X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fcommon.py;h=c0002345870a4490cd3d5904d148b3d272813f28;hb=6ad0219556cefe60239027633193cc9f1dc9fb1d;hp=0891309dd506dab2f7c44c256fdd6203e846be80;hpb=37768f92422c2cf61a961dbaf54d61dabd364506;p=youtube-dl diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index 0891309dd..c00023458 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -662,35 +662,48 @@ class InfoExtractor(object): else: return res - def _get_login_info(self): + def _get_netrc_login_info(self, netrc_machine=None): + username = None + password = None + netrc_machine = netrc_machine or self._NETRC_MACHINE + + if self._downloader.params.get('usenetrc', False): + try: + info = netrc.netrc().authenticators(netrc_machine) + if info is not None: + username = info[0] + password = info[2] + else: + raise netrc.NetrcParseError( + 'No authenticators for %s' % netrc_machine) + except (IOError, netrc.NetrcParseError) as err: + self._downloader.report_warning( + 'parsing .netrc: %s' % error_to_compat_str(err)) + + return username, password + + def _get_login_info(self, username_option='username', password_option='password', netrc_machine=None): """ Get the login info as (username, password) - It will look in the netrc file using the _NETRC_MACHINE value + First look for the manually specified credentials using username_option + and password_option as keys in params dictionary. If no such credentials + available look in the netrc file using the netrc_machine or _NETRC_MACHINE + value. If there's no info available, return (None, None) """ if self._downloader is None: return (None, None) - username = None - password = None downloader_params = self._downloader.params # Attempt to use provided username and password or .netrc data - if downloader_params.get('username') is not None: - username = downloader_params['username'] - password = downloader_params['password'] - elif downloader_params.get('usenetrc', False): - try: - info = netrc.netrc().authenticators(self._NETRC_MACHINE) - if info is not None: - username = info[0] - password = info[2] - else: - raise netrc.NetrcParseError('No authenticators for %s' % self._NETRC_MACHINE) - except (IOError, netrc.NetrcParseError) as err: - self._downloader.report_warning('parsing .netrc: %s' % error_to_compat_str(err)) + if downloader_params.get(username_option) is not None: + username = downloader_params[username_option] + password = downloader_params[password_option] + else: + username, password = self._get_netrc_login_info(netrc_machine) - return (username, password) + return username, password def _get_tfa_info(self, note='two-factor verification code'): """ @@ -816,11 +829,14 @@ class InfoExtractor(object): json_ld = self._search_regex( r'(?s)]+type=(["\'])application/ld\+json\1[^>]*>(?P.+?)', html, 'JSON-LD', group='json_ld', **kwargs) + default = kwargs.get('default', NO_DEFAULT) if not json_ld: - return {} - return self._json_ld( - json_ld, video_id, fatal=kwargs.get('fatal', True), - expected_type=expected_type) + return default if default is not NO_DEFAULT else {} + # JSON-LD may be malformed and thus `fatal` should be respected. + # At the same time `default` may be passed that assumes `fatal=False` + # for _search_regex. Let's simulate the same behavior here as well. + fatal = kwargs.get('fatal', True) if default == NO_DEFAULT else False + return self._json_ld(json_ld, video_id, fatal=fatal, expected_type=expected_type) def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None): if isinstance(json_ld, compat_str): @@ -875,16 +891,16 @@ class InfoExtractor(object): def _hidden_inputs(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): + for input in re.findall(r'(?i)(]+>)', html): + attrs = extract_attributes(input) + if not input: continue - name = re.search(r'(?:name|id)=(["\'])(?P.+?)\1', input) - if not name: + if attrs.get('type') not in ('hidden', 'submit'): continue - value = re.search(r'value=(["\'])(?P.