X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/099764c0c91f4ad7db03d9347798f8619383ea7e..f8df414a4abcde0ddd39325dac26ca071d2d15c6:/youtube_dl/extractor/niconico.py?ds=inline diff --git a/youtube_dl/extractor/niconico.py b/youtube_dl/extractor/niconico.py index 3b5784e..026329d 100644 --- a/youtube_dl/extractor/niconico.py +++ b/youtube_dl/extractor/niconico.py @@ -1,18 +1,27 @@ -# encoding: utf-8 +# coding: utf-8 from __future__ import unicode_literals -import re import json +import datetime from .common import InfoExtractor -from ..utils import ( - compat_urllib_parse, - compat_urllib_request, +from ..compat import ( + compat_parse_qs, compat_urlparse, - unified_strdate, - parse_duration, - int_or_none, +) +from ..utils import ( + determine_ext, + dict_get, ExtractorError, + int_or_none, + float_or_none, + parse_duration, + parse_iso8601, + remove_start, + try_get, + unified_timestamp, + urlencode_postdata, + xpath_text, ) @@ -20,29 +29,116 @@ class NiconicoIE(InfoExtractor): IE_NAME = 'niconico' IE_DESC = 'ニコニコ動画' - _TEST = { + _TESTS = [{ 'url': 'http://www.nicovideo.jp/watch/sm22312215', 'md5': 'd1a75c0823e2f629128c43e1212760f9', 'info_dict': { 'id': 'sm22312215', 'ext': 'mp4', 'title': 'Big Buck Bunny', + 'thumbnail': r're:https?://.*', 'uploader': 'takuya0301', 'uploader_id': '2698420', 'upload_date': '20131123', + 'timestamp': 1385182762, 'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org', 'duration': 33, + 'view_count': int, + 'comment_count': int, }, - 'params': { - 'username': 'ydl.niconico@gmail.com', - 'password': 'youtube-dl', + 'skip': 'Requires an account', + }, { + # File downloaded with and without credentials are different, so omit + # the md5 field + 'url': 'http://www.nicovideo.jp/watch/nm14296458', + 'info_dict': { + 'id': 'nm14296458', + 'ext': 'swf', + 'title': '【鏡音リン】Dance on media【オリジナル】take2!', + 'description': 'md5:689f066d74610b3b22e0f1739add0f58', + 'thumbnail': r're:https?://.*', + 'uploader': 'りょうた', + 'uploader_id': '18822557', + 'upload_date': '20110429', + 'timestamp': 1304065916, + 'duration': 209, }, - } + 'skip': 'Requires an account', + }, { + # 'video exists but is marked as "deleted" + # md5 is unstable + 'url': 'http://www.nicovideo.jp/watch/sm10000', + 'info_dict': { + 'id': 'sm10000', + 'ext': 'unknown_video', + 'description': 'deleted', + 'title': 'ドラえもんエターナル第3話「決戦第3新東京市」＜前編＞', + 'thumbnail': r're:https?://.*', + 'upload_date': '20071224', + 'timestamp': int, # timestamp field has different value if logged in + 'duration': 304, + 'view_count': int, + }, + 'skip': 'Requires an account', + }, { + 'url': 'http://www.nicovideo.jp/watch/so22543406', + 'info_dict': { + 'id': '1388129933', + 'ext': 'mp4', + 'title': '【第1回】RADIOアニメロミックス ラブライブ！～のぞえりRadio Garden～', + 'description': 'md5:b27d224bb0ff53d3c8269e9f8b561cf1', + 'thumbnail': r're:https?://.*', + 'timestamp': 1388851200, + 'upload_date': '20140104', + 'uploader': 'アニメロチャンネル', + 'uploader_id': '312', + }, + 'skip': 'The viewing period of the video you were searching for has expired.', + }, { + # video not available via `getflv`; "old" HTML5 video + 'url': 'http://www.nicovideo.jp/watch/sm1151009', + 'md5': '8fa81c364eb619d4085354eab075598a', + 'info_dict': { + 'id': 'sm1151009', + 'ext': 'mp4', + 'title': 'マスターシステム本体内蔵のスペハリのメインテーマ（ＰＳＧ版）', + 'description': 'md5:6ee077e0581ff5019773e2e714cdd0b7', + 'thumbnail': r're:https?://.*', + 'duration': 184, + 'timestamp': 1190868283, + 'upload_date': '20070927', + 'uploader': 'denden2', + 'uploader_id': '1392194', + 'view_count': int, + 'comment_count': int, + }, + 'skip': 'Requires an account', + }, { + # "New" HTML5 video + 'url': 'http://www.nicovideo.jp/watch/sm31464864', + 'md5': '351647b4917660986dc0fa8864085135', + 'info_dict': { + 'id': 'sm31464864', + 'ext': 'mp4', + 'title': '新作TVアニメ「戦姫絶唱シンフォギアAXZ」PV 最高画質', + 'description': 'md5:e52974af9a96e739196b2c1ca72b5feb', + 'timestamp': 1498514060, + 'upload_date': '20170626', + 'uploader': 'ゲス', + 'uploader_id': '40826363', + 'thumbnail': r're:https?://.