yt_dlp/extractor/nextmedia.py

   1 import urllib.parse
   2
   3 from .common import InfoExtractor
   4 from ..utils import (
   5     clean_html,
   6     get_element_by_class,
   7     int_or_none,
   8     parse_iso8601,
   9     remove_start,
  10     unified_timestamp,
  11 )
  12
  13
  14 class NextMediaIE(InfoExtractor):
  15     IE_DESC = '蘋果日報'
  16     _VALID_URL = r'https?://hk\.apple\.nextmedia\.com/[^/]+/[^/]+/(?P<date>\d+)/(?P<id>\d+)'
  17     _TESTS = [{
  18         'url': 'http://hk.apple.nextmedia.com/realtime/news/20141108/53109199',
  19         'md5': 'dff9fad7009311c421176d1ac90bfe4f',
  20         'info_dict': {
  21             'id': '53109199',
  22             'ext': 'mp4',
  23             'title': '【佔領金鐘】50外國領事議員撐場 讚學生勇敢香港有希望',
  24             'thumbnail': r're:^https?://.*\.jpg$',
  25             'description': 'md5:28222b9912b6665a21011b034c70fcc7',
  26             'timestamp': 1415456273,
  27             'upload_date': '20141108',
  28         },
  29     }]
  30
  31     _URL_PATTERN = r'\{ url: \'(.+)\' \}'
  32
  33     def _real_extract(self, url):
  34         news_id = self._match_id(url)
  35         page = self._download_webpage(url, news_id)
  36         return self._extract_from_nextmedia_page(news_id, url, page)
  37
  38     def _extract_from_nextmedia_page(self, news_id, url, page):
  39         redirection_url = self._search_regex(
  40             r'window\.location\.href\s*=\s*([\'"])(?P<url>(?!\1).+)\1',
  41             page, 'redirection URL', default=None, group='url')
  42         if redirection_url:
  43             return self.url_result(urllib.parse.urljoin(url, redirection_url))
  44
  45         title = self._fetch_title(page)
  46         video_url = self._search_regex(self._URL_PATTERN, page, 'video url')
  47
  48         attrs = {
  49             'id': news_id,
  50             'title': title,
  51             'url': video_url,  # ext can be inferred from url
  52             'thumbnail': self._fetch_thumbnail(page),
  53             'description': self._fetch_description(page),
  54         }
  55
  56         timestamp = self._fetch_timestamp(page)
  57         if timestamp:
  58             attrs['timestamp'] = timestamp
  59         else:
  60             attrs['upload_date'] = self._fetch_upload_date(url)
  61
  62         return attrs
  63
  64     def _fetch_title(self, page):
  65         return self._og_search_title(page)
  66
  67     def _fetch_thumbnail(self, page):
  68         return self._og_search_thumbnail(page)
  69
  70     def _fetch_timestamp(self, page):
  71         date_created = self._search_regex('"dateCreated":"([^"]+)"', page, 'created time')
  72         return parse_iso8601(date_created)
  73
  74     def _fetch_upload_date(self, url):
  75         return self._search_regex(self._VALID_URL, url, 'upload date', group='date')
  76
  77     def _fetch_description(self, page):
  78         return self._og_search_property('description', page)
  79
  80
  81 class NextMediaActionNewsIE(NextMediaIE):  # XXX: Do not subclass from concrete IE
  82     IE_DESC = '蘋果日報 - 動新聞'
  83     _VALID_URL = r'https?://hk\.dv\.nextmedia\.com/actionnews/[^/]+/(?P<date>\d+)/(?P<id>\d+)/\d+'
  84     _TESTS = [{
  85         'url': 'http://hk.dv.nextmedia.com/actionnews/hit/20150121/19009428/20061460',
  86         'md5': '05fce8ffeed7a5e00665d4b7cf0f9201',
  87         'info_dict': {
  88             'id': '19009428',
  89             'ext': 'mp4',
  90             'title': '【壹週刊】細10年男友偷食　50歲邵美琪再失戀',
  91             'thumbnail': r're:^https?://.*\.jpg$',
  92             'description': 'md5:cd802fad1f40fd9ea178c1e2af02d659',
  93             'timestamp': 1421791200,
  94             'upload_date': '20150120',
  95         },
  96     }]
  97
  98     def _real_extract(self, url):
  99         news_id = self._match_id(url)
 100         actionnews_page = self._download_webpage(url, news_id)
 101         article_url = self._og_search_url(actionnews_page)
 102         article_page = self._download_webpage(article_url, news_id)
 103         return self._extract_from_nextmedia_page(news_id, url, article_page)
 104
 105
 106 class AppleDailyIE(NextMediaIE):  # XXX: Do not subclass from concrete IE
 107     IE_DESC = '臺灣蘋果日報'
 108     _VALID_URL = r'https?://(www|ent)\.appledaily\.com\.tw/[^/]+/[^/]+/[^/]+/(?P<date>\d+)/(?P<id>\d+)(/.*)?'
