X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/a070911bf98c751e4425008f1335cc57c86d30fc..d317bf370d3cbd2231f3d6b1970e0a86fddc4e5b:/youtube_dl/extractor/common.py diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index 03f3f18..b9014fc 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -14,6 +14,7 @@ import xml.etree.ElementTree from ..compat import ( compat_cookiejar, + compat_HTTPError, compat_http_client, compat_urllib_error, compat_urllib_parse_urlparse, @@ -21,17 +22,20 @@ from ..compat import ( compat_str, ) from ..utils import ( + NO_DEFAULT, age_restricted, + bug_reports_message, clean_html, compiled_regex_type, + determine_ext, ExtractorError, + fix_xml_ampersands, float_or_none, int_or_none, RegexNotFoundError, sanitize_filename, unescapeHTML, ) -_NO_DEFAULT = object() class InfoExtractor(object): @@ -45,7 +49,7 @@ class InfoExtractor(object): information possibly downloading the video to the file system, among other possible outcomes. - The type field determines the the type of the result. + The type field determines the type of the result. By far the most common value (and the default if _type is missing) is "video", which indicates a single video. @@ -87,7 +91,8 @@ class InfoExtractor(object): * player_url SWF Player URL (used for rtmpdump). * protocol The protocol that will be used for the actual download, lower-case. - "http", "https", "rtsp", "rtmp", "m3u8" or so. + "http", "https", "rtsp", "rtmp", "rtmpe", + "m3u8", or "m3u8_native". * preference Order number of this format. If this field is present and not None, the formats get sorted by this field, regardless of all other values. @@ -108,15 +113,14 @@ class InfoExtractor(object): (quality takes higher priority) -1 for default (order by other properties), -2 or smaller for less than default. - * http_referer HTTP Referer header value to set. - * http_method HTTP method to use for the download. * http_headers A dictionary of additional HTTP headers to add to the request. - * http_post_data Additional data to send with a POST - request. * stretched_ratio If given and not 1, indicates that the - video's pixels are not square. - width : height ratio as float. + video's pixels are not square. + width : height ratio as float. + * no_resume The server does not support resuming the + (HTTP or RTMP) download. Boolean. + url: Final video URL. ext: Video filename extension. format: The video format, defaults to ext (used for --get-format) @@ -130,7 +134,9 @@ class InfoExtractor(object): something like "4234987", title "Dancing naked mole rats", and display_id "dancing-naked-mole-rats" thumbnails: A list of dictionaries, with the following entries: + * "id" (optional, string) - Thumbnail format ID * "url" + * "preference" (optional, int) - quality of the image * "width" (optional, int) * "height" (optional, int) * "resolution" (optional, string "{width}x{height"}, @@ -138,17 +144,25 @@ class InfoExtractor(object): thumbnail: Full URL to a video thumbnail image. description: Full video description. uploader: Full name of the video uploader. + creator: The main artist who created the video. timestamp: UNIX timestamp of the moment the video became available. upload_date: Video upload date (YYYYMMDD). If not explicitly set, calculated from timestamp. uploader_id: Nickname or id of the video uploader. location: Physical location where the video was filmed. - subtitles: The subtitle file contents as a dictionary in the format - {language: subtitles}. + subtitles: The available subtitles as a dictionary in the format + {language: subformats}. "subformats" is a list sorted from + lower to higher preference, each element is a dictionary + with the "ext" entry and one of: + * "data": The subtitles file contents + * "url": A url pointing to the subtitles file + automatic_captions: Like 'subtitles', used by the YoutubeIE for + automatically generated captions duration: Length of the video in seconds, as an integer. view_count: How many users have watched the video on the platform. like_count: Number of positive ratings of the video dislike_count: Number of negative ratings of the video + average_rating: Average rating give by users, the scale used depends on the webpage comment_count: Number of comments on the video comments: A list of comments, each with one or more of the following properties (all but one of text or html optional): @@ -256,8 +270,15 @@ class InfoExtractor(object): def extract(self, url): """Extracts URL information