIGHTCOVE_URL_TEMPLATE % (
+                    bc_attr.get('data-bcaccountid', '3162030207001'),
+                    bc_attr.get('data-noprebcplayerid', 'SyGGpuJy3g'),
+                    bc_attr.get('data-newbcplayerid', 'default'),
+                    bc_attr['data-bcid'])
+                return self.url_result(smuggle_url(bc_url, {'source_url': url}))
+
         # We first look for clipid, because clipprog always appears before
         patterns = [r'id=\'clip(%s)\'\s*value=\'([0-9]+)\'' % t for t in ('id', 'prog')]
         results = list(filter(None, (re.search(p, webpage) for p in patterns)))
@@ -81,7 +116,15 @@ class CSpanIE(InfoExtractor):
                     title = self._og_search_title(webpage)
                     surl = smuggle_url(senate_isvp_url, {'force_title': title})
                     return self.url_result(surl, 'SenateISVP', video_id, title)
+                video_id = self._search_regex(
+                    r'jwsetup\.clipprog\s*=\s*(\d+);',
+                    webpage, 'jwsetup program id', default=None)
+                if video_id:
+                    video_type = 'program'
         if video_type is None or video_id is None:
+            error_message = get_element_by_class('VLplayer-error-message', webpage)
+            if error_message:
+                raise ExtractorError(error_message)
             raise ExtractorError('unable to find video id and type')
 
         def get_text_attr(d, attr):
@@ -108,7 +151,7 @@ class CSpanIE(InfoExtractor):
         entries = []
         for partnum, f in enumerate(files):
             formats = []
-            for quality in f['qualities']:
+            for quality in f.get('qualities', []):
                 formats.append({
                     'format_id': '%s-%sp' % (get_text_attr(quality, 'bitrate'), get_text_attr(quality, 'height')),
                     'url': unescapeHTML(get_text_attr(quality, 'file')),

youtube_dl/extractor/curiositystream.py

@@ -1,6 +1,8 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
 from ..utils import (
     int_or_none,
@@ -46,9 +48,50 @@ class CuriosityStreamBaseIE(InfoExtractor):
 
     def _extract_media_info(self, media):
         video_id = compat_str(media['id'])
-        limelight_media_id = media['limelight_media_id']
         title = media['title']
 
+        formats = []
+        for encoding in media.get('encodings', []):
+            m3u8_url = encoding.get('master_playlist_url')
+            if m3u8_url:
+                formats.extend(self._extract_m3u8_formats(
+                    m3u8_url, video_id, 'mp4', 'm3u8_native',
+                    m3u8_id='hls', fatal=False))
+            encoding_url = encoding.get('url')
+            file_url = encoding.get('file_url')
+            if not encoding_url and not file_url:
+                continue
+            f = {
+                'width': int_or_none(encoding.get('width')),
+                'height': int_or_none(encoding.get('height')),
+                'vbr': int_or_none(encoding.get('video_bitrate')),
+                'abr': int_or_none(encoding.get('audio_bitrate')),
+                'filesize': int_or_none(encoding.get('size_in_bytes')),
+                'vcodec': encoding.get('video_codec'),
+                'acodec': encoding.get('audio_codec'),
+                'container': encoding.get('container_type'),
+            }
+            for f_url in (encoding_url, file_url):
+                if not f_url:
+                    continue
+                fmt = f.copy()
+                rtmp = re.search(r'^(?P<url>rtmpe?://(?P<host>[^/]+)/(?P<app>.+))/(?P<playpath>mp[34]:.+)$', f_url)
+                if rtmp:
+                    fmt.update({
+                        'url': rtmp.group('url'),
+                        'play_path': rtmp.group('playpath'),
+                        'app': rtmp.group('app'),
+                        'ext': 'flv',
+                        'format_id': 'rtmp',
+                    })
+                else:
+                    fmt.update({
+                        'url': f_url,
+                        'format_id': 'http',
+                    })
+                formats.append(fmt)
+        self._sort_formats(formats)
+
         subtitles = {}
         for closed_caption in media.get('closed_captions', []):
             sub_url = closed_caption.get('file')
@@ -60,16 +103,14 @@ class CuriosityStreamBaseIE(InfoExtractor):
             })
 
         return {
-            '_type': 'url_transparent',
             'id': video_id,
-            'url': 'limelight:media:' + limelight_media_id,
+            'formats': formats,
             'title': title,
             'description': media.get('description'),
             'thumbnail': media.get('image_large') or media.get('image_medium') or media.get('image_small'),
             'duration': int_or_none(media.get('duration')),
             'tags': media.get('tags'),
             'subtitles': subtitles,
-            'ie_key': 'LimelightMedia',
         }
 