*', + 'duration': 198, + 'view_count': int, + 'comment_count': int, + }, + 'skip': 'Requires an account', + }, { + 'url': 'http://sp.nicovideo.jp/watch/sm28964488?ss_pos=1&cp_in=wt_tg', + 'only_matching': True, + }] - _VALID_URL = r'https?://(?:www\.|secure\.)?nicovideo\.jp/watch/((?:[a-z]{2})?[0-9]+)' + _VALID_URL = r'https?://(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch/(?P(?:[a-z]{2})?[0-9]+)' _NETRC_MACHINE = 'niconico' - # Determine whether the downloader used authentication to download video - _AUTHENTICATED = False def _real_initialize(self): self._login() @@ -54,97 +150,263 @@ class NiconicoIE(InfoExtractor): return True # Log in + login_ok = True login_form_strs = { - 'mail': username, + 'mail_tel': username, 'password': password, } - # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode - # chokes on unicode - login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items()) - login_data = compat_urllib_parse.urlencode(login_form).encode('utf-8') - request = compat_urllib_request.Request( - 'https://secure.nicovideo.jp/secure/login', login_data) - login_results = self._download_webpage( - request, None, note='Logging in', errnote='Unable to log in') - if re.search(r'(?i)

Log in error

', login_results) is not None: + urlh = self._request_webpage( + 'https://account.nicovideo.jp/api/v1/login', None, + note='Logging in', errnote='Unable to log in', + data=urlencode_postdata(login_form_strs)) + if urlh is False: + login_ok = False + else: + parts = compat_urlparse.urlparse(urlh.geturl()) + if compat_parse_qs(parts.query).get('message', [None])[0] == 'cant_login': + login_ok = False + if not login_ok: self._downloader.report_warning('unable to log in: bad username or password') - return False - # Successful login - self._AUTHENTICATED = True - return True + return login_ok + + def _extract_format_for_quality(self, api_data, video_id, audio_quality, video_quality): + def yesno(boolean): + return 'yes' if boolean else 'no' + + session_api_data = api_data['video']['dmcInfo']['session_api'] + session_api_endpoint = session_api_data['urls'][0] + + format_id = '-'.join(map(lambda s: remove_start(s['id'], 'archive_'), [video_quality, audio_quality])) + + session_response = self._download_json( + session_api_endpoint['url'], video_id, + query={'_format': 'json'}, + headers={'Content-Type': 'application/json'}, + note='Downloading JSON metadata for %s' % format_id, + data=json.dumps({ + 'session': { + 'client_info': { + 'player_id': session_api_data['player_id'], + }, + 'content_auth': { + 'auth_type': session_api_data['auth_types'][session_api_data['protocols'][0]], + 'content_key_timeout': session_api_data['content_key_timeout'], + 'service_id': 'nicovideo', + 'service_user_id': session_api_data['service_user_id'] + }, + 'content_id': session_api_data['content_id'], + 'content_src_id_sets': [{ + 'content_src_ids': [{ + 'src_id_to_mux': { + 'audio_src_ids': [audio_quality['id']], + 'video_src_ids': [video_quality['id']], + } + }] + }], + 'content_type': 'movie', + 'content_uri': '', + 'keep_method': { + 'heartbeat': { + 'lifetime': session_api_data['heartbeat_lifetime'] + } + }, + 'priority': session_api_data['priority'], + 'protocol': { + 'name': 'http', + 'parameters': { + 'http_parameters': { + 'parameters': { + 'http_output_download_parameters': { + 'use_ssl': yesno(session_api_endpoint['is_ssl']), + 'use_well_known_port': yesno(session_api_endpoint['is_well_known_port']), + } + } + } + } + }, + 'recipe_id': session_api_data['recipe_id'], + 'session_operation_auth': { + 'session_operation_auth_by_signature': { + 'signature': session_api_data['signature'], + 'token': session_api_data['token'], + } + }, + 'timing_constraint': 'unlimited' + } + })) + + resolution = video_quality.get('resolution', {}) + + return { + 'url': session_response['data']['session']['content_uri'], + 'format_id': format_id, + 'ext': 'mp4', # Session API are used in HTML5, which always serves mp4 + 'abr': float_or_none(audio_quality.get('bitrate'), 1000), + 'vbr': float_or_none(video_quality.get('bitrate'), 1000), + 'height': resolution.get('height'), + 'width': resolution.get('width'), + } def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - video_id = mobj.group(1) + video_id = self._match_id(url) + + # Get