[cnet] improve extraction

- relex data json regex
- extract the platform metadata once
- extract hds formats
- extract duration
- extract thumbnail
This commit is contained in:
remitamine 2015-12-20 12:43:00 +01:00
parent a641b24592
commit c6ed6fadc2

View File

@ -1,33 +1,33 @@
# coding: utf-8 # coding: utf-8
from __future__ import unicode_literals from __future__ import unicode_literals
import json
from .common import InfoExtractor
from .theplatform import ThePlatformIE from .theplatform import ThePlatformIE
from ..utils import int_or_none
class CNETIE(InfoExtractor): class CNETIE(ThePlatformIE):
_VALID_URL = r'https?://(?:www\.)?cnet\.com/videos/(?P<id>[^/]+)/' _VALID_URL = r'https?://(?:www\.)?cnet\.com/videos/(?P<id>[^/]+)/'
_TESTS = [{ _TESTS = [{
'url': 'http://www.cnet.com/videos/hands-on-with-microsofts-windows-8-1-update/', 'url': 'http://www.cnet.com/videos/hands-on-with-microsofts-windows-8-1-update/',
'info_dict': { 'info_dict': {
'id': '56f4ea68-bd21-4852-b08c-4de5b8354c60', 'id': '56f4ea68-bd21-4852-b08c-4de5b8354c60',
'ext': 'mp4', 'ext': 'flv',
'title': 'Hands-on with Microsoft Windows 8.1 Update', 'title': 'Hands-on with Microsoft Windows 8.1 Update',
'description': 'The new update to the Windows 8 OS brings improved performance for mouse and keyboard users.', 'description': 'The new update to the Windows 8 OS brings improved performance for mouse and keyboard users.',
'uploader_id': '6085384d-619e-11e3-b231-14feb5ca9861', 'uploader_id': '6085384d-619e-11e3-b231-14feb5ca9861',
'uploader': 'Sarah Mitroff', 'uploader': 'Sarah Mitroff',
'duration': 70,
}, },
}, { }, {
'url': 'http://www.cnet.com/videos/whiny-pothole-tweets-at-local-government-when-hit-by-cars-tomorrow-daily-187/', 'url': 'http://www.cnet.com/videos/whiny-pothole-tweets-at-local-government-when-hit-by-cars-tomorrow-daily-187/',
'info_dict': { 'info_dict': {
'id': '56527b93-d25d-44e3-b738-f989ce2e49ba', 'id': '56527b93-d25d-44e3-b738-f989ce2e49ba',
'ext': 'mp4', 'ext': 'flv',
'title': 'Whiny potholes tweet at local government when hit by cars (Tomorrow Daily 187)',
'description': 'Khail and Ashley wonder what other civic woes can be solved by self-tweeting objects, investigate a new kind of VR camera and watch an origami robot self-assemble, walk, climb, dig and dissolve. #TDPothole', 'description': 'Khail and Ashley wonder what other civic woes can be solved by self-tweeting objects, investigate a new kind of VR camera and watch an origami robot self-assemble, walk, climb, dig and dissolve. #TDPothole',
'uploader_id': 'b163284d-6b73-44fc-b3e6-3da66c392d40', 'uploader_id': 'b163284d-6b73-44fc-b3e6-3da66c392d40',
'uploader': 'Ashley Esqueda', 'uploader': 'Ashley Esqueda',
'title': 'Whiny potholes tweet at local government when hit by cars (Tomorrow Daily 187)', 'duration': 1482,
}, },
}] }]
@ -36,10 +36,10 @@ def _real_extract(self, url):
webpage = self._download_webpage(url, display_id) webpage = self._download_webpage(url, display_id)
data_json = self._html_search_regex( data_json = self._html_search_regex(
r"<div class=\"videoPlayer\"\s+.*?data-cnet-video-uvp-options='([^']+)'", r"data-cnet-video(?:-uvp)?-options='([^']+)'",
webpage, 'data json') webpage, 'data json')
data = json.loads(data_json) data = self._parse_json(data_json, display_id)
vdata = data['videos'][0] vdata = data.get('video') or data['videos'][0]
video_id = vdata['id'] video_id = vdata['id']
title = vdata['title'] title = vdata['title']
@ -52,19 +52,22 @@ def _real_extract(self, url):
uploader_id = None uploader_id = None
mpx_account = data['config']['uvpConfig']['default']['mpx_account'] mpx_account = data['config']['uvpConfig']['default']['mpx_account']
tp = ThePlatformIE(self._downloader)
metadata = self.get_metadata('%s/%s' % (mpx_account, list(vdata['files'].values())[0]), video_id)
description = vdata.get('description') or metadata.get('description')
duration = int_or_none(vdata.get('duration')) or metadata.get('duration')
formats = [] formats = []
subtitles = {} subtitles = {}
description = vdata.get('description')
for (fkey, vid) in vdata['files'].items(): for (fkey, vid) in vdata['files'].items():
if fkey == 'hls_phone' and 'hls_tablet' in vdata['files']: if fkey == 'hls_phone' and 'hls_tablet' in vdata['files']:
continue continue
result = tp.extract(('http://link.theplatform.com/s/%s/%s' % (mpx_account, vid))) release_url = 'http://link.theplatform.com/s/%s/%s?format=SMIL&mbr=true' % (mpx_account, vid)
formats.extend(result['formats']) if fkey == 'hds':
subtitles = self._merge_subtitles(subtitles, result['subtitles']) release_url += '&manifest=f4m'
description = description or result.get('description') tp_formats, tp_subtitles = self._extract_theplatform_smil(release_url, video_id, 'Downloading %s SMIL data' % fkey)
formats.extend(tp_formats)
subtitles = self._merge_subtitles(subtitles, tp_subtitles)
self._sort_formats(formats) self._sort_formats(formats)
return { return {
@ -72,6 +75,8 @@ def _real_extract(self, url):
'display_id': display_id, 'display_id': display_id,
'title': title, 'title': title,
'description': description, 'description': description,
'thumbnail': metadata.get('thumbnail'),
'duration': duration,
'uploader': uploader, 'uploader': uploader,
'uploader_id': uploader_id, 'uploader_id': uploader_id,
'subtitles': subtitles, 'subtitles': subtitles,