[mitele] Fix extraction (Closes #6414)

youtube_dl/extractor/mitele.py

@@ -1,17 +1,11 @@
 from __future__ import unicode_literals
 
-import json
-
 from .common import InfoExtractor
-from ..compat import (
+from ..compat import compat_urllib_parse
-    compat_urllib_parse,
-    compat_urllib_parse_unquote,
-    compat_urlparse,
-)
 from ..utils import (
+    encode_dict,
     get_element_by_attribute,
-    parse_duration,
+    int_or_none,
-    strip_jsonp,
 )
 
 
@@ -21,54 +15,71 @@ class MiTeleIE(InfoExtractor):
 
     _TESTS = [{
         'url': 'http://www.mitele.es/programas-tv/diario-de/la-redaccion/programa-144/',
+        'md5': 'ace7635b2a0b286aaa37d3ff192d2a8a',
         'info_dict': {
-            'id': '0fce117d',
+            'id': '0NF1jJnxS1Wu3pHrmvFyw2',
-            'ext': 'mp4',
-            'title': 'Programa 144 - Tor, la web invisible',
-            'description': 'md5:3b6fce7eaa41b2d97358726378d9369f',
             'display_id': 'programa-144',
+            'ext': 'flv',
+            'title': 'Tor, la web invisible',
+            'description': 'md5:3b6fce7eaa41b2d97358726378d9369f',
+            'thumbnail': 're:(?i)^https?://.*\.jpg$',
             'duration': 2913,
         },
-        'params': {
-            # m3u8 download
-            'skip_download': True,
-        },
     }]
 
     def _real_extract(self, url):
-        episode = self._match_id(url)
+        display_id = self._match_id(url)
-        webpage = self._download_webpage(url, episode)
-        embed_data_json = self._search_regex(
-            r'(?s)MSV\.embedData\[.*?\]\s*=\s*({.*?});', webpage, 'embed data',
-        ).replace('\'', '"')
-        embed_data = json.loads(embed_data_json)
 
-        domain = embed_data['mediaUrl']
+        webpage = self._download_webpage(url, display_id)
-        if not domain.startswith('http'):
-            # only happens in telecinco.es videos
-            domain = 'http://' + domain
-        info_url = compat_urlparse.urljoin(
-            domain,
-            compat_urllib_parse_unquote(embed_data['flashvars']['host'])
-        )
-        info_el = self._download_xml(info_url, episode).find('./video/info')
 
-        video_link = info_el.find('videoUrl/link').text
+        config_url = self._search_regex(
-        token_query = compat_urllib_parse.urlencode({'id': video_link})
+            r'data-config\s*=\s*"([^"]+)"', webpage, 'data config url')
-        token_info = self._download_json(
+
-            embed_data['flashvars']['ov_tk'] + '?' + token_query,
+        config = self._download_json(
-            episode,
+            config_url, display_id, 'Downloading config JSON')
-            transform_source=strip_jsonp
+
-        )
+        mmc = self._download_json(
-        formats = self._extract_m3u8_formats(
+            config['services']['mmc'], display_id, 'Downloading mmc JSON')
-            token_info['tokenizedUrl'], episode, ext='mp4')
+
+        formats = []
+        for location in mmc['locations']:
+            gat = self._proto_relative_url(location.get('gat'), 'http:')
+            bas = location.get('bas')
+            loc = location.get('loc')
+            ogn = location.get('ogn')
+            if None in (gat, bas, loc, ogn):
+                continue
+            token_data = {
+                'bas': bas,
+                'icd': loc,
+                'ogn': ogn,
+                'sta': '0',
+            }
+            media = self._download_json(
+                '%s/?%s' % (gat, compat_urllib_parse.urlencode(encode_dict(token_data)).encode('utf-8')),
+                display_id, 'Downloading %s JSON' % location['loc'])
+            file_ = media.get('file')
+            if not file_:
+                continue
+            formats.extend(self._extract_f4m_formats(
+                file_ + '&hdcore=3.2.0&plugin=aasp-3.2.0.77.18',
+                display_id, f4m_id=loc))
+
+        title = self._search_regex(
+            r'class="Destacado-text"[^>]*>\s*<strong>([^<]+)</strong>', webpage, 'title')
+
+        video_id = self._search_regex(
+            r'data-media-id\s*=\s*"([^"]+)"', webpage,
+            'data media id', default=None) or display_id
+        thumbnail = config.get('poster', {}).get('imageUrl')
+        duration = int_or_none(mmc.get('duration'))
 
         return {
-            'id': embed_data['videoId'],
+            'id': video_id,
-            'display_id': episode,
+            'display_id': display_id,
-            'title': info_el.find('title').text,
+            'title': title,
-            'formats': formats,
             'description': get_element_by_attribute('class', 'text', webpage),
-            'thumbnail': info_el.find('thumb').text,
+            'thumbnail': thumbnail,
-            'duration': parse_duration(info_el.find('duration').text),
+            'duration': duration,
+            'formats': formats,
         }