implement ICapVideoProvider.get_video()

weboob/backends/youtube/backend.py

@@ -18,6 +18,8 @@ Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
 
 """
 
+import re
+
 from weboob.backend import BaseBackend
 from weboob.capabilities.video import ICapVideoProvider
 
@@ -44,7 +46,7 @@ class YoutubeBackend(BaseBackend, ICapVideoProvider):
     def need_url(func):
         def inner(self, *args, **kwargs):
             url = args[0]
-            if isinstance(url, (str,unicode)) and not url.isdigit() and u'youtube.com' not in url:
+            if (u'youtube.com' not in url) and not re.match('^\w+$', url):
                 return None
             return func(self, *args, **kwargs)
         return inner

weboob/backends/youtube/browser.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 
 """
-Copyright(C) 2010  Christophe Benz
+Copyright(C) 2010  Christophe Benz, Romain Bignon
 
 This program is free software; you can redistribute it and/or modify
 it under the terms of the GNU General Public License as published by
@@ -49,6 +49,15 @@ class YoutubeBrowser(BaseBrowser):
         assert self.is_on_page(ResultsPage)
         return self.page.iter_videos()
 
+    def get_video(self, _id):
+        if re.match('^\w+$', _id):
+            url = 'http://www.youtube.com/watch?v=%s' % _id
+        else:
+            url = _id
+
+        self.location(url)
+        return self.page.video
+
     def get_video_title(self, page_url):
         self.location(page_url)
         return self.page.title

weboob/backends/youtube/pages/results.py

@@ -24,7 +24,7 @@ from weboob.tools.browser import BasePage
 from weboob.capabilities.video import Video
 
 class ResultsPage(BasePage):
-    WATCH_RE = re.compile('/watch?v=(\w+)')
+    WATCH_RE = re.compile('/watch\?v=(\w+)')
     def iter_videos(self):
         for div in self.document.getroot().cssselect("div[class^=video-entry]"):
             a = div.find('a')

weboob/backends/youtube/pages/video.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 
 """
-Copyright(C) 2010  Christophe Benz
+Copyright(C) 2010  Christophe Benz, Romain Bignon
 
 This program is free software; you can redistribute it and/or modify
 it under the terms of the GNU General Public License as published by
@@ -18,13 +18,53 @@ Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
 
 """
 
+import re
+from logging import warning
+
 from weboob.tools.browser import BasePage
+from weboob.capabilities.video import Video
 
 class VideoPage(BasePage):
+    URL_REGEXP = re.compile("https?://[w\.]*youtube.com/watch\?v=(\w+)")
+
     def on_loaded(self):
+        self.video = Video(self.get_id())
+        self.video.title = self.get_title()
+        self.video.url = self.get_url()
+
+        self.set_details(self.video)
+
+    def get_id(self):
+        m = self.URL_REGEXP.match(self.url)
+        if m:
+            return m.group(1)
+        warning("Unable to parse ID")
+        return 0
+
+    VIDEO_SIGNATURE_RE = re.compile(r'&t=([^ ,&]*)')
+    def get_url(self):
+        video_signature = None
+        for data in self.document.getiterator('script'):
+            if not data.text:
+                continue
+            for m in re.finditer(self.VIDEO_SIGNATURE_RE, data.text):
+                video_signature = m.group(1)
+        return 'http://www.youtube.com/get_video?video_id=%s&t=%s&fmt=18' % (self.video.id, video_signature)
+
+    def get_title(self):
         found = self.document.getroot().cssselect('meta[name=title]')
         if found:
             content = found[0].attrib['content']
-            self.title = unicode(content).strip()
+            return unicode(content).strip()
-        else:
+        return u''
-            self.title = None
+
+    DATE_REGEXP = re.compile("\w+ (\w+) (\d+) (\d+):(\d+):(\d+) (\d+)")
+    MONTH2I = ['', 'Jan', 'Feb', 'Mar', 'Apr', 'May', 'Jun', 'Jul', 'Aug', 'Sep', 'Oct', 'Nov', 'Dec']
+
+    def set_details(self, v):
+        div = self.document.getroot().cssselect('div[id=watch-description-body]')
+        if not div:
+            return
+
+        div = div[0]
+        v.author = div.find('a').find('strong').text