release 2014.01.22.4

Conflicts:
	test/test_utils.py
	youtube_dl/extractor/youtube.py

README.md

@@ -158,7 +158,9 @@ which means you can modify it, redistribute it or use it however you like.
 ## Video Format Options:
     -f, --format FORMAT        video format code, specify the order of
                                preference using slashes: "-f 22/17/18". "-f mp4"
-                               and "-f flv" are also supported
+                               and "-f flv" are also supported. You can also use
+                               the special names "best", "bestaudio", "worst",
+                               and "worstaudio"
     --all-formats              download all available video formats
     --prefer-free-formats      prefer free video formats unless a specific one
                                is requested

test/test_YoutubeDL.py

@@ -1,5 +1,7 @@
 #!/usr/bin/env python
 
+from __future__ import unicode_literals
+
 # Allow direct execution
 import os
 import sys
@@ -30,125 +32,155 @@ class TestFormatSelection(unittest.TestCase):
         ydl = YDL()
         ydl.params['prefer_free_formats'] = True
         formats = [
-            {u'ext': u'webm', u'height': 460},
-            {u'ext': u'mp4',  u'height': 460},
+            {'ext': 'webm', 'height': 460},
+            {'ext': 'mp4',  'height': 460},
         ]
-        info_dict = {u'formats': formats, u'extractor': u'test'}
+        info_dict = {'formats': formats, 'extractor': 'test'}
         yie = YoutubeIE(ydl)
         yie._sort_formats(info_dict['formats'])
         ydl.process_ie_result(info_dict)
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded[u'ext'], u'webm')
+        self.assertEqual(downloaded['ext'], 'webm')
 
         # Different resolution => download best quality (mp4)
         ydl = YDL()
         ydl.params['prefer_free_formats'] = True
         formats = [
-            {u'ext': u'webm', u'height': 720},
-            {u'ext': u'mp4', u'height': 1080},
+            {'ext': 'webm', 'height': 720},
+            {'ext': 'mp4', 'height': 1080},
         ]
-        info_dict[u'formats'] = formats
+        info_dict['formats'] = formats
         yie = YoutubeIE(ydl)
         yie._sort_formats(info_dict['formats'])
         ydl.process_ie_result(info_dict)
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded[u'ext'], u'mp4')
+        self.assertEqual(downloaded['ext'], 'mp4')
 
         # No prefer_free_formats => prefer mp4 and flv for greater compatibilty
         ydl = YDL()
         ydl.params['prefer_free_formats'] = False
         formats = [
-            {u'ext': u'webm', u'height': 720},
-            {u'ext': u'mp4', u'height': 720},
-            {u'ext': u'flv', u'height': 720},
+            {'ext': 'webm', 'height': 720},
+            {'ext': 'mp4', 'height': 720},
+            {'ext': 'flv', 'height': 720},
         ]
-        info_dict[u'formats'] = formats
+        info_dict['formats'] = formats
         yie = YoutubeIE(ydl)
         yie._sort_formats(info_dict['formats'])
         ydl.process_ie_result(info_dict)
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded[u'ext'], u'mp4')
+        self.assertEqual(downloaded['ext'], 'mp4')
 
         ydl = YDL()
         ydl.params['prefer_free_formats'] = False
         formats = [
-            {u'ext': u'flv', u'height': 720},
-            {u'ext': u'webm', u'height': 720},
+            {'ext': 'flv', 'height': 720},
+            {'ext': 'webm', 'height': 720},
         ]
-        info_dict[u'formats'] = formats
+        info_dict['formats'] = formats
         yie = YoutubeIE(ydl)
         yie._sort_formats(info_dict['formats'])
         ydl.process_ie_result(info_dict)
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded[u'ext'], u'flv')
+        self.assertEqual(downloaded['ext'], 'flv')
 
     def test_format_limit(self):
         formats = [
-            {u'format_id': u'meh', u'url': u'http://example.com/meh', 'preference': 1},
-            {u'format_id': u'good', u'url': u'http://example.com/good', 'preference': 2},
-            {u'format_id': u'great', u'url': u'http://example.com/great', 'preference': 3},
-            {u'format_id': u'excellent', u'url': u'http://example.com/exc', 'preference': 4},
+            {'format_id': 'meh', 'url': 'http://example.com/meh', 'preference': 1},
+            {'format_id': 'good', 'url': 'http://example.com/good', 'preference': 2},
+            {'format_id': 'great', 'url': 'http://example.com/great', 'preference': 3},
+            {'format_id': 'excellent', 'url': 'http://example.com/exc', 'preference': 4},
         ]
         info_dict = {
-            u'formats': formats, u'extractor': u'test', 'id': 'testvid'}
+            'formats': formats, 'extractor': 'test', 'id': 'testvid'}
 