video webpage. We are not actually interested in it for normal + # cases, but need the cookies in order to be able to download the + # info webpage + webpage, handle = self._download_webpage_handle( + 'http://www.nicovideo.jp/watch/' + video_id, video_id) + if video_id.startswith('so'): + video_id = self._match_id(handle.geturl()) - # Get video webpage. We are not actually interested in it, but need - # the cookies in order to be able to download the info webpage - self._download_webpage('http://www.nicovideo.jp/watch/' + video_id, video_id) + api_data = self._parse_json(self._html_search_regex( + 'data-api-data="([^"]+)"', webpage, + 'API data', default='{}'), video_id) - video_info = self._download_xml( - 'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id, video_id, - note='Downloading video info page') + def _format_id_from_url(video_url): + return 'economy' if video_real_url.endswith('low') else 'normal' - if self._AUTHENTICATED: + try: + video_real_url = api_data['video']['smileInfo']['url'] + except KeyError: # Flash videos # Get flv info flv_info_webpage = self._download_webpage( - 'http://flapi.nicovideo.jp/api/getflv?v=' + video_id, + 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1', video_id, 'Downloading flv info') + + flv_info = compat_urlparse.parse_qs(flv_info_webpage) + if 'url' not in flv_info: + if 'deleted' in flv_info: + raise ExtractorError('The video has been deleted.', + expected=True) + elif 'closed' in flv_info: + raise ExtractorError('Niconico videos now require logging in', + expected=True) + elif 'error' in flv_info: + raise ExtractorError('%s reports error: %s' % ( + self.IE_NAME, flv_info['error'][0]), expected=True) + else: + raise ExtractorError('Unable to find video URL') + + video_info_xml = self._download_xml( + 'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id, + video_id, note='Downloading video info page') + + def get_video_info(items): + if not isinstance(items, list): + items = [items] + for item in items: + ret = xpath_text(video_info_xml, './/' + item) + if ret: + return ret + + video_real_url = flv_info['url'][0] + + extension = get_video_info('movie_type') + if not extension: + extension = determine_ext(video_real_url) + + formats = [{ + 'url': video_real_url, + 'ext': extension, + 'format_id': _format_id_from_url(video_real_url), + }] else: - # Get external player info - ext_player_info = self._download_webpage( - 'http://ext.nicovideo.jp/thumb_watch/' + video_id, video_id) - thumb_play_key = self._search_regex( - r'\'thumbPlayKey\'\s*:\s*\'(.*?)\'', ext_player_info, 'thumbPlayKey') + formats = [] - # Get flv info - flv_info_data = compat_urllib_parse.urlencode({ - 'k': thumb_play_key, - 'v': video_id - }) - flv_info_request = compat_urllib_request.Request( - 'http://ext.nicovideo.jp/thumb_watch', flv_info_data, - {'Content-Type': 'application/x-www-form-urlencoded'}) - flv_info_webpage = self._download_webpage( - flv_info_request, video_id, - note='Downloading flv info', errnote='Unable to download flv info') + dmc_info = api_data['video'].get('dmcInfo') + if dmc_info: # "New" HTML5 videos + quality_info = dmc_info['quality'] + for audio_quality in quality_info['audios']: + for video_quality in quality_info['videos']: + if not audio_quality['available'] or not video_quality['available']: + continue + formats.append(self._extract_format_for_quality( + api_data, video_id, audio_quality, video_quality)) + + self._sort_formats(formats) + else: # "Old" HTML5 videos + formats = [{ + 'url': video_real_url, + 'ext': 'mp4', + 'format_id': _format_id_from_url(video_real_url), + }] - if 'deleted=' in flv_info_webpage: - raise ExtractorError('The video has been deleted.', - expected=True) - video_real_url = compat_urlparse.parse_qs(flv_info_webpage)['url'][0] + def get_video_info(items): + return dict_get(api_data['video'], items) # Start extracting information - title = video_info.find('.//title').text - extension = video_info.find('.//movie_type').text - video_format = extension.upper() - thumbnail = video_info.find('.//thumbnail_url').text - description = video_info.find('.//description').text - upload_date = unified_strdate(video_info.find('.