Raphaël G. Git Repositories - youtubedl/blob - youtube_dl/extractor/nextmedia.py

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 from .common import InfoExtractor
   5 from ..compat import compat_urlparse
   6 from ..utils import (
   7     clean_html,
   8     get_element_by_class,
   9     int_or_none,
  10     parse_iso8601,
  11     remove_start,
  12     unified_timestamp,
  13 )
  14
  15
  16 class NextMediaIE(InfoExtractor):
  17     IE_DESC = '蘋果日報'
  18     _VALID_URL = r'https?://hk\.apple\.nextmedia\.com/[^/]+/[^/]+/(?P<date>\d+)/(?P<id>\d+)'
  19     _TESTS = [{
  20         'url': 'http://hk.apple.nextmedia.com/realtime/news/20141108/53109199',
  21         'md5': 'dff9fad7009311c421176d1ac90bfe4f',
  22         'info_dict': {
  23             'id': '53109199',
  24             'ext': 'mp4',
  25             'title': '【佔領金鐘】50外國領事議員撐場 讚學生勇敢香港有希望',
  26             'thumbnail': r're:^https?://.*\.jpg$',
  27             'description': 'md5:28222b9912b6665a21011b034c70fcc7',
  28             'timestamp': 1415456273,
  29             'upload_date': '20141108',
  30         }
  31     }]
  32
  33     _URL_PATTERN = r'\{ url: \'(.+)\' \}'
  34
  35     def _real_extract(self, url):
  36         news_id = self._match_id(url)
  37         page = self._download_webpage(url, news_id)
  38         return self._extract_from_nextmedia_page(news_id, url, page)
  39
  40     def _extract_from_nextmedia_page(self, news_id, url, page):
  41         redirection_url = self._search_regex(
  42             r'window\.location\.href\s*=\s*([\'"])(?P<url>(?!\1).+)\1',
  43             page, 'redirection URL', default=None, group='url')
  44         if redirection_url:
  45             return self.url_result(compat_urlparse.urljoin(url, redirection_url))
  46
  47         title = self._fetch_title(page)
  48         video_url = self._search_regex(self._URL_PATTERN, page, 'video url')
  49
  50         attrs = {
  51             'id': news_id,
  52             'title': title,
  53             'url': video_url,  # ext can be inferred from url
  54             'thumbnail': self._fetch_thumbnail(page),
  55             'description': self._fetch_description(page),
  56         }
  57
  58         timestamp = self._fetch_timestamp(page)
  59         if timestamp:
  60             attrs['timestamp'] = timestamp
  61         else:
  62             attrs['upload_date'] = self._fetch_upload_date(url)
  63
  64         return attrs
  65
  66     def _fetch_title(self, page):
  67         return self._og_search_title(page)
  68
  69     def _fetch_thumbnail(self, page):
  70         return self._og_search_thumbnail(page)
  71
  72     def _fetch_timestamp(self, page):
  73         dateCreated = self._search_regex('"dateCreated":"([^"]+)"', page, 'created time')
  74         return parse_iso8601(dateCreated)
  75
  76     def _fetch_upload_date(self, url):
  77         return self._search_regex(self._VALID_URL, url, 'upload date', group='date')
  78
  79     def _fetch_description(self, page):
  80         return self._og_search_property('description', page)
  81
  82
  83 class NextMediaActionNewsIE(NextMediaIE):
  84     IE_DESC = '蘋果日報 - 動新聞'
  85     _VALID_URL = r'https?://hk\.dv\.nextmedia\.com/actionnews/[^/]+/(?P<date>\d+)/(?P<id>\d+)/\d+'
  86     _TESTS = [{
  87         'url': 'http://hk.dv.nextmedia.com/actionnews/hit/20150121/19009428/20061460',
  88         'md5': '05fce8ffeed7a5e00665d4b7cf0f9201',
  89         'info_dict': {
  90             'id': '19009428',
  91             'ext': 'mp4',
  92             'title': '【壹週刊】細10年男友偷食　50歲邵美琪再失戀',
  93             'thumbnail': r're:^https?://.*\.jpg$',
  94             'description': 'md5:cd802fad1f40fd9ea178c1e2af02d659',
  95             'timestamp': 1421791200,
  96             'upload_date': '20150120',
  97         }
  98     }]
  99
 100     def _real_extract(self, url):
 101         news_id = self._match_id(url)
 102         actionnews_page = self._download_webpage(url, news_id)
 103         article_url = self._og_search_url(actionnews_page)
 104         article_page = self._download_webpage(article_url, news_id)
 105         return self._extract_from_nextmedia_page(news_id, url, article_page)
 106
 107
 108 class AppleDailyIE(NextMediaIE):
 109     IE_DESC = '臺灣蘋果日報'
 110     _VALID_URL = r'https?://(www|ent)\.appledaily\.com\.tw/[^/]+/[^/]+/[^/]+/(?P<date>\d+)/(?P<id>\d+)(/.*)?'
