X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fcommon.py;h=eb9bfa3d15a2c5084fbf67f05a401474ad2f881d;hb=5c2266df4b9aeb7881ed8c026a038e2a25e43734;hp=16ae4b98ffe09c97f604981bf6c2ce9dc1e44e03;hpb=11bed5827dace09b5483b159476ce9f8c29d6078;p=youtube-dl diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index 16ae4b98f..eb9bfa3d1 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -10,19 +10,18 @@ import re import socket import sys import time -import xml.etree.ElementTree from ..compat import ( compat_cookiejar, compat_cookies, - compat_HTTPError, + compat_getpass, compat_http_client, compat_urllib_error, compat_urllib_parse, compat_urllib_parse_urlparse, - compat_urllib_request, compat_urlparse, compat_str, + compat_etree_fromstring, ) from ..utils import ( NO_DEFAULT, @@ -37,7 +36,9 @@ from ..utils import ( int_or_none, RegexNotFoundError, sanitize_filename, + sanitized_Request, unescapeHTML, + unified_strdate, url_basename, xpath_text, xpath_with_ns, @@ -151,6 +152,7 @@ class InfoExtractor(object): description: Full video description. uploader: Full name of the video uploader. creator: The main artist who created the video. + release_date: The date (YYYYMMDD) when the video was released. timestamp: UNIX timestamp of the moment the video became available. upload_date: Video upload date (YYYYMMDD). If not explicitly set, calculated from timestamp. @@ -162,12 +164,14 @@ class InfoExtractor(object): with the "ext" entry and one of: * "data": The subtitles file contents * "url": A URL pointing to the subtitles file + "ext" will be calculated from URL if missing 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 + repost_count: Number of reposts 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 @@ -306,11 +310,11 @@ class InfoExtractor(object): @classmethod def ie_key(cls): """A string for getting the InfoExtractor with get_info_extractor""" - return cls.__name__[:-2] + return compat_str(cls.__name__[:-2]) @property def IE_NAME(self): - return type(self).__name__[:-2] + return compat_str(type(self).__name__[:-2]) def _request_webpage(self, url_or_request, video_id, note=None, errnote=None, fatal=True): """ Returns the response handle """ @@ -457,7 +461,7 @@ class InfoExtractor(object): return xml_string if transform_source: xml_string = transform_source(xml_string) - return xml.etree.ElementTree.fromstring(xml_string.encode('utf-8')) + return compat_etree_fromstring(xml_string.encode('utf-8')) def _download_json(self, url_or_request, video_id, note='Downloading JSON metadata', @@ -509,6 +513,18 @@ class InfoExtractor(object): """Report attempt to log in.""" self.to_screen('Logging in') + @staticmethod + def raise_login_required(msg='This video is only available for registered users'): + raise ExtractorError( + '%s. Use --username and --password or --netrc to provide account credentials.' % msg, + expected=True) + + @staticmethod + def raise_geo_restricted(msg='This video is not available from your location due to geo restriction'): + raise ExtractorError( + '%s. You might want to use --proxy to workaround.' % msg, + expected=True) + # Methods for following #608 @staticmethod def url_result(url, ie=None, video_id=None, video_title=None): @@ -610,7 +626,7 @@ class InfoExtractor(object): return (username, password) - def _get_tfa_info(self): + def _get_tfa_info(self, note='two-factor verification code'): """ Get the two-factor authentication info TODO - asking the user will be required for sms/phone verify @@ -624,13 +640,14 @@ class InfoExtractor(object): if downloader_params.get('twofactor', None) is not None: return downloader_params['twofactor'] - return None + return compat_getpass('Type %s and press [Return]: ' % note) # Helper functions for extracting OpenGraph info @staticmethod def _og_regexes(prop): - content_re = r'content=(?:"([^>]+?)"|\'([^>]+?)\')' - property_re = r'(?:name|property)=[\'"]og:%s[\'"]' % re.escape(prop) + content_re = r'content=(?:"([^"]+?)"|\'([^\']+?)\'|\s*([^\s"\'=<>`]+?))' + property_re = (r'(?:name|property)=(?:\'og:%(prop)s\'|"og:%(prop)s"|\s*og:%(prop)s\b)' + % {'prop': re.escape(prop)}) template = r']+?%s[^>]+?%s' return [ template % (property_re, content_re), @@ -724,20 +741,23 @@ class InfoExtractor(object): @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) - ]) + html = re.sub(r'', '', html) + hidden_inputs = {} + for input in re.findall(r'(?i)]+)>', html): + if not re.search(r'type=(["\'])(?:hidden|submit)\1', input): + continue + name = re.search(r'name=(["\'])(?P.+?)\1', input) + if not name: + continue + value = re.search(r'value=(["\'])(?P.*?)\1', input) + if not value: + continue + hidden_inputs[name.group('value')] = value.group('value') + return hidden_inputs def _form_hidden_inputs(self, form_id, html): form = self._search_regex( - r'(?s)]+?id=(["\'])%s\1[^>]*>(?P
.+?)
