X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/46113edab215c2211a604c06245c16d5d4e57dcf..c64c42cb1f9487398f8b91a51266ff6e79115b90:/youtube_dl/extractor/niconico.py diff --git a/youtube_dl/extractor/niconico.py b/youtube_dl/extractor/niconico.py index 62d5707..dd75a48 100644 --- a/youtube_dl/extractor/niconico.py +++ b/youtube_dl/extractor/niconico.py @@ -3,15 +3,22 @@ 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_urllib_parse_urlencode, compat_urlparse, - unified_strdate, - parse_duration, +) +from ..utils import ( + ExtractorError, int_or_none, + parse_duration, + parse_iso8601, + sanitized_Request, + xpath_text, + determine_ext, + urlencode_postdata, ) @@ -19,7 +26,7 @@ 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': { @@ -29,16 +36,53 @@ class NiconicoIE(InfoExtractor): 'uploader': 'takuya0301', 'uploader_id': '2698420', 'upload_date': '20131123', + 'timestamp': 1385182762, 'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org', 'duration': 33, }, - 'params': { - 'username': 'ydl.niconico@gmail.com', - 'password': 'youtube-dl', + }, { + # 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', + 'uploader': 'りょうた', + 'uploader_id': '18822557', + 'upload_date': '20110429', + 'timestamp': 1304065916, + 'duration': 209, }, - } + }, { + # '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＞', + 'upload_date': '20071224', + 'timestamp': 1198527840, # timestamp field has different value if logged in + 'duration': 304, + }, + }, { + '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', + 'timestamp': 1388851200, + 'upload_date': '20140104', + 'uploader': 'アニメロチャンネル', + 'uploader_id': '312', + } + }] - _VALID_URL = r'https?://(?:www\.|secure\.)?nicovideo\.jp/watch/((?:[a-z]{2})?[0-9]+)' + _VALID_URL = r'https?://(?:www\.|secure\.)?nicovideo\.jp/watch/(?P(?:[a-z]{2})?[0-9]+)' _NETRC_MACHINE = 'niconico' # Determine whether the downloader used authentication to download video _AUTHENTICATED = False @@ -57,11 +101,8 @@ class NiconicoIE(InfoExtractor): 'mail': 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( + login_data = urlencode_postdata(login_form_strs) + request = sanitized_Request( 'https://secure.nicovideo.jp/secure/login', login_data) login_results = self._download_webpage( request, None, note='Logging in', errnote='Unable to log in') @@ -73,12 +114,15 @@ class NiconicoIE(InfoExtractor): return True 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, 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) + # 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()) video_info = self._download_xml( 'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id, video_id, @@ -87,7 +131,7 @@ class NiconicoIE(InfoExtractor): if self._AUTHENTICATED: # 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') else: # Get external player info @@ -97,30 +141,88 @@ class NiconicoIE(InfoExtractor): r'\'thumbPlayKey\'\s*:\s*\'(.*?)\'', ext_player_info, 'thumbPlayKey') # Get flv info - flv_info_data = compat_urllib_parse.urlencode({ + flv_info_data = compat_urllib_parse_urlencode({ 'k': thumb_play_key, 'v': video_id }) - flv_info_request = compat_urllib_request.Request( + flv_info_request = sanitized_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') - video_real_url = compat_urlparse.parse_qs(flv_info_webpage)['url'][0] + 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) + else: + raise ExtractorError('Unable to find video URL') + + video_real_url = flv_info['url'][0] # 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 + title = xpath_text(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', {}) + + extension = xpath_text(video_info, './/movie_type') + if not extension: + extension = determine_ext(video_real_url) + + thumbnail = ( + xpath_text(video_info, './/thumbnail_url') or + self._html_search_meta('image', webpage, 'thumbnail', default=None) or + video_detail.get('thumbnail')) + + description = xpath_text(video_info, './/description') + + timestamp = parse_iso8601(xpath_text(video_info, './/first_retrieve')) + 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(xpath_text(video_info, './/view_counter')) + 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(xpath_text(video_info, './/comment_num')) + 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(',', '')) + comment_count = comment_count or video_detail.get('commentCount') + + duration = (parse_duration( + xpath_text(video_info, './/length') or + self._html_search_meta( + 'video:duration', webpage, 'video duration', default=None)) or + video_detail.get('length')) + + webpage_url = xpath_text(video_info, './/watch_url') or url if video_info.find('.//ch_id') is not None: uploader_id = video_info.find('.//ch_id').text @@ -136,11 +238,11 @@ class NiconicoIE(InfoExtractor): 'url': video_real_url, 'title': title, 'ext': extension, - 'format': video_format, + 'format_id': 'economy' if video_real_url.endswith('low') else 'normal', 'thumbnail': thumbnail, 'description': description, 'uploader': uploader, - 'upload_date': upload_date, + 'timestamp': timestamp, 'uploader_id': uploader_id, 'view_count': view_count, 'comment_count': comment_count, @@ -166,12 +268,13 @@ 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_id'], + 'url': ('http://www.nicovideo.jp/watch/%s' % + entry['item_data']['video_id']), } for entry in entries] return {