         ydl = YDL()
         ydl.process_ie_result(info_dict)
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded[u'format_id'], u'excellent')
+        self.assertEqual(downloaded['format_id'], 'excellent')
 
         ydl = YDL({'format_limit': 'good'})
         assert ydl.params['format_limit'] == 'good'
         ydl.process_ie_result(info_dict.copy())
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded[u'format_id'], u'good')
+        self.assertEqual(downloaded['format_id'], 'good')
 
         ydl = YDL({'format_limit': 'great', 'format': 'all'})
         ydl.process_ie_result(info_dict.copy())
-        self.assertEqual(ydl.downloaded_info_dicts[0][u'format_id'], u'meh')
-        self.assertEqual(ydl.downloaded_info_dicts[1][u'format_id'], u'good')
-        self.assertEqual(ydl.downloaded_info_dicts[2][u'format_id'], u'great')
+        self.assertEqual(ydl.downloaded_info_dicts[0]['format_id'], 'meh')
+        self.assertEqual(ydl.downloaded_info_dicts[1]['format_id'], 'good')
+        self.assertEqual(ydl.downloaded_info_dicts[2]['format_id'], 'great')
         self.assertTrue('3' in ydl.msgs[0])
 
         ydl = YDL()
         ydl.params['format_limit'] = 'excellent'
         ydl.process_ie_result(info_dict.copy())
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded[u'format_id'], u'excellent')
+        self.assertEqual(downloaded['format_id'], 'excellent')
 
     def test_format_selection(self):
         formats = [
-            {u'format_id': u'35', u'ext': u'mp4', 'preference': 1},
-            {u'format_id': u'45', u'ext': u'webm', 'preference': 2},
-            {u'format_id': u'47', u'ext': u'webm', 'preference': 3},
-            {u'format_id': u'2', u'ext': u'flv', 'preference': 4},
+            {'format_id': '35', 'ext': 'mp4', 'preference': 1},
+            {'format_id': '45', 'ext': 'webm', 'preference': 2},
+            {'format_id': '47', 'ext': 'webm', 'preference': 3},
+            {'format_id': '2', 'ext': 'flv', 'preference': 4},
         ]
-        info_dict = {u'formats': formats, u'extractor': u'test'}
+        info_dict = {'formats': formats, 'extractor': 'test'}
 
-        ydl = YDL({'format': u'20/47'})
+        ydl = YDL({'format': '20/47'})
         ydl.process_ie_result(info_dict.copy())
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded['format_id'], u'47')
+        self.assertEqual(downloaded['format_id'], '47')
 
-        ydl = YDL({'format': u'20/71/worst'})
+        ydl = YDL({'format': '20/71/worst'})
         ydl.process_ie_result(info_dict.copy())
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded['format_id'], u'35')
+        self.assertEqual(downloaded['format_id'], '35')
 
         ydl = YDL()
         ydl.process_ie_result(info_dict.copy())
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded['format_id'], u'2')
+        self.assertEqual(downloaded['format_id'], '2')
 
-        ydl = YDL({'format': u'webm/mp4'})
+        ydl = YDL({'format': 'webm/mp4'})
         ydl.process_ie_result(info_dict.copy())
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded['format_id'], u'47')
+        self.assertEqual(downloaded['format_id'], '47')
 
