diff options
Diffstat (limited to 'youtube_dl')
24 files changed, 756 insertions, 291 deletions
diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py index 702a6ad50..cad6b026e 100755 --- a/youtube_dl/YoutubeDL.py +++ b/youtube_dl/YoutubeDL.py @@ -21,24 +21,24 @@ import subprocess import socket import sys import time +import tokenize import traceback if os.name == 'nt': import ctypes from .compat import ( - compat_basestring, compat_cookiejar, compat_expanduser, compat_get_terminal_size, compat_http_client, compat_kwargs, compat_str, + compat_tokenize_tokenize, compat_urllib_error, compat_urllib_request, ) from .utils import ( - escape_url, ContentTooShortError, date_from_str, DateRange, @@ -49,7 +49,6 @@ from .utils import ( ExtractorError, format_bytes, formatSeconds, - HEADRequest, locked_file, make_HTTPS_handler, MaxDownloadsReached, @@ -853,8 +852,8 @@ class YoutubeDL(object): else: raise Exception('Invalid result type: %s' % result_type) - def _apply_format_filter(self, format_spec, available_formats): - " Returns a tuple of the remaining format_spec and filtered formats " + def _build_format_filter(self, filter_spec): + " Returns a function to filter the formats according to the filter_spec " OPERATORS = { '<': operator.lt, @@ -864,13 +863,13 @@ class YoutubeDL(object): '=': operator.eq, '!=': operator.ne, } - operator_rex = re.compile(r'''(?x)\s*\[ + operator_rex = re.compile(r'''(?x)\s* (?P<key>width|height|tbr|abr|vbr|asr|filesize|fps) \s*(?P<op>%s)(?P<none_inclusive>\s*\?)?\s* (?P<value>[0-9.]+(?:[kKmMgGtTpPeEzZyY]i?[Bb]?)?) - \]$ + $ ''' % '|'.join(map(re.escape, OPERATORS.keys()))) - m = operator_rex.search(format_spec) + m = operator_rex.search(filter_spec) if m: try: comparison_value = int(m.group('value')) @@ -881,7 +880,7 @@ class YoutubeDL(object): if comparison_value is None: raise ValueError( 'Invalid value %r in format specification %r' % ( - m.group('value'), format_spec)) + m.group('value'), filter_spec)) op = OPERATORS[m.group('op')] if not m: @@ -889,85 +888,283 @@ class YoutubeDL(object): '=': operator.eq, '!=': operator.ne, } - str_operator_rex = re.compile(r'''(?x)\s*\[ + str_operator_rex = re.compile(r'''(?x) \s*(?P<key>ext|acodec|vcodec|container|protocol) \s*(?P<op>%s)(?P<none_inclusive>\s*\?)? \s*(?P<value>[a-zA-Z0-9_-]+) - \s*\]$ + \s*$ ''' % '|'.join(map(re.escape, STR_OPERATORS.keys()))) - m = str_operator_rex.search(format_spec) + m = str_operator_rex.search(filter_spec) if m: comparison_value = m.group('value') op = STR_OPERATORS[m.group('op')] if not m: - raise ValueError('Invalid format specification %r' % format_spec) + raise ValueError('Invalid filter specification %r' % filter_spec) def _filter(f): actual_value = f.get(m.group('key')) if actual_value is None: return m.group('none_inclusive') return op(actual_value, comparison_value) - new_formats = [f for f in available_formats if _filter(f)] + return _filter + + def build_format_selector(self, format_spec): + def syntax_error(note, start): + message = ( + 'Invalid format specification: ' + '{0}\n\t{1}\n\t{2}^'.format(note, format_spec, ' ' * start[1])) + return SyntaxError(message) + + PICKFIRST = 'PICKFIRST' + MERGE = 'MERGE' + SINGLE = 'SINGLE' + GROUP = 'GROUP' + FormatSelector = collections.namedtuple('FormatSelector', ['type', 'selector', 'filters']) + + def _parse_filter(tokens): + filter_parts = [] + for type, string, start, _, _ in tokens: + if type == tokenize.OP and string == ']': + return ''.join(filter_parts) + else: + filter_parts.append(string) + + def _remove_unused_ops(tokens): + # Remove operators that we don't use and join them with the sourrounding strings + # for example: 'mp4' '-' 'baseline' '-' '16x9' is converted to 'mp4-baseline-16x9' + ALLOWED_OPS = ('/', '+', ',', '(', ')') + last_string, last_start, last_end, last_line = None, None, None, None + for type, string, start, end, line in tokens: + if type == tokenize.OP and string == '[': + if last_string: + yield tokenize.NAME, last_string, last_start, last_end, last_line + last_string = None + yield type, string, start, end, line + # everything inside brackets will be handled by _parse_filter + for type, string, start, end, line in tokens: + yield type, string, start, end, line + if type == tokenize.OP and string == ']': + break + elif type == tokenize.OP and string in ALLOWED_OPS: + if last_string: + yield tokenize.NAME, last_string, last_start, last_end, last_line + last_string = None + yield type, string, start, end, line + elif type in [tokenize.NAME, tokenize.NUMBER, tokenize.OP]: + if not last_string: + last_string = string + last_start = start + last_end = end + else: + last_string += string + if last_string: + yield tokenize.NAME, last_string, last_start, last_end, last_line + + def _parse_format_selection(tokens, inside_merge=False, inside_choice=False, inside_group=False): + selectors = [] + current_selector = None + for type, string, start, _, _ in tokens: + # ENCODING is only defined in python 3.x + if type == getattr(tokenize, 'ENCODING', None): + continue + elif type in [tokenize.NAME, tokenize.NUMBER]: + current_selector = FormatSelector(SINGLE, string, []) + elif type == tokenize.OP: + if string == ')': + if not inside_group: + # ')' will be handled by the parentheses group + tokens.restore_last_token() + break + elif inside_merge and string in ['/', ',']: + tokens.restore_last_token() + break + elif inside_choice and string == ',': + tokens.restore_last_token() + break + elif string == ',': + if not current_selector: + raise syntax_error('"," must follow a format selector', start) + selectors.append(current_selector) + current_selector = None + elif string == '/': + if not current_selector: + raise syntax_error('"/" must follow a format selector', start) + first_choice = current_selector + second_choice = _parse_format_selection(tokens, inside_choice=True) + current_selector = FormatSelector(PICKFIRST, (first_choice, second_choice), []) + elif string == '[': + if not current_selector: + current_selector = FormatSelector(SINGLE, 'best', []) + format_filter = _parse_filter(tokens) + current_selector.filters.append(format_filter) + elif string == '(': + if current_selector: + raise syntax_error('Unexpected "("', start) + group = _parse_format_selection(tokens, inside_group=True) + current_selector = FormatSelector(GROUP, group, []) + elif string == '+': + video_selector = current_selector + audio_selector = _parse_format_selection(tokens, inside_merge=True) + if not video_selector or not audio_selector: + raise syntax_error('"+" must be between two format selectors', start) + current_selector = FormatSelector(MERGE, (video_selector, audio_selector), []) + else: + raise syntax_error('Operator not recognized: "{0}"'.format(string), start) + elif type == tokenize.ENDMARKER: + break + if current_selector: + selectors.append(current_selector) + return selectors + + def _build_selector_function(selector): + if isinstance(selector, list): + fs = [_build_selector_function(s) for s in selector] + + def selector_function(formats): + for f in fs: + for format in f(formats): + yield format + return selector_function + elif selector.type == GROUP: + selector_function = _build_selector_function(selector.selector) + elif selector.type == PICKFIRST: + fs = [_build_selector_function(s) for s in selector.selector] + + def selector_function(formats): + for f in fs: + picked_formats = list(f(formats)) + if picked_formats: + return picked_formats + return [] + elif selector.type == SINGLE: + format_spec = selector.selector + + def selector_function(formats): + formats = list(formats) + if not formats: + return + if format_spec == 'all': + for f in formats: + yield f + elif format_spec in ['best', 'worst', None]: + format_idx = 0 if format_spec == 'worst' else -1 + audiovideo_formats = [ + f for f in formats + if f.get('vcodec') != 'none' and f.get('acodec') != 'none'] + if audiovideo_formats: + yield audiovideo_formats[format_idx] + # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format + elif (all(f.get('acodec') != 'none' for f in formats) or + all(f.get('vcodec') != 'none' for f in formats)): + yield formats[format_idx] + elif format_spec == 'bestaudio': + audio_formats = [ + f for f in formats + if f.get('vcodec') == 'none'] + if audio_formats: + yield audio_formats[-1] + elif format_spec == 'worstaudio': + audio_formats = [ + f for f in formats + if f.get('vcodec') == 'none'] + if audio_formats: + yield audio_formats[0] + elif format_spec == 'bestvideo': + video_formats = [ + f for f in formats + if f.get('acodec') == 'none'] + if video_formats: + yield video_formats[-1] + elif format_spec == 'worstvideo': + video_formats = [ + f for f in formats + if f.get('acodec') == 'none'] + if video_formats: + yield video_formats[0] + else: + extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] + if format_spec in extensions: + filter_f = lambda f: f['ext'] == format_spec + else: + filter_f = lambda f: f['format_id'] == format_spec + matches = list(filter(filter_f, formats)) + if matches: + yield matches[-1] + elif selector.type == MERGE: + def _merge(formats_info): + format_1, format_2 = [f['format_id'] for f in formats_info] + # The first format must contain the video and the + # second the audio + if formats_info[0].get('vcodec') == 'none': + self.report_error('The first format must ' + 'contain the video, try using ' + '"-f %s+%s"' % (format_2, format_1)) + return + output_ext = ( + formats_info[0]['ext'] + if self.params.get('merge_output_format') is None + else self.params['merge_output_format']) + return { + 'requested_formats': formats_info, + 'format': '%s+%s' % (formats_info[0].get('format'), + formats_info[1].get('format')), + 'format_id': '%s+%s' % (formats_info[0].get('format_id'), + formats_info[1].get('format_id')), + 'width': formats_info[0].get('width'), + 'height': formats_info[0].get('height'), + 'resolution': formats_info[0].get('resolution'), + 'fps': formats_info[0].get('fps'), + 'vcodec': formats_info[0].get('vcodec'), + 'vbr': formats_info[0].get('vbr'), + 'stretched_ratio': formats_info[0].get('stretched_ratio'), + 'acodec': formats_info[1].get('acodec'), + 'abr': formats_info[1].get('abr'), + 'ext': output_ext, + } + video_selector, audio_selector = map(_build_selector_function, selector.selector) - new_format_spec = format_spec[:-len(m.group(0))] - if not new_format_spec: - new_format_spec = 'best' + def selector_function(formats): + formats = list(formats) + for pair in itertools.product(video_selector(formats), audio_selector(formats)): + yield _merge(pair) - return (new_format_spec, new_formats) + filters = [self._build_format_filter(f) for f in selector.filters] - def select_format(self, format_spec, available_formats): - while format_spec.endswith(']'): - format_spec, available_formats = self._apply_format_filter( - format_spec, available_formats) - if not available_formats: - return None + def final_selector(formats): + for _filter in filters: + formats = list(filter(_filter, formats)) + return selector_function(formats) + return final_selector - if format_spec in ['best', 'worst', None]: - format_idx = 0 if format_spec == 'worst' else -1 - audiovideo_formats = [ - f for f in available_formats - if f.get('vcodec') != 'none' and f.get('acodec') != 'none'] - if audiovideo_formats: - return audiovideo_formats[format_idx] - # for audio only (soundcloud) or video only (imgur) urls, select the best/worst audio format - elif (all(f.get('acodec') != 'none' for f in available_formats) or - all(f.get('vcodec') != 'none' for f in available_formats)): - return available_formats[format_idx] - elif format_spec == 'bestaudio': - audio_formats = [ - f for f in available_formats - if f.get('vcodec') == 'none'] - if audio_formats: - return audio_formats[-1] - elif format_spec == 'worstaudio': - audio_formats = [ - f for f in available_formats - if f.get('vcodec') == 'none'] - if audio_formats: - return audio_formats[0] - elif format_spec == 'bestvideo': - video_formats = [ - f for f in available_formats - if f.get('acodec') == 'none'] - if video_formats: - return video_formats[-1] - elif format_spec == 'worstvideo': - video_formats = [ - f for f in available_formats - if f.get('acodec') == 'none'] - if video_formats: - return video_formats[0] - else: - extensions = ['mp4', 'flv', 'webm', '3gp', 'm4a', 'mp3', 'ogg', 'aac', 'wav'] - if format_spec in extensions: - filter_f = lambda f: f['ext'] == format_spec - else: - filter_f = lambda f: f['format_id'] == format_spec - matches = list(filter(filter_f, available_formats)) - if matches: - return matches[-1] - return None + stream = io.BytesIO(format_spec.encode('utf-8')) + try: + tokens = list(_remove_unused_ops(compat_tokenize_tokenize(stream.readline))) + except tokenize.TokenError: + raise syntax_error('Missing closing/opening brackets or parenthesis', (0, len(format_spec))) + + class TokenIterator(object): + def __init__(self, tokens): + self.tokens = tokens + self.counter = 0 + + def __iter__(self): + return self + + def __next__(self): + if self.counter >= len(self.tokens): + raise StopIteration() + value = self.tokens[self.counter] + self.counter += 1 + return value + + next = __next__ + + def restore_last_token(self): + self.counter -= 1 + + parsed_selector = _parse_format_selection(iter(TokenIterator(tokens))) + return _build_selector_function(parsed_selector) def _calc_headers(self, info_dict): res = std_headers.copy() @@ -1111,56 +1308,8 @@ class YoutubeDL(object): req_format_list.append('bestvideo+bestaudio') req_format_list.append('best') req_format = '/'.join(req_format_list) - formats_to_download = [] - if req_format == 'all': - formats_to_download = formats - else: - for rfstr in req_format.split(','): - # We can accept formats requested in the format: 34/5/best, we pick - # the first that is available, starting from left - req_formats = rfstr.split('/') - for rf in req_formats: - if re.match(r'.