and returns it in list of dicts.""" - self.initialize() - return self._real_extract(url) + try: + self.initialize() + return self._real_extract(url) + except ExtractorError: + raise + except compat_http_client.IncompleteRead as e: + raise ExtractorError('A network error has occured.', cause=e, expected=True) + except (KeyError, StopIteration) as e: + raise ExtractorError('An extractor error has occured.', cause=e) def set_downloader(self, downloader): """Sets the downloader for this IE.""" @@ -303,7 +324,7 @@ class InfoExtractor(object): self._downloader.report_warning(errmsg) return False - def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True): + def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None): """ Returns a tuple (page content as string, URL handle) """ # Strip hashes from the URL (#1038) if isinstance(url_or_request, (compat_str, str)): @@ -313,14 +334,11 @@ class InfoExtractor(object): if urlh is False: assert not fatal return False - content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal) + content = self._webpage_read_content(urlh, url_or_request, video_id, note, errnote, fatal, encoding=encoding) return (content, urlh) - def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None): - content_type = urlh.headers.get('Content-Type', '') - webpage_bytes = urlh.read() - if prefix is not None: - webpage_bytes = prefix + webpage_bytes + @staticmethod + def _guess_encoding_from_content(content_type, webpage_bytes): m = re.match(r'[a-zA-Z0-9_.-]+/[a-zA-Z0-9_.-]+\s*;\s*charset=(.+)', content_type) if m: encoding = m.group(1) @@ -333,6 +351,16 @@ class InfoExtractor(object): encoding = 'utf-16' else: encoding = 'utf-8' + + return encoding + + def _webpage_read_content(self, urlh, url_or_request, video_id, note=None, errnote=None, fatal=True, prefix=None, encoding=None): + content_type = urlh.headers.get('Content-Type', '') + webpage_bytes = urlh.read() + if prefix is not None: + webpage_bytes = prefix + webpage_bytes + if not encoding: + encoding = self._guess_encoding_from_content(content_type, webpage_bytes) if self._downloader.params.get('dump_intermediate_pages', False): try: url = url_or_request.get_full_url() @@ -376,16 +404,26 @@ class InfoExtractor(object): if blocked_iframe: msg += ' Visit %s for more details' % blocked_iframe raise ExtractorError(msg, expected=True) + if 'The URL you requested has been blocked' in content[:512]: + msg = ( + 'Access to this webpage has been blocked by Indian censorship. ' + 'Use a VPN or proxy server (with --proxy) to route around it.') + block_msg = self._html_search_regex( + r'

(.*?)

', + content, 'block message', default=None) + if block_msg: + msg += ' (Message: "%s")' % block_msg.replace('\n', ' ') + raise ExtractorError(msg, expected=True) return content - def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5): + def _download_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True, tries=1, timeout=5, encoding=None): """ Returns the data of the page as a string """ success = False try_count = 0 while success is False: try: - res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal) + res = self._download_webpage_handle(url_or_request, video_id, note, errnote, fatal, encoding=encoding) success = True except compat_http_client.IncompleteRead as e: try_count += 1 @@ -400,10 +438,10 @@ class InfoExtractor(object): def _download_xml(self, url_or_request, video_id, note='Downloading XML', errnote='Unable to download XML', - transform_source=None, fatal=True): + transform_source=None, fatal=True, encoding=None): """Return the xml as an xml.etree.ElementTree.Element""" xml_string = self._download_webpage( - url_or_request, video_id, note, errnote, fatal=fatal) + url_or_request, video_id, note, errnote, fatal=fatal, encoding=encoding) if xml_string is False: return xml_string if transform_source: @@ -414,9 +452,10 @@ class InfoExtractor(object): note='Downloading JSON metadata', errnote='Unable to download JSON metadata', transform_source=None, - fatal=True): + fatal=True, encoding=None): json_string = self._download_webpage( - url_or_request, video_id, note, errnote, fatal=fatal) + url_or_request, video_id, note, errnote, fatal=fatal, + encoding=encoding) if (not fatal) and json_string is False: return None return self._parse_json( @@ -461,7 +500,7 @@ class InfoExtractor(object): # Methods for following #608 @staticmethod - def url_result(url, ie=None, video_id=None): + def url_result(url, ie=None, video_id=None, video_title=None): """Returns