d/p/remove-autoupdate-mechanism.patch: Update patch with metadata at the top.

[youtubedl] / youtube_dl / extractor / niconico.py

diff --git a/youtube_dl/extractor/niconico.py b/youtube_dl/extractor/niconico.py
index bda1cff056d35ffd9cb876a0cc2fc141a6ce2ba1..df7f528be2d4c8da7cfe826609608754f2f8e088 100644 (file)
--- a/youtube_dl/extractor/niconico.py
+++ b/youtube_dl/extractor/niconico.py
@@ -1,24 +1,27 @@
-# encoding: utf-8
+# coding: utf-8

 from __future__ import unicode_literals
 
 from __future__ import unicode_literals
 

-import re

 import json
 import datetime
 
 from .common import InfoExtractor
 from ..compat import (
 import json
 import datetime
 
 from .common import InfoExtractor
 from ..compat import (

-    compat_urllib_parse,
-    compat_urllib_request,
+    compat_parse_qs,

     compat_urlparse,
 )
 from ..utils import (
     compat_urlparse,
 )
 from ..utils import (

-    encode_dict,
+    determine_ext,
+    dict_get,

     ExtractorError,
     int_or_none,
     ExtractorError,
     int_or_none,

+    float_or_none,

     parse_duration,
     parse_iso8601,
     parse_duration,
     parse_iso8601,

+    remove_start,
+    try_get,
+    unified_timestamp,
+    urlencode_postdata,

     xpath_text,
     xpath_text,

-    determine_ext,

 )
 
 
 )
 
 


@@ -33,13 +36,17 @@ class NiconicoIE(InfoExtractor):
             'id': 'sm22312215',
             'ext': 'mp4',
             'title': 'Big Buck Bunny',
             'id': 'sm22312215',
             'ext': 'mp4',
             'title': 'Big Buck Bunny',

+            'thumbnail': r're:https?://.*',

             'uploader': 'takuya0301',
             'uploader_id': '2698420',
             'upload_date': '20131123',
             'uploader': 'takuya0301',
             'uploader_id': '2698420',
             'upload_date': '20131123',

-            'timestamp': 1385182762,
+            'timestamp': int,  # timestamp is unstable

             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
             'duration': 33,
             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
             'duration': 33,

+            'view_count': int,
+            'comment_count': int,

         },
         },

+        'skip': 'Requires an account',

     }, {
         # File downloaded with and without credentials are different, so omit
         # the md5 field
     }, {
         # File downloaded with and without credentials are different, so omit
         # the md5 field


@@ -49,12 +56,14 @@ class NiconicoIE(InfoExtractor):
             'ext': 'swf',
             'title': '【鏡音リン】Dance on media【オリジナル】take2!',
             'description': 'md5:689f066d74610b3b22e0f1739add0f58',
             '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,
         },
             '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
     }, {
         # 'video exists but is marked as "deleted"
         # md5 is unstable


@@ -64,10 +73,13 @@ class NiconicoIE(InfoExtractor):
             '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＞',
             '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',
             'upload_date': '20071224',

-            'timestamp': 1198527840,  # timestamp field has different value if logged in
+            'timestamp': int,  # timestamp field has different value if logged in

             'duration': 304,
             'duration': 304,

+            'view_count': int,

         },
         },

+        'skip': 'Requires an account',

     }, {
         'url': 'http://www.nicovideo.jp/watch/so22543406',
         'info_dict': {
     }, {
         'url': 'http://www.nicovideo.jp/watch/so22543406',
         'info_dict': {


@@ -75,17 +87,77 @@ class NiconicoIE(InfoExtractor):
             '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',
             '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',
             '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
+        # md5 is unstable
+        'url': 'http://www.nicovideo.jp/watch/sm31464864',
+        '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': 'ゲスト',
+            'uploader_id': '40826363',
+            'thumbnail': r're:https?://.*',
+            'duration': 198,
+            'view_count': int,
+            'comment_count': int,
+        },
+        'skip': 'Requires an account',
+    }, {
+        # Video without owner
+        'url': 'http://www.nicovideo.jp/watch/sm18238488',
+        'md5': 'd265680a1f92bdcbbd2a507fc9e78a9e',
+        'info_dict': {
+            'id': 'sm18238488',
+            'ext': 'mp4',
+            'title': '【実写版】ミュータントタートルズ',
+            'description': 'md5:15df8988e47a86f9e978af2064bf6d8e',
+            'timestamp': 1341160408,
+            'upload_date': '20120701',
+            'uploader': None,
+            'uploader_id': None,
+            'thumbnail': r're:https?://.*',
+            'duration': 5271,
+            '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/(?P<id>(?:[a-z]{2})?[0-9]+)'
+    _VALID_URL = r'https?://(?:www\.|secure\.|sp\.)?nicovideo\.jp/watch/(?P<id>(?:[a-z]{2})?[0-9]+)'

     _NETRC_MACHINE = 'niconico'
     _NETRC_MACHINE = 'niconico'

-    # Determine whether the downloader used authentication to download video
-    _AUTHENTICATED = False

 
     def _real_initialize(self):
         self._login()
 
     def _real_initialize(self):
         self._login()