+?\+.+?', rf) is not None: - # Two formats have been requested like '137+139' - format_1, format_2 = rf.split('+') - formats_info = (self.select_format(format_1, formats), - self.select_format(format_2, formats)) - if all(formats_info): - # The first format must contain the video and the - # second the audio - if formats_info[0].get('vcodec') == 'none': - self.report_error('The first format must ' - 'contain the video, try using ' - '"-f %s+%s"' % (format_2, format_1)) - return - output_ext = ( - formats_info[0]['ext'] - if self.params.get('merge_output_format') is None - else self.params['merge_output_format']) - selected_format = { - 'requested_formats': formats_info, - 'format': '%s+%s' % (formats_info[0].get('format'), - formats_info[1].get('format')), - 'format_id': '%s+%s' % (formats_info[0].get('format_id'), - formats_info[1].get('format_id')), - 'width': formats_info[0].get('width'), - 'height': formats_info[0].get('height'), - 'resolution': formats_info[0].get('resolution'), - 'fps': formats_info[0].get('fps'), - 'vcodec': formats_info[0].get('vcodec'), - 'vbr': formats_info[0].get('vbr'), - 'stretched_ratio': formats_info[0].get('stretched_ratio'), - 'acodec': formats_info[1].get('acodec'), - 'abr': formats_info[1].get('abr'), - 'ext': output_ext, - } - else: - selected_format = None - else: - selected_format = self.select_format(rf, formats) - if selected_format is not None: - formats_to_download.append(selected_format) - break + format_selector = self.build_format_selector(req_format) + formats_to_download = list(format_selector(formats)) if not formats_to_download: raise ExtractorError('requested format not available', expected=True) @@ -1708,27 +1857,6 @@ class YoutubeDL(object): def urlopen(self, req): """ Start an HTTP download """ - - # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not - # always respected by websites, some tend to give out URLs with non percent-encoded - # non-ASCII characters (see telemb.py, ard.py [#3412]) - # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991) - # To work around aforementioned issue we will replace request's original URL with - # percent-encoded one - req_is_string = isinstance(req, compat_basestring) - url = req if req_is_string else req.get_full_url() - url_escaped = escape_url(url) - - # Substitute URL if any change after escaping - if url != url_escaped: - if req_is_string: - req = url_escaped - else: - req_type = HEADRequest if req.get_method() == 'HEAD' else compat_urllib_request.Request - req = req_type( - url_escaped, data=req.data, headers=req.headers, - origin_req_host=req.origin_req_host, unverifiable=req.unverifiable) - return self._opener.open(req, timeout=self._socket_timeout) def print_debug_header(self): diff --git a/youtube_dl/compat.py b/youtube_dl/compat.py index e4b9286c0..ace5bd716 100644 --- a/youtube_dl/compat.py +++ b/youtube_dl/compat.py @@ -436,6 +436,11 @@ except TypeError: # Python 2.6 yield n n += step +if sys.version_info >= (3, 0): + from tokenize import tokenize as compat_tokenize_tokenize +else: + from tokenize import generate_tokens as compat_tokenize_tokenize + __all__ = [ 'compat_HTTPError', 'compat_basestring', @@ -457,6 +462,7 @@ __all__ = [ 'compat_socket_create_connection', 'compat_str', 'compat_subprocess_get_DEVNULL', + 'compat_tokenize_tokenize', 'compat_urllib_error', 'compat_urllib_parse', 'compat_urllib_parse_unquote', diff --git a/youtube_dl/downloader/external.py b/youtube_dl/downloader/external.py index 1d5cc9904..30699934b 100644 --- a/youtube_dl/downloader/external.py +++ b/youtube_dl/downloader/external.py @@ -83,6 +83,16 @@ class CurlFD(ExternalFD): return cmd +class AxelFD(ExternalFD): + def _make_cmd(self, tmpfilename, info_dict): + cmd = [self.exe, '-o', tmpfilename] + for key, val in info_dict['http_headers'].items(): + cmd += ['-H', '%s: %s' % (key, val)] + cmd += self._configuration_args() + cmd += ['--', info_dict['url']] + return cmd + + class WgetFD(ExternalFD): def _make_cmd(self, tmpfilename, info_dict): cmd = [self.exe, '-O', tmpfilename, '-nv', '--no-cookies'] diff --git a/youtube_dl/downloader/http.py b/youtube_dl/downloader/http.py index b7f144af9..a29f5cf31 100644 --- a/youtube_dl/downloader/http.py +++ b/youtube_dl/downloader/http.py @@ -4,6 +4,7 @@ import errno import os import socket import time +import re from .common import FileDownloader from ..compat import ( @@ -57,6 +58,24 @@ class HttpFD(FileDownloader): # Establish connection try: 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 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 + # 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 except (compat_urllib_error.HTTPError, ) as err: if (err.code < 500 or err.code >= 600) and err.code != 416: diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py index 7e5c90829..e38e77a27 100644 --- a/youtube_dl/extractor/__init__.py +++ b/youtube_dl/extractor/__init__.py @@ -118,6 +118,7 @@ from .dailymotion import ( ) from .daum import DaumIE from .dbtv import DBTVIE +from .dcn import DCNIE from .dctp import DctpTvIE from .deezer import DeezerPlaylistIE from .dfb import DFBIE @@ -431,6 +432,10 @@ from .orf import ( from .parliamentliveuk import ParliamentLiveUKIE from .patreon import PatreonIE from .pbs import PBSIE +from .periscope import ( + PeriscopeIE, + QuickscopeIE, +) from .philharmoniedeparis import PhilharmonieDeParisIE from .phoenix import PhoenixIE from .photobucket import PhotobucketIE diff --git a/youtube_dl/extractor/bbc.py b/youtube_dl/extractor/bbc.py index 9a1b6e3dc..abc5a44a1 100644 --- a/youtube_dl/extractor/bbc.py +++ b/youtube_dl/extractor/bbc.py @@ -527,6 +527,18 @@ class BBCIE(BBCCoUkIE): 'skip_download': True, } }, { + # single video from video playlist embedded with vxp-playlist-data JSON + 'url': 'http://www.bbc.com/news/video_and_audio/must_see/33376376', + 'info_dict': { + 'id': 'p02w6qjc', + 'ext': 'mp4', + 'title': '''Judge Mindy Glazer: "I'm sorry to see you here... I always wondered what happened to you"''', + 'duration': 56, + }, + 'params': { + 'skip_download': True, + } + }, { # single video story with digitalData 'url': 'http://www.bbc.com/travel/story/20150625-sri-lankas-spicy-secret', 'info_dict': { @@ -695,13 +707,36 @@ class BBCIE(BBCCoUkIE): if not medias: # Single video article (e.g. http://www.bbc.com/news/video_and_audio/international) - media_asset_page = self._parse_json( + media_asset = self._search_regex( + r'mediaAssetPage\.init\(\s*({.+?}), "/', + webpage, 'media asset', default=None) + if media_asset: + media_asset_page = self._parse_json(media_asset, playlist_id, fatal=False) + medias = [] + for video in media_asset_page.get('videos', {}).values(): + medias.extend(video.values()) + + if not medias: + # Multiple video playlist with single `now playing` entry (e.g. + # http://www.bbc.com/news/video_and_audio/must_see/33767813) + vxp_playlist = self._parse_json( self._search_regex( - r'mediaAssetPage\.init\(\s*({.