youtube_dl/extractor/niconico.py

   1 # encoding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import re
   5
   6 from .common import InfoExtractor
   7 from ..utils import (
   8     compat_urllib_parse,
   9     compat_urllib_request,
  10     compat_urlparse,
  11     unified_strdate,
  12     parse_duration,
  13     int_or_none,
  14 )
  15
  16
  17 class NiconicoIE(InfoExtractor):
  18     IE_NAME = 'niconico'
  19     IE_DESC = 'ニコニコ動画'
  20
  21     _TEST = {
  22         'url': 'http://www.nicovideo.jp/watch/sm22312215',
  23         'md5': 'd1a75c0823e2f629128c43e1212760f9',
  24         'info_dict': {
  25             'id': 'sm22312215',
  26             'ext': 'mp4',
  27             'title': 'Big Buck Bunny',
  28             'uploader': 'takuya0301',
  29             'uploader_id': '2698420',
  30             'upload_date': '20131123',
  31             'description': '(c) copyright 2008, Blender Foundation / www.bigbuckbunny.org',
  32             'duration': 33,
  33         },
  34         'params': {
  35             'username': 'ydl.niconico@gmail.com',
  36             'password': 'youtube-dl',
  37         },
  38     }
  39
  40     _VALID_URL = r'https?://(?:www\.|secure\.)?nicovideo\.jp/watch/((?:[a-z]{2})?[0-9]+)'
  41     _NETRC_MACHINE = 'niconico'
  42     # Determine whether the downloader uses authentication to download video
  43     _AUTHENTICATE = False
  44
  45     def _real_initialize(self):
  46         if self._downloader.params.get('username', None) is not None:
  47             self._AUTHENTICATE = True
  48
  49         if self._AUTHENTICATE:
  50             self._login()
  51
  52     def _login(self):
  53         (username, password) = self._get_login_info()
  54
  55         # Log in
  56         login_form_strs = {
  57             'mail': username,
  58             'password': password,
  59         }
  60         # Convert to UTF-8 *before* urlencode because Python 2.x's urlencode
  61         # chokes on unicode
  62         login_form = dict((k.encode('utf-8'), v.encode('utf-8')) for k, v in login_form_strs.items())
  63         login_data = compat_urllib_parse.urlencode(login_form).encode('utf-8')
  64         request = compat_urllib_request.Request(
  65             'https://secure.nicovideo.jp/secure/login', login_data)
  66         login_results = self._download_webpage(
  67             request, None, note='Logging in', errnote='Unable to log in')
  68         if re.search(r'(?i)<h1 class="mb8p4">Log in error</h1>', login_results) is not None:
  69             self._downloader.report_warning('unable to log in: bad username or password')
  70             return False
  71         return True
  72
  73     def _real_extract(self, url):
  74         mobj = re.match(self._VALID_URL, url)
  75         video_id = mobj.group(1)
  76
  77         # Get video webpage. We are not actually interested in it, but need
  78         # the cookies in order to be able to download the info webpage
  79         self._download_webpage('http://www.nicovideo.jp/watch/' + video_id, video_id)
  80
  81         video_info = self._download_xml(
  82             'http://ext.nicovideo.jp/api/getthumbinfo/' + video_id, video_id,
  83             note='Downloading video info page')
  84
  85         if self._AUTHENTICATE:
  86             # Get flv info
  87             flv_info_webpage = self._download_webpage(
  88                 'http://flapi.nicovideo.jp/api/getflv?v=' + video_id,
  89                 video_id, 'Downloading flv info')
  90         else:
  91             # Get external player info
  92             ext_player_info = self._download_webpage(
  93                 'http://ext.nicovideo.jp/thumb_watch/' + video_id, video_id)
  94             thumb_play_key = self._search_regex(
  95                 r'\'thumbPlayKey\'\s*:\s*\'(.*?)\'', ext_player_info, 'thumbPlayKey')
  96
  97             # Get flv info
  98             flv_info_data = compat_urllib_parse.urlencode({
  99                 'k': thumb_play_key,
 100                 'v': video_id
 101             })
 102             flv_info_request = compat_urllib_request.Request(
 103                 'http://ext.nicovideo.jp/thumb_watch', flv_info_data,
 104                 {'Content-Type': 'application/x-www-form-urlencoded'})
 105             flv_info_webpage = self._download_webpage(
 106                 flv_info_request, video_id,
 107                 note='Downloading flv info', errnote='Unable to download flv info')
 108
 109         video_real_url = compat_urlparse.parse_qs(flv_info_webpage)['url'][0]
 110
 111         # Start extracting information
 112         title = video_info.find('.//title').text
 113         extension = video_info.find('.//movie_type').text
 114         video_format = extension.upper()
 115         thumbnail = video_info.find('.//thumbnail_url').text
 116         description = video_info.find('.//description').text
 117         upload_date = unified_strdate(video_info.find('.//first_retrieve').text.split('+')[0])
 118         view_count = int_or_none(video_info.find('.//view_counter').text)
 119         comment_count = int_or_none(video_info.find('.//comment_num').text)
 120         duration = parse_duration(video_info.find('.//length').text)
 121         webpage_url = video_info.find('.//watch_url').text
 122
 123         if video_info.find('.//ch_id') is not None:
 124             uploader_id = video_info.find('.//ch_id').text
 125             uploader = video_info.find('.//ch_name').text
 126         elif video_info.find('.//user_id') is not None:
 127             uploader_id = video_info.find('.//user_id').text
 128             uploader = video_info.find('.//user_nickname').text
 129         else:
 130             uploader_id = uploader = None
 131
 132         return {
 133             'id': video_id,
 134             'url': video_real_url,
 135             'title': title,
 136             'ext': extension,
 137             'format': video_format,
 138             'thumbnail': thumbnail,
 139             'description': description,
 140             'uploader': uploader,
 141             'upload_date': upload_date,
 142             'uploader_id': uploader_id,
 143             'view_count': view_count,
 144             'comment_count': comment_count,
 145             'duration': duration,
 146             'webpage_url': webpage_url,
 147         }