diff options
Diffstat (limited to 'youtube_dl/extractor')
-rw-r--r-- | youtube_dl/extractor/__init__.py | 1 | ||||
-rw-r--r-- | youtube_dl/extractor/miomio_tv.py | 74 |
2 files changed, 75 insertions, 0 deletions
diff --git a/youtube_dl/extractor/__init__.py b/youtube_dl/extractor/__init__.py index f3ebe4ee2..0b9736f2d 100644 --- a/youtube_dl/extractor/__init__.py +++ b/youtube_dl/extractor/__init__.py @@ -274,6 +274,7 @@ from .mdr import MDRIE from .metacafe import MetacafeIE from .metacritic import MetacriticIE from .mgoon import MgoonIE +from .miomio_tv import MiomioTvIE from .minhateca import MinhatecaIE from .ministrygrid import MinistryGridIE from .mit import TechTVMITIE, MITIE, OCWMITIE diff --git a/youtube_dl/extractor/miomio_tv.py b/youtube_dl/extractor/miomio_tv.py new file mode 100644 index 000000000..ae20a32fa --- /dev/null +++ b/youtube_dl/extractor/miomio_tv.py @@ -0,0 +1,74 @@ +# coding: utf-8 +from __future__ import unicode_literals + +from .common import InfoExtractor + + +class MiomioTvIE(InfoExtractor): + IE_NAME = 'miomio.tv' + _VALID_URL = r'https?://(?:www\.)?miomio\.tv/watch/cc(?P<id>[0-9]+)' + _TEST = { + 'url': 'http://www.miomio.tv/watch/cc179734/', + 'md5': '48de02137d0739c15b440a224ad364b9', + 'info_dict': { + 'id': '179734', + 'title': u'\u624b\u7ed8\u52a8\u6f2b\u9b3c\u6ce3\u4f46\u4e01\u5168\u7a0b\u753b\u6cd5', + 'ext': 'flv' + } + } + + def _real_extract(self, url): + video_id = self._match_id(url) + webpage = self._download_webpage(url, video_id) + + title = self._html_search_regex(r'<meta\s+name="description"\s+content="\s*([^"]*)\s*"', webpage, 'title') + ref_path = self._search_regex(r'src="(/mioplayer/.*?)"', webpage, 'ref_path') + referer = 'http://www.miomio.tv{0}'.format(ref_path) + xml_config = self._search_regex(r'flashvars="type=sina&(.*?)&cid=', webpage, 'xml config') + + # skipping the following page causes lags and eventually connection drop-outs + # id is normally a rotating three digit value but a fixed value always appears to work + self._request_webpage("http://www.miomio.tv/mioplayer/mioplayerconfigfiles/xml.php?id={0}&r=cc{1}".format(id, 945), video_id) + + # the following xml contains the actual configuration information on the video file(s) + xml_url = 'http://www.miomio.tv/mioplayer/mioplayerconfigfiles/sina.php?{0}'.format(xml_config) + vidconfig = self._download_xml(xml_url, video_id) + + file_els = vidconfig.findall('.//durl') + + entries = [] + + for file_el in file_els: + segment_id = file_el.find('order').text.strip() + segment_title = '_'.join([title, segment_id]) + segment_duration = file_el.find('length').text.strip() + segment_url = file_el.find('url').text.strip() + + entries.append({ + 'id': segment_id, + 'title': segment_title, + 'duration': segment_duration, + 'url': segment_url + }) + + http_headers = { + 'Referer': referer, + 'Accept-Encoding': 'gzip, deflate', + 'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8' + } + + if len(entries) == 1: + return { + 'id': video_id, + 'title': title, + 'url': entries[0]['url'], + 'http_headers': http_headers + } + + return { + '_type': 'multi_video', + 'id': video_id, + 'title': title, + 'entries': entries, + 'http_headers': http_headers + } |