WatIE: support videos divided in multiple parts (closes #222 and #659)

The id for the videos is now the full id, no the one in the webpage url.
Also extract more information: description, view_count and upload_date

youtube_dl/extractor/tf1.py

@@ -8,15 +8,16 @@ from .common import InfoExtractor
 class TF1IE(InfoExtractor):
     """
     TF1 uses the wat.tv player, currently it can only download videos with the
-    html5 player enabled, it cannot download HD videos or the news.
+    html5 player enabled, it cannot download HD videos.
     """
     _VALID_URL = r'http://videos.tf1.fr/.*-(.*?).html'
     _TEST = {
         u'url': u'http://videos.tf1.fr/auto-moto/citroen-grand-c4-picasso-2013-presentation-officielle-8062060.html',
-        u'file': u'6bysb.mp4',
+        u'file': u'10635995.mp4',
         u'md5': u'66789d3e91278d332f75e1feb7aea327',
         u'info_dict': {
-            u"title": u"Citroën Grand C4 Picasso 2013 : présentation officielle"
+            u'title': u'Citroën Grand C4 Picasso 2013 : présentation officielle',
+            u'description': u'Vidéo officielle du nouveau Citroën Grand C4 Picasso, lancé à l\'automne 2013.',
         }
     }
 

youtube_dl/extractor/wat.py

@@ -1,3 +1,5 @@
+# coding: utf-8
+
 import json
 import re
 
@@ -5,6 +7,7 @@ from .common import InfoExtractor
 
 from ..utils import (
     compat_urllib_parse,
+    unified_strdate,
 )
 
 
@@ -13,36 +16,69 @@ class WatIE(InfoExtractor):
     IE_NAME = 'wat.tv'
     _TEST = {
         u'url': u'http://www.wat.tv/video/world-war-philadelphia-vost-6bv55_2fjr7_.html',
-        u'file': u'6bv55.mp4',
+        u'file': u'10631273.mp4',
         u'md5': u'0a4fe7870f31eaeabb5e25fd8da8414a',
         u'info_dict': {
-            u"title": u"World War Z - Philadelphia VOST"
+            u'title': u'World War Z - Philadelphia VOST',
+            u'description': u'La menace est partout. Que se passe-t-il à Philadelphia ?\r\nWORLD WAR Z, avec Brad Pitt, au cinéma le 3 juillet.\r\nhttp://www.worldwarz.fr',
         }
     }
     
+    def download_video_info(self, real_id):
+        # 'contentv4' is used in the website, but it also returns the related
+        # videos, we don't need them
+        info = self._download_webpage('http://www.wat.tv/interface/contentv3/' + real_id, real_id, 'Downloading video info')
+        info = json.loads(info)
+        return info['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)
         short_id = mobj.group('shortID')
+        webpage = self._download_webpage(url, short_id)
+        real_id = self._search_regex(r'xtpage = ".*-(.*?)";', webpage, 'real id')
 
+        video_info = self.download_video_info(real_id)
+        chapters = video_info['chapters']
+        first_chapter = chapters[0]
+
+        if real_id_for_chapter(first_chapter) != real_id:
+            self.to_screen('Multipart video detected')
+            chapter_urls = []
+            for chapter in chapters:
+                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'])
+
+        # Otherwise we can continue and extract just one part, we have to use
+        # the short id for getting the video url
         player_data = compat_urllib_parse.urlencode({'shortVideoId': short_id,
                                                      'html5': '1'})
         player_info = self._download_webpage('http://www.wat.tv/player?' + player_data,
-                                             short_id, u'Downloading player info')
+                                             real_id, u'Downloading player info')
         player = json.loads(player_info)['player']
         html5_player = self._html_search_regex(r'iframe src="(.*?)"', player,
                                                'html5 player')
-        player_webpage = self._download_webpage(html5_player, short_id,
+        player_webpage = self._download_webpage(html5_player, real_id,
                                                 u'Downloading player webpage')
 
         video_url = self._search_regex(r'urlhtml5 : "(.*?)"', player_webpage,
                                        'video url')
-        title = self._search_regex(r'contentTitle : "(.*?)"', player_webpage,
+        info = {'id': real_id,
-                                   'title')
-        thumbnail = self._search_regex(r'previewMedia : "(.*?)"', player_webpage,
-                                       'thumbnail')
-        return {'id': short_id,
                 'url': video_url,
                 'ext': 'mp4',
-                'title': title,
+                'title': first_chapter['title'],
-                'thumbnail': thumbnail,
+                'thumbnail': first_chapter['preview'],
+                'description': first_chapter['description'],
+                'view_count': video_info['views'],
                 }
+        if 'date_diffusion' in first_chapter:
+            info['upload_date'] = unified_strdate(first_chapter['date_diffusion'])
+
+        return info