' % form_id, + r'(?is)]+?id=(["\'])%s\1[^>]*>(?P
.+?)
' % form_id, html, '%s form' % form_id, group='form') return self._hidden_inputs(form) @@ -821,7 +841,7 @@ class InfoExtractor(object): self._request_webpage(url, video_id, 'Checking %s URL' % item) return True except ExtractorError as e: - if isinstance(e.cause, compat_HTTPError): + if isinstance(e.cause, compat_urllib_error.URLError): self.to_screen( '%s: %s URL is invalid, skipping' % (video_id, item)) return False @@ -852,13 +872,18 @@ class InfoExtractor(object): time.sleep(timeout) def _extract_f4m_formats(self, manifest_url, video_id, preference=None, f4m_id=None, - transform_source=lambda s: fix_xml_ampersands(s).strip()): + transform_source=lambda s: fix_xml_ampersands(s).strip(), + fatal=True): manifest = self._download_xml( manifest_url, video_id, 'Downloading 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) + transform_source=transform_source, + fatal=fatal) + + if manifest is False: + return manifest formats = [] manifest_version = '1.0' @@ -866,6 +891,11 @@ class InfoExtractor(object): if not media_nodes: manifest_version = '2.0' media_nodes = manifest.findall('{http://ns.adobe.com/f4m/2.0}media') + base_url = xpath_text( + manifest, ['{http://ns.adobe.com/f4m/1.0}baseURL', '{http://ns.adobe.com/f4m/2.0}baseURL'], + 'base URL', default=None) + if base_url: + base_url = base_url.strip() for i, media_el in enumerate(media_nodes): if manifest_version == '2.0': media_url = media_el.attrib.get('href') or media_el.attrib.get('url') @@ -873,13 +903,16 @@ class InfoExtractor(object): continue manifest_url = ( media_url if media_url.startswith('http://') or media_url.startswith('https://') - else ('/'.join(manifest_url.split('/')[:-1]) + '/' + media_url)) + else ((base_url or '/'.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)) + f4m_formats = self._extract_f4m_formats( + manifest_url, video_id, preference, f4m_id, fatal=fatal) + if f4m_formats: + formats.extend(f4m_formats) continue tbr = int_or_none(media_el.attrib.get('bitrate')) formats.append({ @@ -915,13 +948,15 @@ class InfoExtractor(object): if re.match(r'^https?://', u) else compat_urlparse.urljoin(m3u8_url, u)) - m3u8_doc = self._download_webpage( + res = self._download_webpage_handle( m3u8_url, video_id, note=note or 'Downloading m3u8 information', errnote=errnote or 'Failed to download m3u8 information', fatal=fatal) - if m3u8_doc is False: - return m3u8_doc + if res is False: + return res + m3u8_doc, urlh = res + m3u8_url = urlh.geturl() last_info = None last_media = None kv_rex = re.compile( @@ -1027,6 +1062,7 @@ class InfoExtractor(object): video_id = os.path.splitext(url_basename(smil_url))[0] title = None description = None + upload_date = None for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): name = meta.attrib.get('name') content = meta.attrib.get('content') @@ -1036,11 +1072,22 @@ class InfoExtractor(object): title = content elif not description and name in ('description', 'abstract'): description = content + elif not upload_date and name == 'date': + upload_date = unified_strdate(content) + + thumbnails = [{ + 'id': image.get('type'), + 'url': image.get('src'), + 'width': int_or_none(image.get('width')), + 'height': int_or_none(image.get('height')), + } for image in smil.findall(self._xpath_ns('.//image', namespace)) if image.get('src')] return { 'id': video_id, 'title': title or video_id, 'description': description, + 'upload_date': upload_date, + 'thumbnails': thumbnails, 'formats': formats, 'subtitles': subtitles, } @@ -1049,7 +1096,7 @@ class InfoExtractor(object): return self._search_regex( r'(?i)^{([^}]+)?