*?)\1', input) - if not value: - continue - hidden_inputs[name.group('value')] = value.group('value') + name = attrs.get('name') or attrs.get('id') + value = attrs.get('value') + if name and value is not None: + hidden_inputs[name] = value return hidden_inputs def _form_hidden_inputs(self, form_id, html): @@ -1150,13 +1166,6 @@ class InfoExtractor(object): m3u8_id=None, note=None, errnote=None, fatal=True, live=False): - formats = [self._m3u8_meta_format(m3u8_url, ext, preference, m3u8_id)] - - format_url = lambda u: ( - u - if re.match(r'^https?://', u) - else compat_urlparse.urljoin(m3u8_url, u)) - res = self._download_webpage_handle( m3u8_url, video_id, note=note or 'Downloading m3u8 information', @@ -1167,6 +1176,13 @@ class InfoExtractor(object): m3u8_doc, urlh = res m3u8_url = urlh.geturl() + formats = [self._m3u8_meta_format(m3u8_url, ext, preference, m3u8_id)] + + format_url = lambda u: ( + u + if re.match(r'^https?://', u) + else compat_urlparse.urljoin(m3u8_url, u)) + # We should try extracting formats only from master playlists [1], i.e. # playlists that describe available qualities. On the other hand media # playlists [2] should be returned as is since they contain just the media @@ -1188,27 +1204,44 @@ class InfoExtractor(object): 'protocol': entry_protocol, 'preference': preference, }] - last_info = None - last_media = None + last_info = {} + last_media = {} for line in m3u8_doc.splitlines(): if line.startswith('#EXT-X-STREAM-INF:'): last_info = parse_m3u8_attributes(line) elif line.startswith('#EXT-X-MEDIA:'): - last_media = parse_m3u8_attributes(line) + media = parse_m3u8_attributes(line) + media_type = media.get('TYPE') + if media_type in ('VIDEO', 'AUDIO'): + media_url = media.get('URI') + if media_url: + format_id = [] + for v in (media.get('GROUP-ID'), media.get('NAME')): + if v: + format_id.append(v) + formats.append({ + 'format_id': '-'.join(format_id), + 'url': format_url(media_url), + 'language': media.get('LANGUAGE'), + 'vcodec': 'none' if media_type == 'AUDIO' else None, + 'ext': ext, + 'protocol': entry_protocol, + 'preference': preference, + }) + else: + # When there is no URI in EXT-X-MEDIA let this tag's + # data be used by regular URI lines below + last_media = media elif line.startswith('#') or not line.strip(): continue else: - if last_info is None: - formats.append({'url': format_url(line)}) - continue - tbr = int_or_none(last_info.get('BANDWIDTH'), scale=1000) + tbr = int_or_none(last_info.get('AVERAGE-BANDWIDTH') or 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') not in ('SUBTITLES', 'CLOSED-CAPTIONS') else None # Despite specification does not mention NAME attribute for # EXT-X-STREAM-INF it still sometimes may be present - stream_name = last_info.get('NAME') or last_media_name + stream_name = last_info.get('NAME') or last_media.get('NAME') # Bandwidth of live streams may differ over time thus making # format_id unpredictable. So it's better to keep provided # format_id intact. @@ -1239,11 +1272,9 @@ class InfoExtractor(object): 'abr': abr, }) f.update(parse_codecs(last_info.get('CODECS'))) - if last_media is not None: - f['m3u8_media'] = last_media - last_media = None formats.append(f) last_info = {} + last_media = {} return formats @staticmethod @@ -1682,7 +1713,7 @@ class InfoExtractor(object): self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type) return formats - def _parse_html5_media_entries(self, base_url, webpage): + def _parse_html5_media_entries(self, base_url, webpage, video_id, m3u8_id=None, m3u8_entry_protocol='m3u8'): def absolute_url(video_url): return compat_urlparse.urljoin(base_url, video_url) @@ -1697,6 +1728,21 @@ class InfoExtractor(object): return f return {} + def _media_formats(src, cur_media_type): + full_url = absolute_url(src) + if determine_ext(full_url) == 'm3u8': + is_plain_url = False + formats = self._extract_m3u8_formats( + full_url, video_id, ext='mp4', + entry_protocol=m3u8_entry_protocol, m3u8_id=m3u8_id) + else: + is_plain_url = True + formats = [{ + 'url': full_url, + 'vcodec': 'none' if cur_media_type == 'audio' else None, + }] + return is_plain_url, formats + entries = [] for media_tag, media_type, media_content in re.findall(r'(?s)(<(?Pvideo|audio)[^>]*>)(.