-        ydl = YDL({'format': u'3gp/40/mp4'})
+        ydl = YDL({'format': '3gp/40/mp4'})
         ydl.process_ie_result(info_dict.copy())
         downloaded = ydl.downloaded_info_dicts[0]
-        self.assertEqual(downloaded['format_id'], u'35')
+        self.assertEqual(downloaded['format_id'], '35')
+
+    def test_format_selection_audio(self):
+        formats = [
+            {'format_id': 'audio-low', 'ext': 'webm', 'preference': 1, 'vcodec': 'none'},
+            {'format_id': 'audio-mid', 'ext': 'webm', 'preference': 2, 'vcodec': 'none'},
+            {'format_id': 'audio-high', 'ext': 'flv', 'preference': 3, 'vcodec': 'none'},
+            {'format_id': 'vid', 'ext': 'mp4', 'preference': 4},
+        ]
+        info_dict = {'formats': formats, 'extractor': 'test'}
+
+        ydl = YDL({'format': 'bestaudio'})
+        ydl.process_ie_result(info_dict.copy())
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'audio-high')
+
+        ydl = YDL({'format': 'worstaudio'})
+        ydl.process_ie_result(info_dict.copy())
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'audio-low')
+
+        formats = [
+            {'format_id': 'vid-low', 'ext': 'mp4', 'preference': 1},
+            {'format_id': 'vid-high', 'ext': 'mp4', 'preference': 2},
+        ]
+        info_dict = {'formats': formats, 'extractor': 'test'}
+
+        ydl = YDL({'format': 'bestaudio/worstaudio/best'})
+        ydl.process_ie_result(info_dict.copy())
+        downloaded = ydl.downloaded_info_dicts[0]
+        self.assertEqual(downloaded['format_id'], 'vid-high')
 
     def test_youtube_format_selection(self):
         order = [
@@ -200,17 +232,17 @@ class TestFormatSelection(unittest.TestCase):
 
     def test_prepare_filename(self):
         info = {
-            u'id': u'1234',
-            u'ext': u'mp4',
-            u'width': None,
+            'id': '1234',
+            'ext': 'mp4',
+            'width': None,
         }
         def fname(templ):
             ydl = YoutubeDL({'outtmpl': templ})
             return ydl.prepare_filename(info)
-        self.assertEqual(fname(u'%(id)s.%(ext)s'), u'1234.mp4')
-        self.assertEqual(fname(u'%(id)s-%(width)s.%(ext)s'), u'1234-NA.mp4')
+        self.assertEqual(fname('%(id)s.%(ext)s'), '1234.mp4')
+        self.assertEqual(fname('%(id)s-%(width)s.%(ext)s'), '1234-NA.mp4')
         # Replace missing fields with 'NA'
-        self.assertEqual(fname(u'%(uploader_date)s-%(id)s.%(ext)s'), u'NA-1234.mp4')
+        self.assertEqual(fname('%(uploader_date)s-%(id)s.%(ext)s'), 'NA-1234.mp4')
 
 
 if __name__ == '__main__':

test/test_utils.py

@@ -19,6 +19,7 @@ from youtube_dl.utils import (
     fix_xml_ampersands,
     get_meta_content,
     orderedSet,
+    PagedList,
     parse_duration,
     sanitize_filename,
     shell_quote,
@@ -214,5 +215,26 @@ class TestUtil(unittest.TestCase):
             fix_xml_ampersands('Ӓ᪼'), 'Ӓ᪼')
         self.assertEqual(fix_xml_ampersands('&#&#'), '&#&#')
 
+    def test_paged_list(self):
+        def testPL(size, pagesize, sliceargs, expected):
+            def get_page(pagenum):
+                firstid = pagenum * pagesize
+                upto = min(size, pagenum * pagesize + pagesize)
+                for i in range(firstid, upto):
+                    yield i
+
+            pl = PagedList(get_page, pagesize)
+            got = pl.getslice(*sliceargs)
+            self.assertEqual(got, expected)
+
+        testPL(5, 2, (), [0, 1, 2, 3, 4])
+        testPL(5, 2, (1,), [1, 2, 3, 4])
+        testPL(5, 2, (2,), [2, 3, 4])
+        testPL(5, 2, (4,), [4])
+        testPL(5, 2, (0, 3), [0, 1, 2])
+        testPL(5, 2, (1, 4), [1, 2, 3])
+        testPL(5, 2, (2, 99), [2, 3, 4])
+        testPL(5, 2, (20, 99), [])
+
 if __name__ == '__main__':
     unittest.main()

youtube_dl/YoutubeDL.py

@@ -39,6 +39,7 @@ from .utils import (
     locked_file,
     make_HTTPS_handler,
     MaxDownloadsReached,
+    PagedList,
     PostProcessingError,
     platform_name,
     preferredencoding,
@@ -578,19 +579,27 @@ class YoutubeDL(object):
 
             playlist_results = []
 
-            n_all_entries = len(ie_result['entries'])
             playliststart = self.params.get('playliststart', 1) - 1
             playlistend = self.params.get('playlistend', None)
             # For backwards compatibility, interpret -1 as whole list
             if playlistend == -1:
                 playlistend = None
 