+?}), "/', webpage, 'media asset'), + r'<script[^>]+class="vxp-playlist-data"[^>]+type="application/json"[^>]*>([^<]+)</script>', + webpage, 'playlist data'), playlist_id) - medias = [] - for video in media_asset_page.get('videos', {}).values(): - medias.extend(video.values()) + playlist_medias = [] + for item in vxp_playlist: + media = item.get('media') + if not media: + continue + playlist_medias.append(media) + # Download single video if found media with asset id matching the video id from URL + if item.get('advert', {}).get('assetId') == playlist_id: + medias = [media] + break + # Fallback to the whole playlist + if not medias: + medias = playlist_medias entries = [] for num, media_meta in enumerate(medias, start=1): diff --git a/youtube_dl/extractor/clipfish.py b/youtube_dl/extractor/clipfish.py index a5c3cb7c6..7af903571 100644 --- a/youtube_dl/extractor/clipfish.py +++ b/youtube_dl/extractor/clipfish.py @@ -1,53 +1,68 @@ from __future__ import unicode_literals import re -import time -import xml.etree.ElementTree from .common import InfoExtractor from ..utils import ( - ExtractorError, - parse_duration, + determine_ext, + int_or_none, + js_to_json, + parse_iso8601, + remove_end, ) class ClipfishIE(InfoExtractor): - IE_NAME = 'clipfish' - - _VALID_URL = r'^https?://(?:www\.)?clipfish\.de/.*?/video/(?P<id>[0-9]+)/' + _VALID_URL = r'https?://(?:www\.)?clipfish\.de/(?:[^/]+/)+video/(?P<id>[0-9]+)' _TEST = { 'url': 'http://www.clipfish.de/special/game-trailer/video/3966754/fifa-14-e3-2013-trailer/', - 'md5': '2521cd644e862936cf2e698206e47385', + 'md5': '79bc922f3e8a9097b3d68a93780fd475', 'info_dict': { 'id': '3966754', 'ext': 'mp4', 'title': 'FIFA 14 - E3 2013 Trailer', + 'timestamp': 1370938118, + 'upload_date': '20130611', 'duration': 82, - }, - 'skip': 'Blocked in the US' + } } def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - video_id = mobj.group(1) - - info_url = ('http://www.clipfish.de/devxml/videoinfo/%s?ts=%d' % - (video_id, int(time.time()))) - doc = self._download_xml( - info_url, video_id, note='Downloading info page') - title = doc.find('title').text - video_url = doc.find('filename').text - if video_url is None: - xml_bytes = xml.etree.ElementTree.tostring(doc) - raise ExtractorError('Cannot find video URL in document %r' % - xml_bytes) - thumbnail = doc.find('imageurl').text - duration = parse_duration(doc.find('duration').text) + video_id = self._match_id(url) + + webpage = self._download_webpage(url, video_id) + + video_info = self._parse_json( + js_to_json(self._html_search_regex( + '(?s)videoObject\s*=\s*({.+?});', webpage, 'video object')), + video_id) + + formats = [] + for video_url in re.findall(r'var\s+videourl\s*=\s*"([^"]+)"', webpage): + ext = determine_ext(video_url) + if ext == 'm3u8': + formats.append({ + 'url': video_url.replace('de.hls.fra.clipfish.de', 'hls.fra.clipfish.de'), + 'ext': 'mp4', + 'format_id': 'hls', + }) + else: + formats.append({ + 'url': video_url, + 'format_id': ext, + }) + self._sort_formats(formats) + + title = remove_end(self._og_search_title(webpage), ' - Video') + thumbnail = self._og_search_thumbnail(webpage) + duration = int_or_none(video_info.get('length')) + timestamp = parse_iso8601(self._html_search_meta('uploadDate', webpage, 'upload date')) return { 'id': video_id, 'title': title, - 'url': video_url, + 'formats': formats, 'thumbnail': thumbnail, 'duration': duration, + 'timestamp': timestamp, } diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index 717dcec7b..def6caa0d 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -638,7 +638,7 @@ class InfoExtractor(object): @staticmethod def _meta_regex(prop): return r'''(?isx)<meta - (?=[^>]+(?:itemprop|name|property)=(["\']?)%s\1) + (?=[^>]+(?:itemprop|name|property|id)=(["\']?)%s\1) [^>]+?content=(["\'])(?P<content>.*?)\2''' % re.escape(prop) def _og_search_property(self, prop, html, name=None, **kargs): diff --git a/youtube_dl/extractor/dailymotion.py b/youtube_dl/extractor/dailymotion.py index 85d945509..2d90b2224 100644 --- a/youtube_dl/extractor/dailymotion.py +++ b/youtube_dl/extractor/dailymotion.py @@ -15,7 +15,6 @@ from ..utils import ( ExtractorError, determine_ext, int_or_none, - orderedSet, parse_iso8601, str_to_int, unescapeHTML, @@ -278,7 +277,7 @@ class DailymotionPlaylistIE(DailymotionBaseInfoExtractor): }] def _extract_entries(self, id): - video_ids = [] + video_ids = set() processed_urls = set() for pagenum in itertools.count(1): page_url = self._PAGE_TEMPLATE % (id, pagenum) @@ -291,12 +290,13 @@ class DailymotionPlaylistIE(DailymotionBaseInfoExtractor): processed_urls.add(urlh.geturl()) - video_ids.extend(re.findall(r'data-xid="(.+?)"', webpage)) + for video_id in re.findall(r'data-xid="(.+?)"', webpage): + if video_id not in video_ids: + yield self.url_result('http://www.dailymotion.com/video/%s' % video_id, 'Dailymotion') + video_ids.add(video_id) if re.search(self._MORE_PAGES_INDICATOR, webpage) is None: break - return [self.url_result('http://www.dailymotion.com/video/%s' % video_id, 'Dailymotion') - for video_id in orderedSet(video_ids)] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) diff --git a/youtube_dl/extractor/dcn.py b/youtube_dl/extractor/dcn.py new file mode 100644 index 000000000..82261e25c --- /dev/null +++ b/youtube_dl/extractor/dcn.py @@ -0,0 +1,84 @@ +# coding: utf-8 +from __future__ import unicode_literals + +from .common import InfoExtractor +from ..compat import ( + compat_urllib_parse, + compat_urllib_request, +) +from ..utils import ( + int_or_none, + parse_iso8601, +) + + +class DCNIE(InfoExtractor): + _VALID_URL = r'https?://(?:www\.)?dcndigital\.ae/(?:#/)?(?:video/.+|show/\d+/.+?)/(?P<id>\d+)' + _TEST = { + 'url': 'http://www.dcndigital.ae/#/show/199074/%D8%B1%D8%AD%D9%84%D8%A9-%D8%A7%D9%84%D8%B9%D9%85%D8%B1-%D8%A7%D9%84%D8%AD%D9%84%D9%82%D8%A9-1/17375/6887', + 'info_dict': + { + 'id': '17375', + 'ext': 'mp4', + 'title': 'رحلة العمر : الحلقة 1', + 'description': 'md5:0156e935d870acb8ef0a66d24070c6d6', + 'thumbnail': 're:^https?://.