*?)', webpage): media_info = { @@ -1706,10 +1752,8 @@ class InfoExtractor(object): media_attributes = extract_attributes(media_tag) src = media_attributes.get('src') if src: - media_info['formats'].append({ - 'url': absolute_url(src), - 'vcodec': 'none' if media_type == 'audio' else None, - }) + _, formats = _media_formats(src, media_type) + media_info['formats'].extend(formats) media_info['thumbnail'] = media_attributes.get('poster') if media_content: for source_tag in re.findall(r']+>', media_content): @@ -1717,12 +1761,13 @@ class InfoExtractor(object): src = source_attributes.get('src') if not src: continue - f = parse_content_type(source_attributes.get('type')) - f.update({ - 'url': absolute_url(src), - 'vcodec': 'none' if media_type == 'audio' else None, - }) - media_info['formats'].append(f) + is_plain_url, formats = _media_formats(src, media_type) + if is_plain_url: + f = parse_content_type(source_attributes.get('type')) + f.update(formats[0]) + media_info['formats'].append(f) + else: + media_info['formats'].extend(formats) for track_tag in re.findall(r']+>', media_content): track_attributes = extract_attributes(track_tag) kind = track_attributes.get('kind') @@ -1738,6 +1783,61 @@ class InfoExtractor(object): entries.append(media_info) return entries + def _extract_akamai_formats(self, manifest_url, video_id): + formats = [] + f4m_url = re.sub(r'(https?://.+?)/i/', r'\1/z/', manifest_url).replace('/master.m3u8', '/manifest.f4m') + formats.extend(self._extract_f4m_formats( + update_url_query(f4m_url, {'hdcore': '3.7.0'}), + video_id, f4m_id='hds', fatal=False)) + m3u8_url = re.sub(r'(https?://.+?)/z/', r'\1/i/', manifest_url).replace('/manifest.f4m', '/master.m3u8') + formats.extend(self._extract_m3u8_formats( + m3u8_url, video_id, 'mp4', 'm3u8_native', + m3u8_id='hls', fatal=False)) + return formats + + def _extract_wowza_formats(self, url, video_id, m3u8_entry_protocol='m3u8_native', skip_protocols=[]): + url = re.sub(r'/(?:manifest|playlist|jwplayer)\.(?:m3u8|f4m|mpd|smil)', '', url) + url_base = self._search_regex(r'(?:https?|rtmp|rtsp)(://[^?]+)', url, 'format url') + http_base_url = 'http' + url_base + formats = [] + if 'm3u8' not in skip_protocols: + formats.extend(self._extract_m3u8_formats( + http_base_url + '/playlist.m3u8', video_id, 'mp4', + m3u8_entry_protocol, m3u8_id='hls', fatal=False)) + if 'f4m' not in skip_protocols: + formats.extend(self._extract_f4m_formats( + http_base_url + '/manifest.f4m', + video_id, f4m_id='hds', fatal=False)) + if re.search(r'(?:/smil:|\.smil)', url_base): + if 'dash' not in skip_protocols: + formats.extend(self._extract_mpd_formats( + http_base_url + '/manifest.mpd', + video_id, mpd_id='dash', fatal=False)) + if 'smil' not in skip_protocols: + rtmp_formats = self._extract_smil_formats( + http_base_url + '/jwplayer.smil', + video_id, fatal=False) + for rtmp_format in rtmp_formats: + rtsp_format = rtmp_format.copy() + rtsp_format['url'] = '%s/%s' % (rtmp_format['url'], rtmp_format['play_path']) + del rtsp_format['play_path'] + del rtsp_format['ext'] + rtsp_format.update({ + 'url': rtsp_format['url'].replace('rtmp://', 'rtsp://'), + 'format_id': rtmp_format['format_id'].replace('rtmp', 'rtsp'), + 'protocol': 'rtsp', + }) + formats.extend([rtmp_format, rtsp_format]) + else: + for protocol in ('rtmp', 'rtsp'): + if protocol not in skip_protocols: + formats.append({ + 'url': protocol + url_base, + 'format_id': protocol, + 'protocol': protocol, + }) + return formats + def _live_title(self, name): """ Generate the title for a live video """ now = datetime.datetime.now()