@@ -97,21 +169,102 @@ class NiconicoIE(InfoExtractor):
             return True
 
         # Log in
             return True
 
         # Log in

+        login_ok = True

         login_form_strs = {
         login_form_strs = {

-            'mail': username,
+            'mail_tel': username,

             'password': password,
         }
             'password': password,
         }

-        login_data = compat_urllib_parse.urlencode(encode_dict(login_form_strs)).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)<h1 class="mb8p4">Log in error</h1>', 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')
             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):
         video_id = self._match_id(url)
 
     def _real_extract(self, url):
         video_id = self._match_id(url)


@@ -124,46 +277,84 @@ class NiconicoIE(InfoExtractor):
         if video_id.startswith('so'):
             video_id = self._match_id(handle.geturl())
 
         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,
-            note='Downloading video info page')
+        api_data = self._parse_json(self._html_search_regex(
+            'data-api-data="([^"]+)"', webpage,
+            'API data', default='{}'), video_id)

 
 

-        if self._AUTHENTICATED:
+        def _format_id_from_url(video_url):
+            return 'economy' if video_real_url.endswith('low') else 'normal'
+
+        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/' + video_id + '?as3=1',
                 video_id, 'Downloading flv info')
             # Get flv info
             flv_info_webpage = self._download_webpage(
                 'http://flapi.nicovideo.jp/api/getflv/' + video_id + '?as3=1',
                 video_id, 'Downloading flv info')

-        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')

 
 

-            # 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')
-
-        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]
+            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:
+            formats = []
+
+            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),
+                }]
+
+            def get_video_info(items):
+                return dict_get(api_data['video'], items)

 
         # Start extracting information
 
         # Start extracting information

-        title = xpath_text(video_info, './/title')
+        title = get_video_info('title')

         if not title:
             title = self._og_search_title(webpage, default=None)
         if not title:
         if not title:
             title = self._og_search_title(webpage, default=None)
         if not title:


@@ -177,18 +368,15 @@ class NiconicoIE(InfoExtractor):
         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', {})
 
         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 = (
         thumbnail = (

-            xpath_text(video_info, './/thumbnail_url') or
+            get_video_info(['thumbnail_url', 'thumbnailURL']) or

             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
             video_detail.get('thumbnail'))
 
             self._html_search_meta('image', webpage, 'thumbnail', default=None) or
             video_detail.get('thumbnail'))
 

-        description = xpath_text(video_info, './/description')
+        description = get_video_info('description')

 
 

-        timestamp = parse_iso8601(xpath_text(video_info, './/first_retrieve'))
+        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:
         if not timestamp:
             match = self._html_search_meta('datePublished', webpage, 'date published', default=None)
             if match:


@@ -198,7 +386,7 @@ class NiconicoIE(InfoExtractor):
                 video_detail['postedAt'].replace('/', '-'),
                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 
                 video_detail['postedAt'].replace('/', '-'),
                 delimiter=' ', timezone=datetime.timedelta(hours=9))
 

-        view_count = int_or_none(xpath_text(video_info, './/view_counter'))
+        view_count = int_or_none(get_video_info(['view_counter', 'viewCount']))

         if not view_count:
             match = self._html_search_regex(
                 r'>Views: <strong[^>]*>([^<]+)</strong>',
         if not view_count:
             match = self._html_search_regex(
                 r'>Views: <strong[^>]*>([^<]+)</strong>',


@@ -207,38 +395,35 @@ class NiconicoIE(InfoExtractor):
                 view_count = int_or_none(match.replace(',', ''))
         view_count = view_count or video_detail.get('viewCount')
 
                 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'))
+        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: <strong[^>]*>([^<]+)</strong>',
                 webpage, 'comment count', default=None)
             if match:
                 comment_count = int_or_none(match.replace(',', ''))
         if not comment_count:
             match = self._html_search_regex(
                 r'>Comments: <strong[^>]*>([^<]+)</strong>',
                 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(
 
         duration = (parse_duration(

-            xpath_text(video_info, './/length') or
+            get_video_info('length') or

             self._html_search_meta(
                 'video:duration', webpage, 'video duration', default=None)) or
             self._html_search_meta(
                 'video:duration', webpage, 'video duration', default=None)) or

-            video_detail.get('length'))
+            video_detail.get('length') or
+            get_video_info('duration'))

 
 

-        webpage_url = xpath_text(video_info, './/watch_url') or url
+        webpage_url = get_video_info('watch_url') or url

 
 

-        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
+        # Note: cannot use api_data.get('owner', {}) because owner may be set to "null"
+        # in the JSON, which will cause None to be returned instead of {}.
+        owner = try_get(api_data, lambda x: x.get('owner'), dict) or {}
+        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,
 
         return {
             'id': video_id,

-            'url': video_real_url,

             'title': title,
             'title': title,

-            'ext': extension,
-            'format_id': 'economy' if video_real_url.endswith('low') else 'normal',
+            'formats': formats,

             'thumbnail': thumbnail,
             'description': description,
             'uploader': uploader,
             'thumbnail': thumbnail,
             'description': description,
             'uploader': uploader,


@@ -252,7 +437,7 @@ class NiconicoIE(InfoExtractor):
 
 
 class NiconicoPlaylistIE(InfoExtractor):
 
 
 class NiconicoPlaylistIE(InfoExtractor):

-    _VALID_URL = r'https?://www\.nicovideo\.jp/mylist/(?P<id>\d+)'
+    _VALID_URL = r'https?://(?:www\.)?nicovideo\.jp/mylist/(?P<id>\d+)'

 
     _TEST = {
         'url': 'http://www.nicovideo.jp/mylist/27411728',
 
     _TEST = {
         'url': 'http://www.nicovideo.jp/mylist/27411728',