youtube-dl

commit ffbc3901d276a4fafc9423c51dc2c6aa5d108d91
parent 482a1258de6af0a15b6e7859d244f9125cadef47
Author: Yen Chi Hsuan <yan12125@gmail.com>
Date:   Thu, 30 Apr 2015 23:33:49 +0800

Merge remote-tracking branch 'upstream/master'

Diffstat:
M
test/test_YoutubeDL.py
 | 
4
++--
M
youtube_dl/YoutubeDL.py
 | 
11
++++++-----
M
youtube_dl/extractor/vporn.py
 | 
30
+++++++++++-------------------

3 files changed, 19 insertions(+), 26 deletions(-)
diff --git a/test/test_YoutubeDL.py b/test/test_YoutubeDL.py
@@ -237,7 +237,7 @@ class TestFormatSelection(unittest.TestCase):
             f2['url'] = 'url:' + f2id
 
             info_dict = _make_result([f1, f2], extractor='youtube')
-            ydl = YDL()
+            ydl = YDL({'format': 'best/bestvideo'})
             yie = YoutubeIE(ydl)
             yie._sort_formats(info_dict['formats'])
             ydl.process_ie_result(info_dict)
@@ -245,7 +245,7 @@ class TestFormatSelection(unittest.TestCase):
             self.assertEqual(downloaded['format_id'], f1id)
 
             info_dict = _make_result([f2, f1], extractor='youtube')
-            ydl = YDL()
+            ydl = YDL({'format': 'best/bestvideo'})
             yie = YoutubeIE(ydl)
             yie._sort_formats(info_dict['formats'])
             ydl.process_ie_result(info_dict)
diff --git a/youtube_dl/YoutubeDL.py b/youtube_dl/YoutubeDL.py
@@ -914,15 +914,16 @@ class YoutubeDL(object):
         if not available_formats:
             return None
 
-        if format_spec == 'best' or format_spec is None:
-            return available_formats[-1]
-        elif format_spec == 'worst':
+        if format_spec in ['best', 'worst', None]:
+            format_idx = 0 if format_spec == 'worst' else -1
             audiovideo_formats = [
                 f for f in available_formats
                 if f.get('vcodec') != 'none' and f.get('acodec') != 'none']
             if audiovideo_formats:
-                return audiovideo_formats[0]
-            return available_formats[0]
+                return audiovideo_formats[format_idx]
+            # for audio only urls, select the best/worst audio format
+            elif all(f.get('acodec') != 'none' for f in available_formats):
+                return available_formats[format_idx]
         elif format_spec == 'bestaudio':
             audio_formats = [
                 f for f in available_formats
diff --git a/youtube_dl/extractor/vporn.py b/youtube_dl/extractor/vporn.py
@@ -27,9 +27,6 @@ class VpornIE(InfoExtractor):
                 'duration': 393,
                 'age_limit': 18,
                 'view_count': int,
-                'like_count': int,
-                'dislike_count': int,
-                'comment_count': int,
             }
         },
         {
@@ -47,9 +44,6 @@ class VpornIE(InfoExtractor):
                 'duration': 588,
                 'age_limit': 18,
                 'view_count': int,
-                'like_count': int,
-                'dislike_count': int,
-                'comment_count': int,
             }
         },
     ]
@@ -64,29 +58,29 @@ class VpornIE(InfoExtractor):
         title = self._html_search_regex(
             r'videoname\s*=\s*\'([^\']+)\'', webpage, 'title').strip()
         description = self._html_search_regex(
-            r'<div class="description_txt">(.*?)</div>', webpage, 'description', fatal=False)
+            r'class="(?:descr|description_txt)">(.*?)</div>',
+            webpage, 'description', fatal=False)
         thumbnail = self._html_search_regex(
             r'flashvars\.imageUrl\s*=\s*"([^"]+)"', webpage, 'description', fatal=False, default=None)
         if thumbnail:
             thumbnail = 'http://www.vporn.com' + thumbnail
 
         uploader = self._html_search_regex(
-            r'(?s)UPLOADED BY.*?<a href="/user/[^"]+">([^<]+)</a>',
+            r'(?s)Uploaded by:.*?<a href="/user/[^"]+"[^>]*>(.+?)</a>',
             webpage, 'uploader', fatal=False)
 
-        categories = re.findall(r'<a href="/cat/[^"]+">([^<]+)</a>', webpage)
+        categories = re.findall(r'<a href="/cat/[^"]+"[^>]*>([^<]+)</a>', webpage)
 
         duration = parse_duration(self._search_regex(
-            r'duration (\d+ min \d+ sec)', webpage, 'duration', fatal=False))
+            r'Runtime:\s*</span>\s*(\d+ min \d+ sec)',
+            webpage, 'duration', fatal=False))
 
-        view_count = str_to_int(self._html_search_regex(
-            r'<span>([\d,\.]+) VIEWS</span>', webpage, 'view count', fatal=False))
-        like_count = str_to_int(self._html_search_regex(
-            r'<span id="like" class="n">([\d,\.]+)</span>', webpage, 'like count', fatal=False))
-        dislike_count = str_to_int(self._html_search_regex(
-            r'<span id="dislike" class="n">([\d,\.]+)</span>', webpage, 'dislike count', fatal=False))
+        view_count = str_to_int(self._search_regex(
+            r'class="views">([\d,\.]+) [Vv]iews<',
+            webpage, 'view count', fatal=False))
         comment_count = str_to_int(self._html_search_regex(
-            r'<h4>Comments \(<b>([\d,\.]+)</b>\)</h4>', webpage, 'comment count', fatal=False))
+            r"'Comments \(([\d,\.]+)\)'",
+            webpage, 'comment count', default=None))
 
         formats = []
 
@@ -117,8 +111,6 @@ class VpornIE(InfoExtractor):
             'categories': categories,
             'duration': duration,
             'view_count': view_count,
-            'like_count': like_count,
-            'dislike_count': dislike_count,
             'comment_count': comment_count,
             'age_limit': 18,
             'formats': formats,