a url that points to a page that should be processed""" # TODO: ie should be the class used for getting the info video_info = {'_type': 'url', @@ -469,6 +508,8 @@ class InfoExtractor(object): 'ie_key': ie} if video_id is not None: video_info['id'] = video_id + if video_title is not None: + video_info['title'] = video_title return video_info @staticmethod @@ -484,7 +525,7 @@ class InfoExtractor(object): video_info['description'] = playlist_description return video_info - def _search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None): + def _search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): """ Perform a regex search on the given string, using a single or a list of patterns returning the first matching group. @@ -499,7 +540,7 @@ class InfoExtractor(object): if mobj: break - if os.name != 'nt' and sys.stderr.isatty(): + if not self._downloader.params.get('no_color') and os.name != 'nt' and sys.stderr.isatty(): _name = '\033[0;34m%s\033[0m' % name else: _name = name @@ -510,16 +551,15 @@ class InfoExtractor(object): return next(g for g in mobj.groups() if g is not None) else: return mobj.group(group) - elif default is not _NO_DEFAULT: + elif default is not NO_DEFAULT: return default elif fatal: raise RegexNotFoundError('Unable to extract %s' % _name) else: - self._downloader.report_warning('unable to extract %s; ' - 'please report this issue on http://yt-dl.org/bug' % _name) + self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message()) return None - def _html_search_regex(self, pattern, string, name, default=_NO_DEFAULT, fatal=True, flags=0, group=None): + def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None): """ Like _search_regex, but strips HTML tags and unescapes entities. """ @@ -531,7 +571,7 @@ class InfoExtractor(object): def _get_login_info(self): """ - Get the the login info as (username, password) + Get the login info as (username, password) It will look in the netrc file using the _NETRC_MACHINE value If there's no info available, return (None, None) """ @@ -648,11 +688,45 @@ class InfoExtractor(object): } return RATING_TABLE.get(rating.lower(), None) + def _family_friendly_search(self, html): + # See http://schema.org/VideoObject + family_friendly = self._html_search_meta('isFamilyFriendly', html) + + if not family_friendly: + return None + + RATING_TABLE = { + '1': 0, + 'true': 0, + '0': 18, + 'false': 18, + } + return RATING_TABLE.get(family_friendly.lower(), None) + def _twitter_search_player(self, html): return self._html_search_meta('twitter:player', html, 'twitter card player') - def _sort_formats(self, formats): + @staticmethod + def _hidden_inputs(html): + return dict([ + (input.group('name'), input.group('value')) for input in re.finditer( + r'''(?x) + ["\'])hidden(?P=q_hidden)\s+ + name=(?P["\'])(?P.+?)(?P=q_name)\s+ + (?:id=(?P["\']).+?(?P=q_id)\s+)? + value=(?P["\'])(?P.*?)(?P=q_value) + ''', html) + ]) + + def _form_hidden_inputs(self, form_id, html): + form = self._search_regex( + r'(?s)]+?id=(["\'])%s\1[^>]*>(?P
.+?)
' % form_id, + html, '%s form' % form_id, group='form') + return self._hidden_inputs(form) + + def _sort_formats(self, formats, field_preference=None): if not formats: raise ExtractorError('No video formats found') @@ -662,6 +736,9 @@ class InfoExtractor(object): if not f.get('ext') and 'url' in f: f['ext'] = determine_ext(f['url']) + if isinstance(field_preference, (list, tuple)): + return tuple(f.get(field) if f.get(field) is not None else -1 for field in field_preference) + preference = f.get('preference') if preference is None: proto = f.get('protocol') @@ -697,21 +774,44 @@ class InfoExtractor(object): preference, f.get('language_preference') if f.get('language_preference') is not None else -1, f.get('quality') if f.get('quality') is not None else -1, + f.get('tbr') if f.get('tbr') is not None else -1, + f.get('filesize') if f.get('filesize') is not None else -1, + f.get('vbr') if f.get('vbr') is not None else -1, f.get('height') if f.get('height') is not None else -1, f.get('width') if f.get('width') is not None else -1, ext_preference, - f.get('tbr') if f.get('tbr') is not None else -1, - f.get('vbr') if f.get('vbr') is not None else -1, f.get('abr') if f.get('abr') is not None else -1, audio_ext_preference, f.get('fps') if f.get('fps') is not None else -1, - f.get('filesize') if f.get('filesize') is not None else -1, f.get('filesize_approx') if f.get('filesize_approx') is not None