*\.jpg$', + 'duration': 2041, + 'timestamp': 1227504126, + 'upload_date': '20081124', + }, + 'params': { + # m3u8 download + 'skip_download': True, + }, + } + + def _real_extract(self, url): + video_id = self._match_id(url) + + request = compat_urllib_request.Request( + 'http://admin.mangomolo.com/analytics/index.php/plus/video?id=%s' % video_id, + headers={'Origin': 'http://www.dcndigital.ae'}) + + video = self._download_json(request, video_id) + title = video.get('title_en') or video['title_ar'] + + webpage = self._download_webpage( + 'http://admin.mangomolo.com/analytics/index.php/customers/embed/video?' + + compat_urllib_parse.urlencode({ + 'id': video['id'], + 'user_id': video['user_id'], + 'signature': video['signature'], + 'countries': 'Q0M=', + 'filter': 'DENY', + }), video_id) + + m3u8_url = self._html_search_regex(r'file:\s*"([^"]+)', webpage, 'm3u8 url') + formats = self._extract_m3u8_formats( + m3u8_url, video_id, 'mp4', entry_protocol='m3u8_native', m3u8_id='hls') + + rtsp_url = self._search_regex( + r'<a[^>]+href="(rtsp://[^"]+)"', webpage, 'rtsp url', fatal=False) + if rtsp_url: + formats.append({ + 'url': rtsp_url, + 'format_id': 'rtsp', + }) + + self._sort_formats(formats) + + img = video.get('img') + thumbnail = 'http://admin.mangomolo.com/analytics/%s' % img if img else None + duration = int_or_none(video.get('duration')) + description = video.get('description_en') or video.get('description_ar') + timestamp = parse_iso8601(video.get('create_time') or video.get('update_time'), ' ') + + return { + 'id': video_id, + 'title': title, + 'description': description, + 'thumbnail': thumbnail, + 'duration': duration, + 'timestamp': timestamp, + 'formats': formats, + } diff --git a/youtube_dl/extractor/facebook.py b/youtube_dl/extractor/facebook.py index e17bb9aea..178a7ca4c 100644 --- a/youtube_dl/extractor/facebook.py +++ b/youtube_dl/extractor/facebook.py @@ -17,6 +17,8 @@ from ..utils import ( int_or_none, limit_length, urlencode_postdata, + get_element_by_id, + clean_html, ) @@ -42,6 +44,7 @@ class FacebookIE(InfoExtractor): 'id': '637842556329505', 'ext': 'mp4', 'title': 're:Did you know Kei Nishikori is the first Asian man to ever reach a Grand Slam', + 'uploader': 'Tennis on Facebook', } }, { 'note': 'Video without discernible title', @@ -50,6 +53,7 @@ class FacebookIE(InfoExtractor): 'id': '274175099429670', 'ext': 'mp4', 'title': 'Facebook video #274175099429670', + 'uploader': 'Asif Nawab Butt', }, 'expected_warnings': [ 'title' @@ -161,6 +165,7 @@ class FacebookIE(InfoExtractor): video_title = limit_length(video_title, 80) if not video_title: video_title = 'Facebook video #%s' % video_id + uploader = clean_html(get_element_by_id('fbPhotoPageAuthorName', webpage)) return { 'id': video_id, @@ -168,4 +173,5 @@ class FacebookIE(InfoExtractor): 'formats': formats, 'duration': int_or_none(video_data.get('video_duration')), 'thumbnail': video_data.get('thumbnail_src'), + 'uploader': uploader, } diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py index 27584c44c..901f77304 100644 --- a/youtube_dl/extractor/generic.py +++ b/youtube_dl/extractor/generic.py @@ -304,6 +304,19 @@ class GenericIE(InfoExtractor): }, 'add_ie': ['Ooyala'], }, + { + # ooyala video embedded with http://player.ooyala.com/iframe.js + 'url': 'http://www.macrumors.com/2015/07/24/steve-jobs-the-man-in-the-machine-first-trailer/', + 'info_dict': { + 'id': 'p0MGJndjoG5SOKqO_hZJuZFPB-Tr5VgB', + 'ext': 'mp4', + 'title': '"Steve Jobs: Man in the Machine" trailer', + 'description': 'The first trailer for the Alex Gibney documentary "Steve Jobs: Man in the Machine."', + }, + 'params': { + 'skip_download': True, + }, + }, # multiple ooyala embeds on SBN network websites { 'url': 'http://www.sbnation.com/college-football-recruiting/2015/2/3/7970291/national-signing-day-rationalizations-itll-be-ok-itll-be-ok', @@ -1390,7 +1403,7 @@ class GenericIE(InfoExtractor): return self.url_result(mobj.group('url')) # Look for Ooyala videos - mobj = (re.search(r'player\.ooyala\.com/[^"?]+\?[^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or + mobj = (re.search(r'player\.ooyala\.com/[^"?]+[?#][^"]*?(?:embedCode|ec)=(?P<ec>[^"&]+)', webpage) or re.search(r'OO\.Player\.create\([\'"].*?[\'"],\s*[\'"](?P<ec>.{32})[\'"]', webpage) or re.search(r'SBN\.VideoLinkset\.ooyala\([\'"](?P<ec>.{32})[\'"]\)', webpage) or re.search(r'data-ooyala-video-id\s*=\s*[\'"](?P<ec>.{32})[\'"]', webpage)) @@ -1725,7 +1738,7 @@ class GenericIE(InfoExtractor): if not found: # Broaden the findall a little bit: JWPlayer JS loader found = filter_video(re.findall( - r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)) + r'[^A-Za-z0-9]?(?:file|video_url)["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)) if not found: # Flow player found = filter_video(re.findall(r'''(?xs) diff --git a/youtube_dl/extractor/lynda.py b/youtube_dl/extractor/lynda.py index a00f6e5e5..deead220a 100644 --- a/youtube_dl/extractor/lynda.py +++ b/youtube_dl/extractor/lynda.py @@ -17,7 +17,6 @@ from ..utils import ( class LyndaBaseIE(InfoExtractor): _LOGIN_URL = 'https://www.lynda.com/login/login.aspx' - _SUCCESSFUL_LOGIN_REGEX = r'isLoggedIn: true' _ACCOUNT_CREDENTIALS_HINT = 'Use --username and --password options to provide lynda.com account credentials.' _NETRC_MACHINE = 'lynda' @@ -41,7 +40,7 @@ class LyndaBaseIE(InfoExtractor): request, None, 'Logging in as %s' % username) # Not (yet) logged in - m = re.search(r'loginResultJson = \'(?P<json>[^\']+)\';', login_page) + m = re.search(r'loginResultJson\s*=\s*\'(?P<json>[^\']+)\';', login_page) if m is not None: response = m.group('json') response_json = json.loads(response) @@ -70,7 +69,7 @@ class LyndaBaseIE(InfoExtractor): request, None, 'Confirming log in and log out from another device') - if re.search(self._SUCCESSFUL_LOGIN_REGEX, login_page) is None: + if all(not re.search(p, login_page) for p in ('isLoggedIn\s*:\s*true', r'logout\.aspx', r'>Log out<')): raise ExtractorError('Unable to log in') diff --git a/youtube_dl/extractor/mdr.py b/youtube_dl/extractor/mdr.py index 5fdd19027..fc7499958 100644 --- a/youtube_dl/extractor/mdr.py +++ b/youtube_dl/extractor/mdr.py @@ -29,7 +29,7 @@ class MDRIE(InfoExtractor): doc = self._download_xml(domain + xmlurl, video_id) formats = [] for a in doc.findall('./assets/asset'): - url_el = a.find('.//progressiveDownloadUrl') + url_el = a.find('./progressiveDownloadUrl') if url_el is None: continue abr = int(a.find('bitrateAudio').text) // 1000 diff --git a/youtube_dl/extractor/nowtv.py b/youtube_dl/extractor/nowtv.py index 0b5ff4760..66c627bec 100644 --- a/youtube_dl/extractor/nowtv.py +++ b/youtube_dl/extractor/nowtv.py @@ -1,12 +1,11 @@ # coding: utf-8 from __future__ import unicode_literals -import re - from .common import InfoExtractor from ..compat import compat_str from ..utils import ( ExtractorError, + determine_ext, int_or_none, parse_iso8601, parse_duration, @@ -15,7 +14,7 @@ from ..utils import ( class NowTVIE(InfoExtractor): - _VALID_URL = r'https?://(?:www\.)?nowtv\.de/(?P<station>rtl|rtl2|rtlnitro|superrtl|ntv|vox)/(?P<id>.+?)/player' + _VALID_URL = r'https?://(?:www\.)?nowtv\.(?:de|at|ch)/(?:rtl|rtl2|rtlnitro|superrtl|ntv|vox)/(?P<id>.+?)/(?:player|preview)' _TESTS = [{ # rtl @@ -23,7 +22,7 @@ class NowTVIE(InfoExtractor): 'info_dict': { 'id': '203519', 'display_id': 'bauer-sucht-frau/die-neuen-bauern-und-eine-hochzeit', - 'ext': 'mp4', + 'ext': 'flv', 'title': 'Die neuen Bauern und eine Hochzeit', 'description': 'md5:e234e1ed6d63cf06be5c070442612e7e', 'thumbnail': 're:^https?://.