yt_dlp/extractor/infoq.py

   1 import base64
   2 import urllib.parse
   3
   4 from .bokecc import BokeCCBaseIE
   5 from ..utils import (
   6     ExtractorError,
   7     determine_ext,
   8     traverse_obj,
   9     update_url_query,
  10 )
  11
  12
  13 class InfoQIE(BokeCCBaseIE):
  14     _VALID_URL = r'https?://(?:www\.)?infoq\.com/(?:[^/]+/)+(?P<id>[^/]+)'
  15
  16     _TESTS = [{
  17         'url': 'http://www.infoq.com/presentations/A-Few-of-My-Favorite-Python-Things',
  18         'md5': 'b5ca0e0a8c1fed93b0e65e48e462f9a2',
  19         'info_dict': {
  20             'id': 'A-Few-of-My-Favorite-Python-Things',
  21             'ext': 'mp4',
  22             'description': 'Mike Pirnat presents some tips and tricks, standard libraries and third party packages that make programming in Python a richer experience.',
  23             'title': 'A Few of My Favorite [Python] Things',
  24         },
  25     }, {
  26         'url': 'http://www.infoq.com/fr/presentations/changez-avis-sur-javascript',
  27         'only_matching': True,
  28     }, {
  29         'url': 'http://www.infoq.com/cn/presentations/openstack-continued-delivery',
  30         'md5': '4918d0cca1497f2244572caf626687ef',
  31         'info_dict': {
  32             'id': 'openstack-continued-delivery',
  33             'title': 'OpenStack持续交付之路',
  34             'ext': 'flv',
  35             'description': 'md5:308d981fb28fa42f49f9568322c683ff',
  36         },
  37         'skip': 'Sorry, the page you visited does not exist',
  38     }, {
  39         'url': 'https://www.infoq.com/presentations/Simple-Made-Easy',
  40         'md5': '0e34642d4d9ef44bf86f66f6399672db',
  41         'info_dict': {
  42             'id': 'Simple-Made-Easy',
  43             'title': 'Simple Made Easy',
  44             'ext': 'mp3',
  45             'description': 'md5:3e0e213a8bbd074796ef89ea35ada25b',
  46         },
  47         'params': {
  48             'format': 'bestaudio',
  49         },
  50     }]
  51
  52     def _extract_rtmp_video(self, webpage):
  53         # The server URL is hardcoded
  54         video_url = 'rtmpe://videof.infoq.com/cfx/st/'
  55
  56         # Extract video URL
  57         encoded_id = self._search_regex(
  58             r"jsclassref\s*=\s*'([^']*)'", webpage, 'encoded id', default=None)
  59
  60         real_id = urllib.parse.unquote(base64.b64decode(encoded_id).decode('utf-8'))
  61         playpath = 'mp4:' + real_id
  62
  63         return [{
  64             'format_id': 'rtmp_video',
  65             'url': video_url,
  66             'ext': determine_ext(playpath),
  67             'play_path': playpath,
  68         }]
  69
  70     def _extract_cf_auth(self, webpage):
  71         policy = self._search_regex(r'InfoQConstants\.scp\s*=\s*\'([^\']+)\'', webpage, 'policy')
  72         signature = self._search_regex(r'InfoQConstants\.scs\s*=\s*\'([^\']+)\'', webpage, 'signature')
  73         key_pair_id = self._search_regex(r'InfoQConstants\.sck\s*=\s*\'([^\']+)\'', webpage, 'key-pair-id')
  74         return {
  75             'Policy': policy,
  76             'Signature': signature,
  77             'Key-Pair-Id': key_pair_id,
  78         }
  79
  80     def _extract_http_video(self, webpage):
  81         http_video_url = self._search_regex(r'P\.s\s*=\s*\'([^\']+)\'', webpage, 'video URL')
  82         http_video_url = update_url_query(http_video_url, self._extract_cf_auth(webpage))
  83         return [{
  84             'format_id': 'http_video',
  85             'url': http_video_url,
  86             'http_headers': {'Referer': 'https://www.infoq.com/'},
  87         }]
  88
  89     def _extract_http_audio(self, webpage, video_id):
  90         try:
  91             http_audio_url = traverse_obj(self._form_hidden_inputs('mp3Form', webpage), 'filename')
  92         except ExtractorError:
  93             http_audio_url = None
  94         if not http_audio_url:
  95             return []
  96
  97         # base URL is found in the Location header in the response returned by
  98         # GET https://www.infoq.com/mp3download.action?filename=... when logged in.
  99         http_audio_url = urllib.parse.urljoin('http://ress.infoq.com/downloads/mp3downloads/', http_audio_url)
 100         http_audio_url = update_url_query(http_audio_url, self._extract_cf_auth(webpage))
 101
 102         # audio file seem to be missing some times even if there is a download link
 103         # so probe URL to make sure
 104         if not self._is_valid_url(http_audio_url, video_id):
 105             return []
 106
 107         return [{
 108             'format_id': 'http_audio',
 109             'url': http_audio_url,
 110             'vcodec': 'none',
 111         }]
 112
 113     def _real_extract(self, url):
 114         video_id = self._match_id(url)
 115         webpage = self._download_webpage(url, video_id)
 116
 117         video_title = self._html_extract_title(webpage)
 118         video_description = self._html_search_meta('description', webpage, 'description')
 119
 120         if '/cn/' in url:
 121             # for China videos, HTTP video URL exists but always fails with 403
 122             formats = self._extract_bokecc_formats(webpage, video_id)
 123         else:
 124             formats = (
 125                 self._extract_rtmp_video(webpage)
 126                 + self._extract_http_video(webpage)
 127                 + self._extract_http_audio(webpage, video_id))
 128
 129         return {
 130             'id': video_id,
 131             'title': video_title,
 132             'description': video_description,
 133             'formats': formats,
 134         }