//first_retrieve').text.split('+')[0]) - view_count = int_or_none(video_info.find('.//view_counter').text) - comment_count = int_or_none(video_info.find('.//comment_num').text) - duration = parse_duration(video_info.find('.//length').text) - webpage_url = video_info.find('.//watch_url').text - - if video_info.find('.//ch_id') is not None: - uploader_id = video_info.find('.//ch_id').text - uploader = video_info.find('.//ch_name').text - elif video_info.find('.//user_id') is not None: - uploader_id = video_info.find('.//user_id').text - uploader = video_info.find('.//user_nickname').text - else: - uploader_id = uploader = None + title = get_video_info('title') + if not title: + title = self._og_search_title(webpage, default=None) + if not title: + title = self._html_search_regex( + r']+class="videoHeaderTitle"[^>]*>([^<]+)', + webpage, 'video title') + + watch_api_data_string = self._html_search_regex( + r']+id="watchAPIDataContainer"[^>]+>([^<]+)', + webpage, 'watch api data', default=None) + watch_api_data = self._parse_json(watch_api_data_string, video_id) if watch_api_data_string else {} + video_detail = watch_api_data.get('videoDetail', {}) + + thumbnail = ( + get_video_info(['thumbnail_url', 'thumbnailURL']) or + self._html_search_meta('image', webpage, 'thumbnail', default=None) or + video_detail.get('thumbnail')) + + description = get_video_info('description') + + timestamp = (parse_iso8601(get_video_info('first_retrieve')) or + unified_timestamp(get_video_info('postedDateTime'))) + if not timestamp: + match = self._html_search_meta('datePublished', webpage, 'date published', default=None) + if match: + timestamp = parse_iso8601(match.replace('+', ':00+')) + if not timestamp and video_detail.get('postedAt'): + timestamp = parse_iso8601( + video_detail['postedAt'].replace('/', '-'), + delimiter=' ', timezone=datetime.timedelta(hours=9)) + + view_count = int_or_none(get_video_info(['view_counter', 'viewCount'])) + if not view_count: + match = self._html_search_regex( + r'>Views: ]*>([^<]+)', + webpage, 'view count', default=None) + if match: + view_count = int_or_none(match.replace(',', '')) + view_count = view_count or video_detail.get('viewCount') + + comment_count = (int_or_none(get_video_info('comment_num')) or + video_detail.get('commentCount') or + try_get(api_data, lambda x: x['thread']['commentCount'])) + if not comment_count: + match = self._html_search_regex( + r'>Comments: ]*>([^<]+)', + webpage, 'comment count', default=None) + if match: + comment_count = int_or_none(match.replace(',', '')) + + duration = (parse_duration( + get_video_info('length') or + self._html_search_meta( + 'video:duration', webpage, 'video duration', default=None)) or + video_detail.get('length') or + get_video_info('duration')) + + webpage_url = get_video_info('watch_url') or url + + owner = api_data.get('owner', {}) + uploader_id = get_video_info(['ch_id', 'user_id']) or owner.get('id') + uploader = get_video_info(['ch_name', 'user_nickname']) or owner.get('nickname') return { 'id': video_id, - 'url': video_real_url, 'title': title, - 'ext': extension, - 'format': video_format, + 'formats': formats, 'thumbnail': thumbnail, 'description': description, 'uploader': uploader, - 'upload_date': upload_date, + 'timestamp': timestamp, 'uploader_id': uploader_id, 'view_count': view_count, 'comment_count': comment_count, @@ -154,7 +416,7 @@ class NiconicoIE(InfoExtractor): class NiconicoPlaylistIE(InfoExtractor): - _VALID_URL = r'https?://www\.nicovideo\.jp/mylist/(?P\d+)' + _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/mylist/(?P\d+)' _TEST = { 'url': 'http://www.nicovideo.jp/mylist/27411728', @@ -170,13 +432,13 @@ class NiconicoPlaylistIE(InfoExtractor): webpage = self._download_webpage(url, list_id) entries_json = self._search_regex(r'Mylist\.preload\(\d+, (\[.*\])\);', - webpage, 'entries') + webpage, 'entries') entries = json.loads(entries_json) entries = [{ '_type': 'url', 'ie_key': NiconicoIE.ie_key(), 'url': ('http://www.nicovideo.jp/watch/%s' % - entry['item_data']['video_id']), + entry['item_data']['video_id']), } for entry in entries] return {