}smil$', smil.tag, 'namespace', default=None) - def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None): + def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None): base = smil_url for meta in smil.findall(self._xpath_ns('./head/meta', namespace)): b = meta.get('base') or meta.get('httpBase') @@ -1067,7 +1114,7 @@ class InfoExtractor(object): if not src: continue - bitrate = int_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000) + bitrate = float_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000) filesize = int_or_none(video.get('size') or video.get('fileSize')) width = int_or_none(video.get('width')) height = int_or_none(video.get('height')) @@ -1088,13 +1135,21 @@ class InfoExtractor(object): 'width': width, 'height': height, }) + if transform_rtmp_url: + streamer, src = transform_rtmp_url(streamer, src) + formats[-1].update({ + 'url': streamer, + 'play_path': src, + }) continue src_url = src if src.startswith('http') else compat_urlparse.urljoin(base, src) if proto == 'm3u8' or src_ext == 'm3u8': - formats.extend(self._extract_m3u8_formats( - src_url, video_id, ext or 'mp4', m3u8_id='hls')) + m3u8_formats = self._extract_m3u8_formats( + src_url, video_id, ext or 'mp4', m3u8_id='hls', fatal=False) + if m3u8_formats: + formats.extend(m3u8_formats) continue if src_ext == 'f4m': @@ -1106,10 +1161,12 @@ class InfoExtractor(object): } f4m_url += '&' if '?' in f4m_url else '?' f4m_url += compat_urllib_parse.urlencode(f4m_params) - formats.extend(self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds')) + f4m_formats = self._extract_f4m_formats(f4m_url, video_id, f4m_id='hds', fatal=False) + if f4m_formats: + formats.extend(f4m_formats) continue - if src_url.startswith('http'): + if src_url.startswith('http') and self._is_valid_url(src, video_id): http_count += 1 formats.append({ 'url': src_url, @@ -1126,7 +1183,7 @@ class InfoExtractor(object): return formats - def _parse_smil_subtitles(self, smil, namespace=None): + def _parse_smil_subtitles(self, smil, namespace=None, subtitles_lang='en'): subtitles = {} for num, textstream in enumerate(smil.findall(self._xpath_ns('.//textstream', namespace))): src = textstream.get('src') @@ -1135,9 +1192,14 @@ class InfoExtractor(object): ext = textstream.get('ext') or determine_ext(src) if not ext: type_ = textstream.get('type') - if type_ == 'text/srt': - ext = 'srt' - lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') + SUBTITLES_TYPES = { + 'text/vtt': 'vtt', + 'text/srt': 'srt', + 'application/smptett+xml': 'tt', + } + if type_ in SUBTITLES_TYPES: + ext = SUBTITLES_TYPES[type_] + lang = textstream.get('systemLanguage') or textstream.get('systemLanguageName') or textstream.get('lang') or subtitles_lang subtitles.setdefault(lang, []).append({ 'url': src, 'ext': ext, @@ -1223,7 +1285,7 @@ class InfoExtractor(object): def _get_cookies(self, url): """ Return a compat_cookies.SimpleCookie with the cookies for the url """ - req = compat_urllib_request.Request(url) + req = sanitized_Request(url) self._downloader.cookiejar.add_cookie_header(req) return compat_cookies.SimpleCookie(req.get_header('Cookie')) @@ -1265,6 +1327,23 @@ class InfoExtractor(object): def _get_subtitles(self, *args, **kwargs): raise NotImplementedError("This method must be implemented by subclasses") + @staticmethod + def _merge_subtitle_items(subtitle_list1, subtitle_list2): + """ Merge subtitle items for one language. Items with duplicated URLs + will be dropped. """ + list1_urls = set([item['url'] for item in subtitle_list1]) + ret = list(subtitle_list1) + ret.extend([item for item in subtitle_list2 if item['url'] not in list1_urls]) + return ret + + @classmethod + def _merge_subtitles(cls, subtitle_dict1, subtitle_dict2): + """ Merge two subtitle dictionaries, language by language. """ + ret = dict(subtitle_dict1) + for lang in subtitle_dict2: + ret[lang] = cls._merge_subtitle_items(subtitle_dict1.get(lang, []), subtitle_dict2[lang]) + return ret + def extract_automatic_captions(self, *args, **kwargs): if (self._downloader.params.get('writeautomaticsub', False) or self._downloader.params.get('listsubtitles')):