else -1, f.get('source_preference') if f.get('source_preference') is not None else -1, - f.get('format_id'), + f.get('format_id') if f.get('format_id') is not None else '', ) formats.sort(key=_formats_key) + def _check_formats(self, formats, video_id): + if formats: + formats[:] = filter( + lambda f: self._is_valid_url( + f['url'], video_id, + item='%s video format' % f.get('format_id') if f.get('format_id') else 'video'), + formats) + + def _is_valid_url(self, url, video_id, item='video'): + url = self._proto_relative_url(url, scheme='http:') + # For now assume non HTTP(S) URLs always valid + if not (url.startswith('http://') or url.startswith('https://')): + return True + try: + self._request_webpage(url, video_id, 'Checking %s URL' % item) + return True + except ExtractorError as e: + if isinstance(e.cause, compat_HTTPError): + self.to_screen( + '%s: %s URL is invalid, skipping' % (video_id, item)) + return False + raise + def http_scheme(self): """ Either "http:" or "https:", depending on the user's preferences """ return ( @@ -736,10 +836,14 @@ class InfoExtractor(object): self.to_screen(msg) time.sleep(timeout) - def _extract_f4m_formats(self, manifest_url, video_id): + def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None, + transform_source=lambda s: fix_xml_ampersands(s).strip()): manifest = self._download_xml( manifest_url, video_id, 'Downloading f4m manifest', - 'Unable to download f4m manifest') + 'Unable to download f4m manifest', + # Some manifests may be malformed, e.g. prosiebensat1 generated manifests + # (see https://github.com/rg3/youtube-dl/issues/6215#issuecomment-121704244) + transform_source=transform_source) formats = [] manifest_version = '1.0' @@ -749,30 +853,44 @@ class InfoExtractor(object): media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media') for i, media_el in enumerate(media_nodes): if manifest_version == '2.0': - manifest_url = '/'.join(manifest_url.split('/')[:-1]) + '/' + media_el.attrib.get('href') + media_url = media_el.attrib.get('href') or media_el.attrib.get('url') + if not media_url: + continue + manifest_url = ( + media_url if media_url.startswith('http://') or media_url.startswith('https://') + else ('/'.join(manifest_url.split('/')[:-1]) + '/' + media_url)) + # If media_url is itself a f4m manifest do the recursive extraction + # since bitrates in parent manifest (this one) and media_url manifest + # may differ leading to inability to resolve the format by requested + # bitrate in f4m downloader + if determine_ext(manifest_url) == 'f4m': + formats.extend(self._extract_f4m_formats(manifest_url, video_id, preference, f4m_id)) + continue tbr = int_or_none(media_el.attrib.get('bitrate')) - format_id = 'f4m-%d' % (i if tbr is None else tbr) formats.append({ - 'format_id': format_id, + 'format_id': '-'.join(filter(None, [f4m_id, compat_str(i if tbr is None else tbr)])), 'url': manifest_url, 'ext': 'flv', 'tbr': tbr, 'width': int_or_none(media_el.attrib.get('width')), 'height': int_or_none(media_el.attrib.get('height')), + 'preference': preference, }) self._sort_formats(formats) return formats def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None, - entry_protocol='m3u8', preference=None): + entry_protocol='m3u8', preference=None, + m3u8_id=None, note=None, errnote=None, + fatal=True): formats = [{ - 'format_id': 'm3u8-meta', + 'format_id': '-'.join(filter(None, [m3u8_id, 'meta'])), 'url': m3u8_url, 'ext': ext, 'protocol': 'm3u8', - 'preference': -1, + 'preference': preference - 1 if preference else -1, 'resolution': 'multiple', 'format_note': 'Quality selection URL', }] @@ -784,9 +902,13 @@ class InfoExtractor(object): m3u8_doc = self._download_webpage( m3u8_url, video_id, - note='Downloading m3u8 information', - errnote='Failed to download m3u8 information') + note=note or 'Downloading m3u8 information', + errnote=errnote or 'Failed to download m3u8 information', + fatal=fatal) + if m3u8_doc is False: + return m3u8_doc last_info = None + last_media = None kv_rex = re.compile( r'(?P[a-zA-Z_-]+)=(?P"[^"]+"|[^",]+)(?:,|$)') for line in m3u8_doc.splitlines(): @@ -797,6 +919,13 @@ class InfoExtractor(object): if v.startswith('"'): v = v[1:-1] last_info[m.group('key')] = v + elif line.startswith('#EXT-X-MEDIA:'): + last_media = {} + for m in kv_rex.finditer(line): + v = m.group('val') + if v.startswith('"'): + v = v[1:-1] + last_media[m.group('key')] = v elif line.startswith('#') or not line.strip(): continue else: @@ -804,9 +933,13 @@ class InfoExtractor(object): formats.append({'url': format_url(line)}) continue tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000) - + format_id = [] + if m3u8_id: + format_id.append(m3u8_id) + last_media_name = last_media.get('NAME') if last_media and last_media.get('TYPE') != 'SUBTITLES' else None + format_id.append(last_media_name if last_media_name else '%d' % (tbr if tbr else len(formats))) f = { - 'format_id': 'm3u8-%d' % (tbr if tbr else len(formats)), + 'format_id': '-'.join(format_id), 'url': format_url(line.strip()), 'tbr': tbr, 'ext': ext, @@ -826,54 +959,78 @@ class InfoExtractor(object): width_str, height_str = resolution.split('x') f['width'] = int(width_str) f['height'] = int(height_str) + if last_media is not None: + f['m3u8_media'] = last_media + last_media = None formats.append(f) last_info = {} self._sort_formats(formats) return formats # TODO: improve extraction - def _extract_smil_formats(self, smil_url, video_id): + def _extract_smil_formats(self, smil_url, video_id, fatal=True): smil = self._download_xml( smil_url, video_id, 'Downloading SMIL file', - 'Unable to download SMIL file') + 'Unable to download SMIL file', fatal=fatal) + if smil is False: + assert not fatal + return [] base = smil.find('./head/meta').get('base') formats = [] rtmp_count = 0 - for video in smil.findall('./body/switch/video'): - src = video.get('src') - if not src: - continue - bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000) - width = int_or_none(video.get('width')) - height = int_or_none(video.get('height')) - proto = video.get('proto') - if not proto: - if base: - if base.startswith('rtmp'): - proto = 'rtmp' - elif base.startswith('http'): - proto = 'http' - ext = video.get('ext') - if proto == 'm3u8': - formats.extend(self._extract_m3u8_formats(src, video_id, ext)) - elif proto == 'rtmp': - rtmp_count += 1 - streamer = video.get('streamer') or base - formats.append({ - 'url': streamer, - 'play_path': src, - 'ext': 'flv', - 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate), - 'tbr': bitrate, - 'width': width, - 'height': height, - }) + if smil.findall('./body/seq/video'): + video = smil.findall('./body/seq/video')[0] + fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count) + formats.extend(fmts) + else: + for video in smil.findall('./body/switch/video'): + fmts, rtmp_count = self._parse_smil_video(video, video_id, base, rtmp_count) + formats.extend(fmts) + self._sort_formats(formats) return formats + def _parse_smil_video(self, video, video_id, base, rtmp_count): + src = video.get('src') + if not src: + return [], rtmp_count + bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000) + width = int_or_none(video.get('width')) + height = int_or_none(video.get('height')) + proto = video.get('proto') + if not proto: + if base: + if base.startswith('rtmp'): + proto = 'rtmp' + elif base.startswith('http'): + proto = 'http' + ext = video.get('ext') + if proto == 'm3u8': + return self._extract_m3u8_formats(src, video_id, ext), rtmp_count + elif proto == 'rtmp': + rtmp_count += 1 + streamer = video.get('streamer') or base + return ([{ + 'url': streamer, + 'play_path': src, + 'ext': 'flv', + 'format_id': 'rtmp-%d' % (rtmp_count if bitrate is None else bitrate), + 'tbr': bitrate, + 'width': width, + 'height': height, + }], rtmp_count) + elif proto.startswith('http'): + return ([{ + 'url': base + src, + 'ext': ext or 'flv', + 'tbr': bitrate, + 'width': width, + 'height': height, + }], rtmp_count) + def _live_title(self, name): """ Generate the title for a live video """ now = datetime.datetime.now() @@ -937,6 +1094,24 @@ class InfoExtractor(object): any_restricted = any_restricted or is_restricted return not any_restricted + def extract_subtitles(self, *args, **kwargs): + if (self._downloader.params.get('writesubtitles', False) or + self._downloader.params.get('listsubtitles')): + return self._get_subtitles(*args, **kwargs) + return {} + + def _get_subtitles(self, *args, **kwargs): + raise NotImplementedError("This method must be implemented by subclasses") + + def extract_automatic_captions(self, *args, **kwargs): + if (self._downloader.params.get('writeautomaticsub', False) or + self._downloader.params.get('listsubtitles')): + return self._get_automatic_captions(*args, **kwargs) + return {} + + def _get_automatic_captions(self, *args, **kwargs): + raise NotImplementedError("This method must be implemented by subclasses") + class SearchInfoExtractor(InfoExtractor): """