*\.jpg$', @@ -32,7 +31,7 @@ class NowTVIE(InfoExtractor): 'duration': 2786, }, 'params': { - # m3u8 download + # rtmp download 'skip_download': True, }, }, { @@ -41,7 +40,7 @@ class NowTVIE(InfoExtractor): 'info_dict': { 'id': '203481', 'display_id': 'berlin-tag-nacht/berlin-tag-nacht-folge-934', - 'ext': 'mp4', + 'ext': 'flv', 'title': 'Berlin - Tag & Nacht (Folge 934)', 'description': 'md5:c85e88c2e36c552dfe63433bc9506dd0', 'thumbnail': 're:^https?://.*\.jpg$', @@ -50,7 +49,7 @@ class NowTVIE(InfoExtractor): 'duration': 2641, }, 'params': { - # m3u8 download + # rtmp download 'skip_download': True, }, }, { @@ -59,7 +58,7 @@ class NowTVIE(InfoExtractor): 'info_dict': { 'id': '165780', 'display_id': 'alarm-fuer-cobra-11-die-autobahnpolizei/hals-und-beinbruch-2014-08-23-21-10-00', - 'ext': 'mp4', + 'ext': 'flv', 'title': 'Hals- und Beinbruch', 'description': 'md5:b50d248efffe244e6f56737f0911ca57', 'thumbnail': 're:^https?://.*\.jpg$', @@ -68,7 +67,7 @@ class NowTVIE(InfoExtractor): 'duration': 2742, }, 'params': { - # m3u8 download + # rtmp download 'skip_download': True, }, }, { @@ -77,7 +76,7 @@ class NowTVIE(InfoExtractor): 'info_dict': { 'id': '99205', 'display_id': 'medicopter-117/angst', - 'ext': 'mp4', + 'ext': 'flv', 'title': 'Angst!', 'description': 'md5:30cbc4c0b73ec98bcd73c9f2a8c17c4e', 'thumbnail': 're:^https?://.*\.jpg$', @@ -86,7 +85,7 @@ class NowTVIE(InfoExtractor): 'duration': 3025, }, 'params': { - # m3u8 download + # rtmp download 'skip_download': True, }, }, { @@ -95,7 +94,7 @@ class NowTVIE(InfoExtractor): 'info_dict': { 'id': '203521', 'display_id': 'ratgeber-geld/thema-ua-der-erste-blick-die-apple-watch', - 'ext': 'mp4', + 'ext': 'flv', 'title': 'Thema u.a.: Der erste Blick: Die Apple Watch', 'description': 'md5:4312b6c9d839ffe7d8caf03865a531af', 'thumbnail': 're:^https?://.*\.jpg$', @@ -104,7 +103,7 @@ class NowTVIE(InfoExtractor): 'duration': 1083, }, 'params': { - # m3u8 download + # rtmp download 'skip_download': True, }, }, { @@ -113,7 +112,7 @@ class NowTVIE(InfoExtractor): 'info_dict': { 'id': '128953', 'display_id': 'der-hundeprofi/buero-fall-chihuahua-joel', - 'ext': 'mp4', + 'ext': 'flv', 'title': "Büro-Fall / Chihuahua 'Joel'", 'description': 'md5:e62cb6bf7c3cc669179d4f1eb279ad8d', 'thumbnail': 're:^https?://.*\.jpg$', @@ -122,15 +121,19 @@ class NowTVIE(InfoExtractor): 'duration': 3092, }, 'params': { - # m3u8 download + # rtmp download 'skip_download': True, }, + }, { + 'url': 'http://www.nowtv.de/rtl/bauer-sucht-frau/die-neuen-bauern-und-eine-hochzeit/preview', + 'only_matching': True, + }, { + 'url': 'http://www.nowtv.at/rtl/bauer-sucht-frau/die-neuen-bauern-und-eine-hochzeit/preview?return=/rtl/bauer-sucht-frau/die-neuen-bauern-und-eine-hochzeit', + 'only_matching': True, }] def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - display_id = mobj.group('id') - station = mobj.group('station') + display_id = self._match_id(url) info = self._download_json( 'https://api.nowtv.de/v3/movies/%s?fields=id,title,free,geoblocked,articleLong,articleShort,broadcastStartDate,seoUrl,duration,format,files' % display_id, @@ -148,29 +151,19 @@ class NowTVIE(InfoExtractor): raise ExtractorError( 'Video %s is not available for free' % video_id, expected=True) - f = info.get('format', {}) - station = f.get('station') or station - - STATIONS = { - 'rtl': 'rtlnow', - 'rtl2': 'rtl2now', - 'vox': 'voxnow', - 'nitro': 'rtlnitronow', - 'ntv': 'n-tvnow', - 'superrtl': 'superrtlnow' - } - formats = [] for item in files['items']: - item_path = remove_start(item['path'], '/') - tbr = int_or_none(item['bitrate']) - m3u8_url = 'http://hls.fra.%s.de/hls-vod-enc/%s.m3u8' % (STATIONS[station], item_path) - m3u8_url = m3u8_url.replace('now/', 'now/videos/') + if determine_ext(item['path']) != 'f4v': + continue + app, play_path = remove_start(item['path'], '/').split('/', 1) formats.append({ - 'url': m3u8_url, - 'format_id': '%s-%sk' % (item['id'], tbr), - 'ext': 'mp4', - 'tbr': tbr, + 'url': 'rtmpe://fms.rtl.de', + 'app': app, + 'play_path': 'mp4:%s' % play_path, + 'ext': 'flv', + 'page_url': url, + 'player_url': 'http://rtl-now.rtl.de/includes/nc_player.swf', + 'tbr': int_or_none(item.get('bitrate')), }) self._sort_formats(formats) @@ -178,6 +171,8 @@ class NowTVIE(InfoExtractor): description = info.get('articleLong') or info.get('articleShort') timestamp = parse_iso8601(info.get('broadcastStartDate'), ' ') duration = parse_duration(info.get('duration')) + + f = info.get('format', {}) thumbnail = f.get('defaultImage169Format') or f.get('defaultImage169Logo') return { diff --git a/youtube_dl/extractor/periscope.py b/youtube_dl/extractor/periscope.py new file mode 100644 index 000000000..8ad936758 --- /dev/null +++ b/youtube_dl/extractor/periscope.py @@ -0,0 +1,99 @@ +# coding: utf-8 +from __future__ import unicode_literals + +from .common import InfoExtractor +from ..compat import ( + compat_urllib_parse, + compat_urllib_request, +) +from ..utils import parse_iso8601 + + +class PeriscopeIE(InfoExtractor): + IE_DESC = 'Periscope' + _VALID_URL = r'https?://(?:www\.)?periscope\.tv/w/(?P<id>[^/?#]+)' + _TEST = { + 'url': 'https://www.periscope.tv/w/aJUQnjY3MjA3ODF8NTYxMDIyMDl2zCg2pECBgwTqRpQuQD352EMPTKQjT4uqlM3cgWFA-g==', + 'md5': '65b57957972e503fcbbaeed8f4fa04ca', + 'info_dict': { + 'id': '56102209', + 'ext': 'mp4', + 'title': 'Bec Boop - 🚠✈️🇬🇧 Fly above #London in Emirates Air Line cable car at night 🇬🇧✈️🚠 #BoopScope 🎀💗', + 'timestamp': 1438978559, + 'upload_date': '20150807', + 'uploader': 'Bec Boop', + 'uploader_id': '1465763', + }, + 'skip': 'Expires in 24 hours', + } + + def _call_api(self, method, token): + return self._download_json( + 'https://api.periscope.tv/api/v2/%s?token=%s' % (method, token), token) + + def _real_extract(self, url): + token = self._match_id(url) + + broadcast_data = self._call_api('getBroadcastPublic', token) + broadcast = broadcast_data['broadcast'] + status = broadcast['status'] + + uploader = broadcast.get('user_display_name') or broadcast_data.get('user', {}).get('display_name') + uploader_id = broadcast.get('user_id') or broadcast_data.get('user', {}).get('id') + + title = '%s - %s' % (uploader, status) if uploader else status + state = broadcast.get('state').lower() + if state == 'running': + title = self._live_title(title) + timestamp = parse_iso8601(broadcast.get('created_at')) + + thumbnails = [{ + 'url': broadcast[image], + } for image in ('image_url', 'image_url_small') if broadcast.get(image)] + + stream = self._call_api('getAccessPublic', token) + + formats = [] + for format_id in ('replay', 'rtmp', 'hls', 'https_hls'): + video_url = stream.get(format_id + '_url') + if not video_url: + continue + f = { + 'url': video_url, + 'ext': 'flv' if format_id == 'rtmp' else 'mp4', + } + if format_id != 'rtmp': + f['protocol'] = 