 111     _TESTS = [{
 112         'url': 'http://ent.appledaily.com.tw/enews/article/entertainment/20150128/36354694',
 113         'md5': 'a843ab23d150977cc55ef94f1e2c1e4d',
 114         'info_dict': {
 115             'id': '36354694',
 116             'ext': 'mp4',
 117             'title': '周亭羽走過摩鐵陰霾2男陪吃 九把刀孤寒看醫生',
 118             'thumbnail': r're:^https?://.*\.jpg$',
 119             'description': 'md5:2acd430e59956dc47cd7f67cb3c003f4',
 120             'upload_date': '20150128',
 121         }
 122     }, {
 123         'url': 'http://www.appledaily.com.tw/realtimenews/article/strange/20150128/550549/%E4%B8%8D%E6%BB%BF%E8%A2%AB%E8%B8%A9%E8%85%B3%E3%80%80%E5%B1%B1%E6%9D%B1%E5%85%A9%E5%A4%A7%E5%AA%BD%E4%B8%80%E8%B7%AF%E6%89%93%E4%B8%8B%E8%BB%8A',
 124         'md5': '86b4e9132d158279c7883822d94ccc49',
 125         'info_dict': {
 126             'id': '550549',
 127             'ext': 'mp4',
 128             'title': '不滿被踩腳　山東兩大媽一路打下車',
 129             'thumbnail': r're:^https?://.*\.jpg$',
 130             'description': 'md5:175b4260c1d7c085993474217e4ab1b4',
 131             'upload_date': '20150128',
 132         }
 133     }, {
 134         'url': 'http://www.appledaily.com.tw/animation/realtimenews/new/20150128/5003671',
 135         'md5': '03df296d95dedc2d5886debbb80cb43f',
 136         'info_dict': {
 137             'id': '5003671',
 138             'ext': 'mp4',
 139             'title': '20正妹熱舞　《刀龍傳說Online》火辣上市',
 140             'thumbnail': r're:^https?://.*\.jpg$',
 141             'description': 'md5:23c0aac567dc08c9c16a3161a2c2e3cd',
 142             'upload_date': '20150128',
 143         },
 144         'skip': 'redirect to http://www.appledaily.com.tw/animation/',
 145     }, {
 146         # No thumbnail
 147         'url': 'http://www.appledaily.com.tw/animation/realtimenews/new/20150128/5003673/',
 148         'md5': 'b06182cd386ea7bc6115ec7ff0f72aeb',
 149         'info_dict': {
 150             'id': '5003673',
 151             'ext': 'mp4',
 152             'title': '半夜尿尿　好像會看到___',
 153             'description': 'md5:61d2da7fe117fede148706cdb85ac066',
 154             'upload_date': '20150128',
 155         },
 156         'expected_warnings': [
 157             'video thumbnail',
 158         ],
 159         'skip': 'redirect to http://www.appledaily.com.tw/animation/',
 160     }, {
 161         'url': 'http://www.appledaily.com.tw/appledaily/article/supplement/20140417/35770334/',
 162         'md5': 'eaa20e6b9df418c912d7f5dec2ba734d',
 163         'info_dict': {
 164             'id': '35770334',
 165             'ext': 'mp4',
 166             'title': '咖啡占卜測 XU裝熟指數',
 167             'thumbnail': r're:^https?://.*\.jpg$',
 168             'description': 'md5:7b859991a6a4fedbdf3dd3b66545c748',
 169             'upload_date': '20140417',
 170         },
 171     }, {
 172         'url': 'http://www.appledaily.com.tw/actionnews/appledaily/7/20161003/960588/',
 173         'only_matching': True,
 174     }, {
 175         # Redirected from http://ent.appledaily.com.tw/enews/article/entertainment/20150128/36354694
 176         'url': 'http://ent.appledaily.com.tw/section/article/headline/20150128/36354694',
 177         'only_matching': True,
 178     }]
 179
 180     _URL_PATTERN = r'\{url: \'(.+)\'\}'
 181
 182     def _fetch_title(self, page):
 183         return (self._html_search_regex(r'<h1 id="h1">([^<>]+)</h1>', page, 'news title', default=None)
 184                 or self._html_search_meta('description', page, 'news title'))
 185
 186     def _fetch_thumbnail(self, page):
 187         return self._html_search_regex(r"setInitialImage\(\'([^']+)'\)", page, 'video thumbnail', fatal=False)
 188
 189     def _fetch_timestamp(self, page):
 190         return None
 191
 192     def _fetch_description(self, page):
 193         return self._html_search_meta('description', page, 'news description')
 194
 195
 196 class NextTVIE(InfoExtractor):
 197     IE_DESC = '壹電視'
 198     _VALID_URL = r'https?://(?:www\.)?nexttv\.com\.tw/(?:[^/]+/)+(?P<id>\d+)'
 199
 200     _TEST = {
 201         'url': 'http://www.nexttv.com.tw/news/realtime/politics/11779671',
 202         'info_dict': {
 203             'id': '11779671',
 204             'ext': 'mp4',
 205             'title': '「超收稅」近4千億！　藍議員籲發消費券',
 206             'thumbnail': r're:^https?://.*\.jpg$',
 207             'timestamp': 1484825400,
 208             'upload_date': '20170119',
 209             'view_count': int,
 210         },
 211     }
 212
 213     def _real_extract(self, url):
 214         video_id = self._match_id(url)
 215
 216         webpage = self._download_webpage(url, video_id)
 217
 218         title = self._html_search_regex(
 219             r'<h1[^>]*>([^<]+)</h1>', webpage, 'title')
 220
 221         data = self._hidden_inputs(webpage)
 222
 223         video_url = data['ntt-vod-src-detailview']
 224
 225         date_str = get_element_by_class('date', webpage)
 226         timestamp = unified_timestamp(date_str + '+0800') if date_str else None
 227
 228         view_count = int_or_none(remove_start(
 229             clean_html(get_element_by_class('click', webpage)), '點閱：'))
 230
 231         return {
 232             'id': video_id,
 233             'title': title,
 234             'url': video_url,
 235             'thumbnail': data.get('ntt-vod-img-src'),
 236             'timestamp': timestamp,
 237             'view_count': view_count,
 238         }