 109     _TESTS = [{
 110         'url': 'http://ent.appledaily.com.tw/enews/article/entertainment/20150128/36354694',
 111         'md5': 'a843ab23d150977cc55ef94f1e2c1e4d',
 112         'info_dict': {
 113             'id': '36354694',
 114             'ext': 'mp4',
 115             'title': '周亭羽走過摩鐵陰霾2男陪吃 九把刀孤寒看醫生',
 116             'thumbnail': r're:^https?://.*\.jpg$',
 117             'description': 'md5:2acd430e59956dc47cd7f67cb3c003f4',
 118             'upload_date': '20150128',
 119         },
 120     }, {
 121         'url': 'http://www.appledaily.com.tw/realtimenews/article/strange/20150128/550549/%E4%B8%8D%E6%BB%BF%E8%A2%AB%E8%B8%A9%E8%85%B3%E3%80%80%E5%B1%B1%E6%9D%B1%E5%85%A9%E5%A4%A7%E5%AA%BD%E4%B8%80%E8%B7%AF%E6%89%93%E4%B8%8B%E8%BB%8A',
 122         'md5': '86b4e9132d158279c7883822d94ccc49',
 123         'info_dict': {
 124             'id': '550549',
 125             'ext': 'mp4',
 126             'title': '不滿被踩腳　山東兩大媽一路打下車',
 127             'thumbnail': r're:^https?://.*\.jpg$',
 128             'description': 'md5:175b4260c1d7c085993474217e4ab1b4',
 129             'upload_date': '20150128',
 130         },
 131     }, {
 132         'url': 'http://www.appledaily.com.tw/animation/realtimenews/new/20150128/5003671',
 133         'md5': '03df296d95dedc2d5886debbb80cb43f',
 134         'info_dict': {
 135             'id': '5003671',
 136             'ext': 'mp4',
 137             'title': '20正妹熱舞　《刀龍傳說Online》火辣上市',
 138             'thumbnail': r're:^https?://.*\.jpg$',
 139             'description': 'md5:23c0aac567dc08c9c16a3161a2c2e3cd',
 140             'upload_date': '20150128',
 141         },
 142         'skip': 'redirect to http://www.appledaily.com.tw/animation/',
 143     }, {
 144         # No thumbnail
 145         'url': 'http://www.appledaily.com.tw/animation/realtimenews/new/20150128/5003673/',
 146         'md5': 'b06182cd386ea7bc6115ec7ff0f72aeb',
 147         'info_dict': {
 148             'id': '5003673',
 149             'ext': 'mp4',
 150             'title': '半夜尿尿　好像會看到___',
 151             'description': 'md5:61d2da7fe117fede148706cdb85ac066',
 152             'upload_date': '20150128',
 153         },
 154         'expected_warnings': [
 155             'video thumbnail',
 156         ],
 157         'skip': 'redirect to http://www.appledaily.com.tw/animation/',
 158     }, {
 159         'url': 'http://www.appledaily.com.tw/appledaily/article/supplement/20140417/35770334/',
 160         'md5': 'eaa20e6b9df418c912d7f5dec2ba734d',
 161         'info_dict': {
 162             'id': '35770334',
 163             'ext': 'mp4',
 164             'title': '咖啡占卜測 XU裝熟指數',
 165             'thumbnail': r're:^https?://.*\.jpg$',
 166             'description': 'md5:7b859991a6a4fedbdf3dd3b66545c748',
 167             'upload_date': '20140417',
 168         },
 169     }, {
 170         'url': 'http://www.appledaily.com.tw/actionnews/appledaily/7/20161003/960588/',
 171         'only_matching': True,
 172     }, {
 173         # Redirected from http://ent.appledaily.com.tw/enews/article/entertainment/20150128/36354694
 174         'url': 'http://ent.appledaily.com.tw/section/article/headline/20150128/36354694',
 175         'only_matching': True,
 176     }]
 177
 178     _URL_PATTERN = r'\{url: \'(.+)\'\}'
 179
 180     def _fetch_title(self, page):
 181         return (self._html_search_regex(r'<h1 id="h1">([^<>]+)</h1>', page, 'news title', default=None)
 182                 or self._html_search_meta('description', page, 'news title'))
 183
 184     def _fetch_thumbnail(self, page):
 185         return self._html_search_regex(r"setInitialImage\(\'([^']+)'\)", page, 'video thumbnail', fatal=False)
 186
 187     def _fetch_timestamp(self, page):
 188         return None
 189
 190     def _fetch_description(self, page):
 191         return self._html_search_meta('description', page, 'news description')
 192
 193
 194 class NextTVIE(InfoExtractor):
 195     _WORKING = False
 196     _ENABLED = None  # XXX: pass through to GenericIE
 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         }