'm3u8_native' if state == 'ended' else 'm3u8' + formats.append(f) + self._sort_formats(formats) + + return { + 'id': broadcast.get('id') or token, + 'title': title, + 'timestamp': timestamp, + 'uploader': uploader, + 'uploader_id': uploader_id, + 'thumbnails': thumbnails, + 'formats': formats, + } + + +class QuickscopeIE(InfoExtractor): + IE_DESC = 'Quick Scope' + _VALID_URL = r'https?://watchonperiscope\.com/broadcast/(?P<id>\d+)' + _TEST = { + 'url': 'https://watchonperiscope.com/broadcast/56180087', + 'only_matching': True, + } + + def _real_extract(self, url): + broadcast_id = self._match_id(url) + request = compat_urllib_request.Request( + 'https://watchonperiscope.com/api/accessChannel', compat_urllib_parse.urlencode({ + 'broadcast_id': broadcast_id, + 'entry_ticket': '', + 'from_push': 'false', + 'uses_sessions': 'true', + }).encode('utf-8')) + return self.url_result( + self._download_json(request, broadcast_id)['share_url'], 'Periscope') diff --git a/youtube_dl/extractor/pornhub.py b/youtube_dl/extractor/pornhub.py index 0b7886840..7b0cdc41a 100644 --- a/youtube_dl/extractor/pornhub.py +++ b/youtube_dl/extractor/pornhub.py @@ -81,7 +81,7 @@ class PornHubIE(InfoExtractor): comment_count = self._extract_count( r'All Comments\s*<span>\(([\d,.]+)\)', webpage, 'comment') - video_urls = list(map(compat_urllib_parse_unquote, re.findall(r'"quality_[0-9]{3}p":"([^"]+)', webpage))) + video_urls = list(map(compat_urllib_parse_unquote, re.findall(r"player_quality_[0-9]{3}p\s*=\s*'([^']+)'", webpage))) if webpage.find('"encrypted":true') != -1: password = compat_urllib_parse_unquote_plus( self._search_regex(r'"video_title":"([^"]+)', webpage, 'password')) @@ -94,7 +94,7 @@ class PornHubIE(InfoExtractor): format = path.split('/')[5].split('_')[:2] format = "-".join(format) - m = re.match(r'^(?P<height>[0-9]+)P-(?P<tbr>[0-9]+)K$', format) + m = re.match(r'^(?P<height>[0-9]+)[pP]-(?P<tbr>[0-9]+)[kK]$', format) if m is None: height = None tbr = None diff --git a/youtube_dl/extractor/screenwavemedia.py b/youtube_dl/extractor/screenwavemedia.py index d1ab66b32..3bc84989e 100644 --- a/youtube_dl/extractor/screenwavemedia.py +++ b/youtube_dl/extractor/screenwavemedia.py @@ -1,12 +1,11 @@ # encoding: utf-8 from __future__ import unicode_literals -import re - from .common import InfoExtractor from ..utils import ( int_or_none, unified_strdate, + js_to_json, ) @@ -22,59 +21,48 @@ class ScreenwaveMediaIE(InfoExtractor): video_id = self._match_id(url) playerdata = self._download_webpage( - 'http://player.screenwavemedia.com/play/player.php?id=%s' % video_id, + 'http://player.screenwavemedia.com/player.php?id=%s' % video_id, video_id, 'Downloading player webpage') vidtitle = self._search_regex( r'\'vidtitle\'\s*:\s*"([^"]+)"', playerdata, 'vidtitle').replace('\\/', '/') - vidurl = self._search_regex( - r'\'vidurl\'\s*:\s*"([^"]+)"', playerdata, 'vidurl').replace('\\/', '/') - - videolist_url = None - - mobj = re.search(r"'videoserver'\s*:\s*'(?P<videoserver>[^']+)'", playerdata) - if mobj: - videoserver = mobj.group('videoserver') - mobj = re.search(r'\'vidid\'\s*:\s*"(?P<vidid>[^\']+)"', playerdata) - vidid = mobj.group('vidid') if mobj else video_id - videolist_url = 'http://%s/vod/smil:%s.smil/jwplayer.smil' % (videoserver, vidid) - else: - mobj = re.search(r"file\s*:\s*'(?P<smil>http.+?/jwplayer\.smil)'", playerdata) - if mobj: - videolist_url = mobj.group('smil') - - if videolist_url: - videolist = self._download_xml(videolist_url, video_id, 'Downloading videolist XML') - formats = [] - baseurl = vidurl[:vidurl.rfind('/') + 1] - for video in videolist.findall('.//video'): - src = video.get('src') - if not src: - continue - file_ = src.partition(':')[-1] - width = int_or_none(video.get('width')) - height = int_or_none(video.get('height')) - bitrate = int_or_none(video.get('system-bitrate'), scale=1000) - format = { - 'url': baseurl + file_, - 'format_id': src.rpartition('.')[0].rpartition('_')[-1], - } - if width or height: - format.update({ - 'tbr': bitrate, - 'width': width, - 'height': height, - }) - else: - format.update({ - 'abr': bitrate, - 'vcodec': 'none', - }) - formats.append(format) - else: - formats = [{ - 'url': vidurl, - }] + + playerconfig = self._download_webpage( + 'http://player.screenwavemedia.com/player.js', + video_id, 'Downloading playerconfig webpage') + + videoserver = self._search_regex(r"\[ipaddress\]\s*=>\s*([\d\.]+)", playerdata, 'videoserver') + + sources = self._parse_json( + js_to_json( + self._search_regex( + r"sources\s*:\s*(\[[^\]]+?\])", playerconfig, + 'sources', + ).replace( + "' + thisObj.options.videoserver + '", + videoserver + ).replace( + "' + playerVidId + '", + video_id + ) + ), + video_id + ) + + formats = [] + for source in sources: + if source['type'] == 'hls': + formats.extend(self._extract_m3u8_formats(source['file'], video_id)) + else: + format_label = source.get('label') + height = int_or_none(self._search_regex( + r'^(\d+)[pP]', format_label, 'height', default=None)) + formats.append({ + 'url': source['file'], + 'format': format_label, + 'ext': source.get('type'), + 'height': height, + }) self._sort_formats(formats) return { diff --git a/youtube_dl/extractor/southpark.py b/youtube_dl/extractor/southpark.py index 7fb165a87..87b650468 100644 --- a/youtube_dl/extractor/southpark.py +++ b/youtube_dl/extractor/southpark.py @@ -45,6 +45,14 @@ class SouthParkDeIE(SouthParkIE): 'title': 'The Government Won\'t Respect My Privacy', 'description': 'Cartman explains the benefits of "Shitter" to Stan, Kyle and Craig.', }, + }, { + # non-ASCII characters in initial URL + 'url': 'http://www.southpark.de/alle-episoden/s18e09-hashtag-aufwärmen', + 'playlist_count': 4, + }, { + # non-ASCII characters in redirect URL + 'url': 'http://www.southpark.de/alle-episoden/s18e09', + 'playlist_count': 4, }] diff --git a/youtube_dl/extractor/tudou.py b/youtube_dl/extractor/tudou.py index c89de5ba4..84fe71aef 100644 --- a/youtube_dl/extractor/tudou.py +++ b/youtube_dl/extractor/tudou.py @@ -29,6 +29,8 @@ class TudouIE(InfoExtractor): } }] + _PLAYER_URL = 'http://js.tudouui.com/bin/lingtong/PortalPlayer_177.swf' + def _url_for_id(self, id, quality=None): info_url = "http://v2.tudou.com/f?id=" + str(id) if quality: @@ -54,6 +56,10 @@ class TudouIE(InfoExtractor): thumbnail_url = self._search_regex( r",pic:\s*[\"'](.+?)[\"']", webpage, 'thumbnail URL', fatal=False) + player_url = self._search_regex( + r"playerUrl\s*:\s*['\"](.+?\.swf)[\"']", + webpage, 'player URL', default=self._PLAYER_URL) + segs_json = self._search_regex(r'segs: \'(.