-            entries = ie_result['entries'][playliststart:playlistend]
-            n_entries = len(entries)
-
-            self.to_screen(
-                "[%s] playlist '%s': Collected %d video ids (downloading %d of them)" %
-                (ie_result['extractor'], playlist, n_all_entries, n_entries))
+            if isinstance(ie_result['entries'], list):
+                n_all_entries = len(ie_result['entries'])
+                entries = ie_result['entries'][playliststart:playlistend]
+                n_entries = len(entries)
+                self.to_screen(
+                    "[%s] playlist %s: Collected %d video ids (downloading %d of them)" %
+                    (ie_result['extractor'], playlist, n_all_entries, n_entries))
+            else:
+                assert isinstance(ie_result['entries'], PagedList)
+                entries = ie_result['entries'].getslice(
+                    playliststart, playlistend)
+                n_entries = len(entries)
+                self.to_screen(
+                    "[%s] playlist %s: Downloading %d videos" %
+                    (ie_result['extractor'], playlist, n_entries))
 
             for i, entry in enumerate(entries, 1):
                 self.to_screen('[download] Downloading video #%s of %s' % (i, n_entries))
@@ -637,6 +646,18 @@ class YoutubeDL(object):
             return available_formats[-1]
         elif format_spec == 'worst':
             return available_formats[0]
+        elif format_spec == 'bestaudio':
+            audio_formats = [
+                f for f in available_formats
+                if f.get('vcodec') == 'none']
+            if audio_formats:
+                return audio_formats[-1]
+        elif format_spec == 'worstaudio':
+            audio_formats = [
+                f for f in available_formats
+                if f.get('vcodec') == 'none']
+            if audio_formats:
+                return audio_formats[0]
         else:
             extensions = ['mp4', 'flv', 'webm', '3gp']
             if format_spec in extensions:
@@ -701,7 +722,7 @@ class YoutubeDL(object):
             self.list_formats(info_dict)
             return
 
-        req_format = self.params.get('format', 'best')
+        req_format = self.params.get('format')
         if req_format is None:
             req_format = 'best'
         formats_to_download = []
@@ -1094,6 +1115,8 @@ class YoutubeDL(object):
                 res += 'audio'
             if fdict.get('abr') is not None:
                 res += '@%3dk' % fdict['abr']
+            if fdict.get('asr') is not None:
+                res += ' (%5dHz)' % fdict['asr']
             if fdict.get('filesize') is not None:
                 if res:
                     res += ', '

youtube_dl/__init__.py

@@ -256,8 +256,8 @@ def parseOpts(overrideArguments=None):
 
 
     video_format.add_option('-f', '--format',
-            action='store', dest='format', metavar='FORMAT', default='best',
-            help='video format code, specify the order of preference using slashes: "-f 22/17/18". "-f mp4" and "-f flv" are also supported')
+            action='store', dest='format', metavar='FORMAT', default=None,
+            help='video format code, specify the order of preference using slashes: "-f 22/17/18". "-f mp4" and "-f flv" are also supported. You can also use the special names "best", "bestaudio", "worst", and "worstaudio"')
     video_format.add_option('--all-formats',
             action='store_const', dest='format', help='download all available video formats', const='all')
     video_format.add_option('--prefer-free-formats',
@@ -624,6 +624,10 @@ def _real_main(argv=None):
     if opts.default_search not in ('auto', None) and ':' not in opts.default_search:
         parser.error(u'--default-search invalid; did you forget a colon (:) at the end?')
 
+    # Do not download videos when there are audio-only formats
+    if opts.extractaudio and not opts.keepvideo and opts.format is None:
+        opts.format = 'bestaudio/best'
+
     # --all-sub automatically sets --write-sub if --write-auto-sub is not given
     # this was the old behaviour if only --all-sub was given.
     if opts.allsubtitles and (opts.writeautomaticsub == False):

youtube_dl/extractor/common.py

@@ -63,6 +63,7 @@ class InfoExtractor(object):
                     * tbr        Average bitrate of audio and video in KBit/s
                     * abr        Average audio bitrate in KBit/s
                     * acodec     Name of the audio codec in use
+                    * asr        Audio sampling rate in Hertz
                     * vbr        Average video bitrate in KBit/s
                     * vcodec     Name of the video codec in use
                     * filesize   The number of bytes, if known in advance

