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

   1 # coding: utf-8
   2 from __future__ import unicode_literals
   3
   4 import random
   5 import re
   6
   7 from .common import InfoExtractor
   8 from ..compat import compat_urllib_parse_unquote_plus
   9 from ..utils import (
  10     int_or_none,
  11     float_or_none,
  12     timeconvert,
  13     update_url_query,
  14     xpath_text,
  15 )
  16
  17
  18 class KUSIIE(InfoExtractor):
  19     _VALID_URL = r'https?://(?:www\.)?kusi\.com/(?P<path>story/.+|video\?clipId=(?P<clipId>\d+))'
  20     _TESTS = [{
  21         'url': 'http://www.kusi.com/story/31183873/turko-files-case-closed-put-on-hold',
  22         'md5': 'f926e7684294cf8cb7bdf8858e1b3988',
  23         'info_dict': {
  24             'id': '12203019',
  25             'ext': 'mp4',
  26             'title': 'Turko Files: Case Closed! & Put On Hold!',
  27             'duration': 231.0,
  28             'upload_date': '20160210',
  29             'timestamp': 1455087571,
  30             'thumbnail': 're:^https?://.*\.jpg$'
  31         },
  32     }, {
  33         'url': 'http://kusi.com/video?clipId=12203019',
  34         'info_dict': {
  35             'id': '12203019',
  36             'ext': 'mp4',
  37             'title': 'Turko Files: Case Closed! & Put On Hold!',
  38             'duration': 231.0,
  39             'upload_date': '20160210',
  40             'timestamp': 1455087571,
  41             'thumbnail': 're:^https?://.*\.jpg$'
  42         },
  43         'params': {
  44             'skip_download': True,  # Same as previous one
  45         },
  46     }]
  47
  48     def _real_extract(self, url):
  49         mobj = re.match(self._VALID_URL, url)
  50         clip_id = mobj.group('clipId')
  51         video_id = clip_id or mobj.group('path')
  52
  53         webpage = self._download_webpage(url, video_id)
  54
  55         if clip_id is None:
  56             video_id = clip_id = self._html_search_regex(
  57                 r'"clipId"\s*,\s*"(\d+)"', webpage, 'clip id')
  58
  59         affiliate_id = self._search_regex(
  60             r'affiliateId\s*:\s*\'([^\']+)\'', webpage, 'affiliate id')
  61
  62         # See __Packages/worldnow/model/GalleryModel.as of WNGallery.swf
  63         xml_url = update_url_query('http://www.kusi.com/build.asp', {
  64             'buildtype': 'buildfeaturexmlrequest',
  65             'featureType': 'Clip',
  66             'featureid': clip_id,
  67             'affiliateno': affiliate_id,
  68             'clientgroupid': '1',
  69             'rnd': int(round(random.random() * 1000000)),
  70         })
  71
  72         doc = self._download_xml(xml_url, video_id)
  73
  74         video_title = xpath_text(doc, 'HEADLINE', fatal=True)
  75         duration = float_or_none(xpath_text(doc, 'DURATION'), scale=1000)
  76         description = xpath_text(doc, 'ABSTRACT')
  77         thumbnail = xpath_text(doc, './THUMBNAILIMAGE/FILENAME')
  78         createtion_time = timeconvert(xpath_text(doc, 'rfc822creationdate'))
  79
  80         quality_options = doc.find('{http://search.yahoo.com/mrss/}group').findall('{http://search.yahoo.com/mrss/}content')
  81         formats = []
  82         for quality in quality_options:
  83             formats.append({
  84                 'url': compat_urllib_parse_unquote_plus(quality.attrib['url']),
  85                 'height': int_or_none(quality.attrib.get('height')),
  86                 'width': int_or_none(quality.attrib.get('width')),
  87                 'vbr': float_or_none(quality.attrib.get('bitratebits'), scale=1000),
  88             })
  89         self._sort_formats(formats)
  90
  91         return {
  92             'id': video_id,
  93             'title': video_title,
  94             'description': description,
  95             'duration': duration,
  96             'formats': formats,
  97             'thumbnail': thumbnail,
  98             'timestamp': createtion_time,
  99         }