*)\'', webpage, 'segments') segments = json.loads(segs_json) # It looks like the keys are the arguments that have to be passed as @@ -76,6 +82,9 @@ class TudouIE(InfoExtractor): 'ext': ext, 'title': title, 'thumbnail': thumbnail_url, + 'http_headers': { + 'Referer': player_url, + }, } result.append(part_info) diff --git a/youtube_dl/extractor/twitch.py b/youtube_dl/extractor/twitch.py index 73ce335b7..a2b6a35aa 100644 --- a/youtube_dl/extractor/twitch.py +++ b/youtube_dl/extractor/twitch.py @@ -7,12 +7,15 @@ import random from .common import InfoExtractor from ..compat import ( + compat_parse_qs, compat_str, compat_urllib_parse, + compat_urllib_parse_urlparse, compat_urllib_request, ) from ..utils import ( ExtractorError, + parse_duration, parse_iso8601, ) @@ -185,7 +188,7 @@ class TwitchVodIE(TwitchItemBaseIE): _ITEM_SHORTCUT = 'v' _TEST = { - 'url': 'http://www.twitch.tv/riotgames/v/6528877', + 'url': 'http://www.twitch.tv/riotgames/v/6528877?t=5m10s', 'info_dict': { 'id': 'v6528877', 'ext': 'mp4', @@ -197,6 +200,7 @@ class TwitchVodIE(TwitchItemBaseIE): 'uploader': 'Riot Games', 'uploader_id': 'riotgames', 'view_count': int, + 'start_time': 310, }, 'params': { # m3u8 download @@ -216,6 +220,12 @@ class TwitchVodIE(TwitchItemBaseIE): item_id, 'mp4') self._prefer_source(formats) info['formats'] = formats + + parsed_url = compat_urllib_parse_urlparse(url) + query = compat_parse_qs(parsed_url.query) + if 't' in query: + info['start_time'] = parse_duration(query['t'][0]) + return info diff --git a/youtube_dl/extractor/xhamster.py b/youtube_dl/extractor/xhamster.py index b4ad513a0..97315750f 100644 --- a/youtube_dl/extractor/xhamster.py +++ b/youtube_dl/extractor/xhamster.py @@ -4,7 +4,6 @@ import re from .common import InfoExtractor from ..utils import ( - ExtractorError, unified_strdate, str_to_int, int_or_none, @@ -22,7 +21,7 @@ class XHamsterIE(InfoExtractor): 'ext': 'mp4', 'title': 'FemaleAgent Shy beauty takes the bait', 'upload_date': '20121014', - 'uploader_id': 'Ruseful2011', + 'uploader': 'Ruseful2011', 'duration': 893, 'age_limit': 18, } @@ -34,7 +33,7 @@ class XHamsterIE(InfoExtractor): 'ext': 'mp4', 'title': 'Britney Spears Sexy Booty', 'upload_date': '20130914', - 'uploader_id': 'jojo747400', + 'uploader': 'jojo747400', 'duration': 200, 'age_limit': 18, } @@ -46,12 +45,12 @@ class XHamsterIE(InfoExtractor): ] def _real_extract(self, url): - def extract_video_url(webpage): - mp4 = re.search(r'<video\s+.*?file="([^"]+)".*?>', webpage) - if mp4 is None: - raise ExtractorError('Unable to extract media URL') - else: - return mp4.group(1) + def extract_video_url(webpage, name): + return self._search_regex( + [r'''file\s*:\s*(?P<q>["'])(?P<mp4>.+?)(?P=q)''', + r'''<a\s+href=(?P<q>["'])(?P<mp4>.+?)(?P=q)\s+class=["']mp4Thumb''', + r'''<video[^>]+file=(?P<q>["'])(?P<mp4>.+?)(?P=q)[^>]*>'''], + webpage, name, group='mp4') def is_hd(webpage): return '<div class=\'icon iconHD\'' in webpage @@ -75,10 +74,14 @@ class XHamsterIE(InfoExtractor): if upload_date: upload_date = unified_strdate(upload_date) - uploader_id = self._html_search_regex(r'<a href=\'/user/[^>]+>(?P<uploader_id>[^<]+)', - webpage, 'uploader id', default='anonymous') + uploader = self._html_search_regex( + r"<a href='[^']+xhamster\.com/user/[^>]+>(?P<uploader>[^<]+)", + webpage, 'uploader', default='anonymous') - thumbnail = self._html_search_regex(r'<video\s+.*?poster="([^"]+)".*?>', webpage, 'thumbnail', fatal=False) + thumbnail = self._search_regex( + [r'''thumb\s*:\s*(?P<q>["'])(?P<thumbnail>.+?)(?P=q)''', + r'''<video[^>]+poster=(?P<q>["'])(?P<thumbnail>.+?)(?P=q)[^>]*>'''], + webpage, 'thumbnail', fatal=False, group='thumbnail') duration = parse_duration(self._html_search_regex(r'<span>Runtime:</span> (\d+:\d+)</div>', webpage, 'duration', fatal=False)) @@ -97,7 +100,9 @@ class XHamsterIE(InfoExtractor): hd = is_hd(webpage) - video_url = extract_video_url(webpage) + format_id = 'hd' if hd else 'sd' + + video_url = extract_video_url(webpage, format_id) formats = [{ 'url': video_url, 'format_id': 'hd' if hd else 'sd', @@ -108,7 +113,7 @@ class XHamsterIE(InfoExtractor): mrss_url = self._search_regex(r'<link rel="canonical" href="([^"]+)', webpage, 'mrss_url') webpage = self._download_webpage(mrss_url + '?hd', video_id, note='Downloading HD webpage') if is_hd(webpage): - video_url = extract_video_url(webpage) + video_url = extract_video_url(webpage, 'hd') formats.append({ 'url': video_url, 'format_id': 'hd', @@ -122,7 +127,7 @@ class XHamsterIE(InfoExtractor): 'title': title, 'description': description, 'upload_date': upload_date, - 'uploader_id': uploader_id, + 'uploader': uploader, 'thumbnail': thumbnail, 'duration': duration, 'view_count': view_count, diff --git a/youtube_dl/utils.py b/youtube_dl/utils.py index 78dc2b449..e265c7574 100644 --- a/youtube_dl/utils.py +++ b/youtube_dl/utils.py @@ -651,6 +651,26 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): return ret def http_request(self, req): + # According to RFC 3986, URLs can not contain non-ASCII characters, however this is not + # always respected by websites, some tend to give out URLs with non percent-encoded + # non-ASCII characters (see telemb.py, ard.py [#3412]) + # urllib chokes on URLs with non-ASCII characters (see http://bugs.python.org/issue3991) + # To work around aforementioned issue we will replace request's original URL with + # percent-encoded one + # Since redirects are also affected (e.g. http://www.southpark.de/alle-episoden/s18e09) + # the code of this workaround has been moved here from YoutubeDL.urlopen() + url = req.get_full_url() + url_escaped = escape_url(url) + + # Substitute URL if any change after escaping + if url != url_escaped: + req_type = HEADRequest if req.get_method() == 'HEAD' else compat_urllib_request.Request + new_req = req_type( + url_escaped, data=req.data, headers=req.headers, + origin_req_host=req.origin_req_host, unverifiable=req.unverifiable) + new_req.timeout = req.timeout + req = new_req + for h, v in std_headers.items(): # Capitalize is needed because of Python bug 2275: http://bugs.python.org/issue2275 # The dict keys are capitalized because of this bug by urllib @@ -695,6 +715,17 @@ class YoutubeDLHandler(compat_urllib_request.HTTPHandler): gz = io.BytesIO(self.deflate(resp.read())) resp = self.addinfourl_wrapper(gz, old_resp.headers, old_resp.url, old_resp.code) resp.msg = old_resp.msg + # Percent-encode redirect URL of Location HTTP header to satisfy RFC 3986 + if 300 <= resp.code < 400: + location = resp.headers.get('Location') + if location: + # As of RFC 2616 default charset is iso-8859-1 that is respected by python 3 + if sys.version_info >= (3, 0): + location = location.encode('iso-8859-1').decode('utf-8') + location_escaped = escape_url(location) + if location != location_escaped: + del resp.headers['Location'] + resp.headers['Location'] = location_escaped return resp https_request = http_request diff --git a/youtube_dl/version.py b/youtube_dl/version.py index fa157cadb..9f209499c 100644 --- a/youtube_dl/version.py +++ b/youtube_dl/version.py @@ -1,3 +1,3 @@ from __future__ import unicode_literals -__version__ = '2015.07.28' +__version__ = '2015.08.06.1' |