[wat] extract all formats

youtube_dl/extractor/wat.py

@@ -2,25 +2,26 @@
 from __future__ import unicode_literals
 
 import re
-import hashlib
 
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
     ExtractorError,
     unified_strdate,
+    HEADRequest,
+    float_or_none,
 )
 
 
 class WatIE(InfoExtractor):
-    _VALID_URL = r'(?:wat:(?P<real_id>\d{8})|https?://www\.wat\.tv/video/(?P<display_id>.*)-(?P<short_id>.*?)_.*?\.html)'
+    _VALID_URL = r'(?:wat:|https?://(?:www\.)?wat\.tv/video/.*-)(?P<id>[0-9a-z]+)'
     IE_NAME = 'wat.tv'
     _TESTS = [
         {
             'url': 'http://www.wat.tv/video/soupe-figues-l-orange-aux-epices-6z1uz_2hvf7_.html',
-            'md5': 'ce70e9223945ed26a8056d413ca55dc9',
+            'md5': '83d882d9de5c9d97f0bb2c6273cde56a',
             'info_dict': {
                 'id': '11713067',
-                'display_id': 'soupe-figues-l-orange-aux-epices',
                 'ext': 'mp4',
                 'title': 'Soupe de figues à l\'orange et aux épices',
                 'description': 'Retrouvez l\'émission "Petits plats en équilibre", diffusée le 18 août 2014.',
@@ -33,7 +34,6 @@ class WatIE(InfoExtractor):
             'md5': 'fbc84e4378165278e743956d9c1bf16b',
             'info_dict': {
                 'id': '11713075',
-                'display_id': 'gregory-lemarchal-voix-ange',
                 'ext': 'mp4',
                 'title': 'Grégory Lemarchal, une voix d\'ange depuis 10 ans (1/3)',
                 'description': 'md5:b7a849cf16a2b733d9cd10c52906dee3',
@@ -44,96 +44,85 @@ class WatIE(InfoExtractor):
         },
     ]
 
-    def download_video_info(self, real_id):
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        video_id = video_id if video_id.isdigit() and len(video_id) > 6 else compat_str(int(video_id, 36))
+
         # 'contentv4' is used in the website, but it also returns the related
         # videos, we don't need them
-        info = self._download_json('http://www.wat.tv/interface/contentv3/' + real_id, real_id)
+        video_info = self._download_json(
-        return info['media']
+            'http://www.wat.tv/interface/contentv3/' + video_id, video_id)['media']
-
-    def _real_extract(self, url):
-        def real_id_for_chapter(chapter):
-            return chapter['tc_start'].split('-')[0]
-        mobj = re.match(self._VALID_URL, url)
-        display_id = mobj.group('display_id')
-        real_id = mobj.group('real_id')
-        if not real_id:
-            short_id = mobj.group('short_id')
-            webpage = self._download_webpage(url, display_id or short_id)
-            real_id = self._search_regex(r'xtpage = ".*-(.*?)";', webpage, 'real id')
-
-        video_info = self.download_video_info(real_id)
 
         error_desc = video_info.get('error_desc')
         if error_desc:
             raise ExtractorError(
                 '%s returned error: %s' % (self.IE_NAME, error_desc), expected=True)
 
-        geo_list = video_info.get('geoList')
-        country = geo_list[0] if geo_list else ''
-
         chapters = video_info['chapters']
         first_chapter = chapters[0]
-        files = video_info['files']
-        first_file = files[0]
 
-        if real_id_for_chapter(first_chapter) != real_id:
+        def video_id_for_chapter(chapter):
+            return chapter['tc_start'].split('-')[0]
+
+        if video_id_for_chapter(first_chapter) != video_id:
             self.to_screen('Multipart video detected')
-            chapter_urls = []
+            entries = [self.url_result('wat:%s' % video_id_for_chapter(chapter)) for chapter in chapters]
-            for chapter in chapters:
+            return self.playlist_result(entries, video_id, video_info['title'])
-                chapter_id = real_id_for_chapter(chapter)
-                # Yes, when we this chapter is processed by WatIE,
-                # it will download the info again
-                chapter_info = self.download_video_info(chapter_id)
-                chapter_urls.append(chapter_info['url'])
-            entries = [self.url_result(chapter_url) for chapter_url in chapter_urls]
-            return self.playlist_result(entries, real_id, video_info['title'])
-
-        upload_date = None
-        if 'date_diffusion' in first_chapter:
-            upload_date = unified_strdate(first_chapter['date_diffusion'])
         # Otherwise we can continue and extract just one part, we have to use
-        # the short id for getting the video url
+        # the video id for getting the video url
 
-        formats = [{
+        date_diffusion = first_chapter.get('date_diffusion')
-            'url': 'http://wat.tv/get/android5/%s.mp4' % real_id,
+        upload_date = unified_strdate(date_diffusion) if date_diffusion else None
-            'format_id': 'Mobile',
-        }]
 
-        fmts = [('SD', 'web')]
+        def extract_url(path_template, url_type):
-        if first_file.get('hasHD'):
+            req_url = 'http://www.wat.tv/get/%s' % (path_template % video_id)
-            fmts.append(('HD', 'webhd'))
+            head = self._request_webpage(HEADRequest(req_url), video_id, 'Extracting %s url' % url_type)
+            red_url = head.geturl()
+            if req_url == red_url:
+                raise ExtractorError(
+                    '%s said: Sorry, this video is not available from your country.' % self.IE_NAME,
+                    expected=True)
+            return red_url
 
-        def compute_token(param):
+        m3u8_url = extract_url('ipad/%s.m3u8', 'm3u8')
-            timestamp = '%08x' % int(self._download_webpage(
+        http_url = extract_url('android5/%s.mp4', 'http')
-                'http://www.wat.tv/servertime', real_id,
-                'Downloading server time').split('|')[0])
-            magic = '9b673b13fa4682ed14c3cfa5af5310274b514c4133e9b3a81e6e3aba009l2564'
-            return '%s/%s' % (hashlib.md5((magic + param + timestamp).encode('ascii')).hexdigest(), timestamp)
 
-        for fmt in fmts:
+        formats = []
-            webid = '/%s/%s' % (fmt[1], real_id)
+        m3u8_formats = self._extract_m3u8_formats(
-            video_url = self._download_webpage(
+            m3u8_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls')
-                'http://www.wat.tv/get%s?token=%s&getURL=1&country=%s' % (webid, compute_token(webid), country),
+        formats.extend(m3u8_formats)
-                real_id,
+        formats.extend(self._extract_f4m_formats(
-                'Downloading %s video URL' % fmt[0],
+            m3u8_url.replace('ios.', 'web.').replace('.m3u8', '.f4m'),
-                'Failed to download %s video URL' % fmt[0],
+            video_id, f4m_id='hds', fatal=False))
-                False)
+        for m3u8_format in m3u8_formats:
-            if not video_url:
+            mobj = re.search(
+                r'audio.*?%3D(\d+)(?:-video.*?%3D(\d+))?', m3u8_format['url'])
+            if not mobj:
                 continue
-            formats.append({
+            abr, vbr = mobj.groups()
-                'url': video_url,
+            abr, vbr = float_or_none(abr, 1000), float_or_none(vbr, 1000)
-                'ext': 'mp4',
+            m3u8_format.update({
-                'format_id': fmt[0],
+                'vbr': vbr,
+                'abr': abr,
             })
+            if not vbr or not abr:
+                continue
+            f = m3u8_format.copy()
+            f.update({
+                'url': re.sub(r'%s-\d+00-\d+' % video_id, '%s-%d00-%d' % (video_id, round(vbr / 100), round(abr)), http_url),
+                'format_id': f['format_id'].replace('hls', 'http'),
+                'protocol': 'http',
+            })
+            formats.append(f)
+        self._sort_formats(formats)
 
         return {
-            'id': real_id,
+            'id': video_id,
-            'display_id': display_id,
             'title': first_chapter['title'],
             'thumbnail': first_chapter['preview'],
             'description': first_chapter['description'],
             'view_count': video_info['views'],
             'upload_date': upload_date,
-            'duration': first_file['duration'],
+            'duration': video_info['files'][0]['duration'],
             'formats': formats,
         }