 
@@ -78,14 +119,12 @@ class CuriosityStreamIE(CuriosityStreamBaseIE):
     _VALID_URL = r'https?://app\.curiositystream\.com/video/(?P<id>\d+)'
     _TEST = {
         'url': 'https://app.curiositystream.com/video/2',
-        'md5': 'a0074c190e6cddaf86900b28d3e9ee7a',
+        'md5': '262bb2f257ff301115f1973540de8983',
         'info_dict': {
             'id': '2',
             'ext': 'mp4',
             'title': 'How Did You Develop The Internet?',
             'description': 'Vint Cerf, Google\'s Chief Internet Evangelist, describes how he and Bob Kahn created the internet.',
-            'timestamp': 1448388615,
-            'upload_date': '20151124',
         }
     }
 
@@ -105,7 +144,7 @@ class CuriosityStreamCollectionIE(CuriosityStreamBaseIE):
             'title': 'Curious Minds: The Internet',
             'description': 'How is the internet shaping our lives in the 21st Century?',
         },
-        'playlist_mincount': 17,
+        'playlist_mincount': 12,
     }
 
     def _real_extract(self, url):

youtube_dl/extractor/cwtv.py

@@ -82,6 +82,11 @@ class CWTVIE(InfoExtractor):
                             'url': quality_url,
                             'tbr': tbr,
                         })
+        video_metadata = video_data['assetFields']
+        ism_url = video_metadata.get('smoothStreamingUrl')
+        if ism_url:
+            formats.extend(self._extract_ism_formats(
+                ism_url, video_id, ism_id='mss', fatal=False))
         self._sort_formats(formats)
 
         thumbnails = [{
@@ -90,8 +95,6 @@ class CWTVIE(InfoExtractor):
             'height': image.get('height'),
         } for image_id, image in video_data['images'].items() if image.get('uri')] if video_data.get('images') else None
 
-        video_metadata = video_data['assetFields']
-
         subtitles = {
             'en': [{
                 'url': video_metadata['UnicornCcUrl'],

youtube_dl/extractor/dailymail.py

@@ -1,17 +1,21 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import re
+
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
     int_or_none,
     determine_protocol,
+    try_get,
     unescapeHTML,
 )
 
 
 class DailyMailIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?dailymail\.co\.uk/video/[^/]+/video-(?P<id>[0-9]+)'
-    _TEST = {
+    _VALID_URL = r'https?://(?:www\.)?dailymail\.co\.uk/(?:video/[^/]+/video-|embed/video/)(?P<id>[0-9]+)'
+    _TESTS = [{
         'url': 'http://www.dailymail.co.uk/video/tvshowbiz/video-1295863/The-Mountain-appears-sparkling-water-ad-Heavy-Bubbles.html',
         'md5': 'f6129624562251f628296c3a9ffde124',
         'info_dict': {
@@ -20,7 +24,16 @@ class DailyMailIE(InfoExtractor):
             'title': 'The Mountain appears in sparkling water ad for \'Heavy Bubbles\'',
             'description': 'md5:a93d74b6da172dd5dc4d973e0b766a84',
         }
-    }
+    }, {
+        'url': 'http://www.dailymail.co.uk/embed/video/1295863.html',
+        'only_matching': True,
+    }]
+
+    @staticmethod
+    def _extract_urls(webpage):
+        return re.findall(
+            r'<iframe\b[^>]+\bsrc=["\'](?P<url>(?:https?:)?//(?:www\.)?dailymail\.co\.uk/embed/video/\d+\.html)',
+            webpage)
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
@@ -28,8 +41,14 @@ class DailyMailIE(InfoExtractor):
         video_data = self._parse_json(self._search_regex(
             r"data-opts='({.+?})'", webpage, 'video data'), video_id)
         title = unescapeHTML(video_data['title'])
-        video_sources = self._download_json(video_data.get(
-            'sources', {}).get('url') or 'http://www.dailymail.co.uk/api/player/%s/video-sources.json' % video_id, video_id)
+
+        sources_url = (try_get(
+            video_data,
+            (lambda x: x['plugins']['sources']['url'],
+             lambda x: x['sources']['url']), compat_str) or
+            'http://www.dailymail.co.uk/api/player/%s/video-sources.json' % video_id)
+
+        video_sources = self._download_json(sources_url, video_id)
 
         formats = []
         for rendition in video_sources['renditions']: