[cnn] Add multiple formats, duration, and upload_date

test/test_utils.py

@@ -18,6 +18,7 @@ from youtube_dl.utils import (
     find_xpath_attr,
     get_meta_content,
     orderedSet,
+    parse_duration,
     sanitize_filename,
     shell_quote,
     smuggle_url,
@@ -192,5 +193,12 @@ class TestUtil(unittest.TestCase):
             url_basename(u'http://media.w3.org/2010/05/sintel/trailer.mp4'),
             u'trailer.mp4')
 
+    def test_parse_duration(self):
+        self.assertEqual(parse_duration(None), None)
+        self.assertEqual(parse_duration('1'), 1)
+        self.assertEqual(parse_duration('1337:12'), 80232)
+        self.assertEqual(parse_duration('9:12:43'), 33163)
+        self.assertEqual(parse_duration('x:y'), None)
+
 if __name__ == '__main__':
     unittest.main()

youtube_dl/extractor/cnn.py

@@ -1,7 +1,10 @@
 import re
 
 from .common import InfoExtractor
-from ..utils import determine_ext
+from ..utils import (
+    int_or_none,
+    parse_duration,
+)
 
 
 class CNNIE(InfoExtractor):
@@ -15,6 +18,8 @@ class CNNIE(InfoExtractor):
         u'info_dict': {
             u'title': u'Nadal wins 8th French Open title',
             u'description': u'World Sport\'s Amanda Davies chats with 2013 French Open champion Rafael Nadal.',
+            u'duration': 135,
+            u'upload_date': u'20130609',
         },
     },
     {
@@ -35,22 +40,58 @@ class CNNIE(InfoExtractor):
         info = self._download_xml(info_url, page_title)
 
         formats = []
+        rex = re.compile(r'''(?x)
+            (?P<width>[0-9]+)x(?P<height>[0-9]+)
+            (?:_(?P<bitrate>[0-9]+)k)?
+        ''')
         for f in info.findall('files/file'):
-            mf = re.match(r'(\d+)x(\d+)(?:_(.*)k)?',f.attrib['bitrate'])
-            if mf is not None:
-                formats.append((int(mf.group(1)), int(mf.group(2)), int(mf.group(3) or 0), f.text))
-        formats = sorted(formats)
-        (_,_,_, video_path) = formats[-1]
-        video_url = 'http://ht.cdn.turner.com/cnn/big%s' % video_path
+            video_url = 'http://ht.cdn.turner.com/cnn/big%s' % (f.text.strip())
+            fdct = {
+                'format_id': f.attrib['bitrate'],
+                'url': video_url,
+            }
+
+            mf = rex.match(f.attrib['bitrate'])
+            if mf:
+                fdct['width'] = int(mf.group('width'))
+                fdct['height'] = int(mf.group('height'))
+                fdct['tbr'] = int_or_none(mf.group('bitrate'))
+            else:
+                mf = rex.search(f.text)
+                if mf:
+                    fdct['width'] = int(mf.group('width'))
+                    fdct['height'] = int(mf.group('height'))
+                    fdct['tbr'] = int_or_none(mf.group('bitrate'))
+                else:
+                    mi = re.match(r'ios_(audio|[0-9]+)$', f.attrib['bitrate'])
+                    if mi:
+                        if mi.group(1) == 'audio':
+                            fdct['vcodec'] = 'none'
+                            fdct['ext'] = 'm4a'
+                        else:
+                            fdct['tbr'] = int(mi.group(1))
+
+            formats.append(fdct)
+
+        self._sort_formats(formats)
 
         thumbnails = sorted([((int(t.attrib['height']),int(t.attrib['width'])), t.text) for t in info.findall('images/image')])
         thumbs_dict = [{'resolution': res, 'url': t_url} for (res, t_url) in thumbnails]
 
-        return {'id': info.attrib['id'],
-                'title': info.find('headline').text,
-                'url': video_url,
-                'ext': determine_ext(video_url),
-                'thumbnail': thumbnails[-1][1],
-                'thumbnails': thumbs_dict,
-                'description': info.find('description').text,
-                }
+        metas_el = info.find('metas')
+        upload_date = (
+            metas_el.attrib.get('version') if metas_el is not None else None)
+
+        duration_el = info.find('length')
+        duration = parse_duration(duration_el.text)
+
+        return {
+            'id': info.attrib['id'],
+            'title': info.find('headline').text,
+            'formats': formats,
+            'thumbnail': thumbnails[-1][1],
+            'thumbnails': thumbs_dict,
+            'description': info.find('description').text,
+            'duration': duration,
+            'upload_date': upload_date,
+        }

youtube_dl/utils.py

@@ -1102,3 +1102,19 @@ class HEADRequest(compat_urllib_request.Request):
 
 def int_or_none(v):
     return v if v is None else int(v)
+
+
+def parse_duration(s):
+    if s is None:
+        return None
+
+    m = re.match(
+        r'(?:(?:(?P<hours>[0-9]+):)?(?P<mins>[0-9]+):)?(?P<secs>[0-9]+)$', s)
+    if not m:
+        return None
+    res = int(m.group('secs'))
+    if m.group('mins'):
+        res += int(m.group('mins')) * 60
+        if m.group('hours'):
+            res += int(m.group('hours')) * 60 * 60
+    return res