youtube_dl/extractor/cspan.py

@@ -20,6 +20,7 @@ class CSpanIE(InfoExtractor):
             'title': 'Attorney General Eric Holder on Voting Rights Act Decision',
             'description': 'Attorney General Eric Holder spoke to reporters following the Supreme Court decision in [Shelby County v. Holder] in which the court ruled that the preclearance provisions of the Voting Rights Act could not be enforced until Congress established new guidelines for review.',
         },
+        'skip': 'Regularly fails on travis, for unknown reasons',
     }
 
     def _real_extract(self, url):

youtube_dl/extractor/xvideos.py

@@ -1,3 +1,5 @@
+from __future__ import unicode_literals
+
 import re
 
 from .common import InfoExtractor
@@ -9,12 +11,12 @@ from ..utils import (
 class XVideosIE(InfoExtractor):
     _VALID_URL = r'^(?:https?://)?(?:www\.)?xvideos\.com/video([0-9]+)(?:.*)'
     _TEST = {
-        u'url': u'http://www.xvideos.com/video939581/funny_porns_by_s_-1',
-        u'file': u'939581.flv',
-        u'md5': u'1d0c835822f0a71a7bf011855db929d0',
-        u'info_dict': {
-            u"title": u"Funny Porns By >>>>S<<<<<< -1",
-            u"age_limit": 18,
+        'url': 'http://www.xvideos.com/video939581/funny_porns_by_s_-1',
+        'file': '939581.flv',
+        'md5': '1d0c835822f0a71a7bf011855db929d0',
+        'info_dict': {
+            "title": "Funny Porns By >>>>S<<<<<< -1",
+            "age_limit": 18,
         }
     }
 
@@ -27,18 +29,18 @@ class XVideosIE(InfoExtractor):
         self.report_extraction(video_id)
 
         # Extract video URL
-        video_url = compat_urllib_parse.unquote(self._search_regex(r'flv_url=(.+?)&',
-            webpage, u'video URL'))
+        video_url = compat_urllib_parse.unquote(
+            self._search_regex(r'flv_url=(.+?)&', webpage, 'video URL'))
 
         # Extract title
-        video_title = self._html_search_regex(r'<title>(.*?)\s+-\s+XVID',
-            webpage, u'title')
+        video_title = self._html_search_regex(
+            r'<title>(.*?)\s+-\s+XVID', webpage, 'title')
 
         # Extract video thumbnail
-        video_thumbnail = self._search_regex(r'http://(?:img.*?\.)xvideos.com/videos/thumbs/[a-fA-F0-9]+/[a-fA-F0-9]+/[a-fA-F0-9]+/[a-fA-F0-9]+/([a-fA-F0-9.]+jpg)',
-            webpage, u'thumbnail', fatal=False)
+        video_thumbnail = self._search_regex(
+            r'url_bigthumb=(.+?)&amp', webpage, 'thumbnail', fatal=False)
 
-        info = {
+        return {
             'id': video_id,
             'url': video_url,
             'uploader': None,
@@ -49,5 +51,3 @@ class XVideosIE(InfoExtractor):
             'description': None,
             'age_limit': 18,
         }
-
-        return [info]

youtube_dl/extractor/youtube.py

@@ -27,6 +27,8 @@ from ..utils import (
     get_element_by_id,
     get_element_by_attribute,
     ExtractorError,
+    int_or_none,
+    PagedList,
     RegexNotFoundError,
     unescapeHTML,
     unified_strdate,
@@ -270,6 +272,21 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
                 u"uploader_id": u"setindia"
             }
         },
+        {
+            u"url": u"http://www.youtube.com/watch?v=a9LDPn-MO4I",
+            u"file": u"a9LDPn-MO4I.m4a",
+            u"note": u"256k DASH audio (format 141) via DASH manifest",
+            u"params": {
+                u"format": "141"
+            },
+            u"info_dict": {
+                u"upload_date": "20121002",
+                u"uploader_id": "8KVIDEO",
+                u"description": "No description available.",
+                u"uploader": "8KVIDEO",
+                u"title": "UHDTV TEST 8K VIDEO.mp4"
+            }
+        },
     ]
 
 
@@ -1067,18 +1084,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
         video_id = mobj.group(2)
         return video_id
 
-    def _get_video_url_list(self, url_map):
-        """
-        Transform a dictionary in the format {itag:url} to a list of (itag, url)
-        with the requested formats.
-        """
-        existing_formats = [x for x in self._formats if x in url_map]
-        if len(existing_formats) == 0:
-            raise ExtractorError(u'no known formats available for video')
-        video_url_list = [(f, url_map[f]) for f in existing_formats] # All formats
-        video_url_list.reverse() # order worst to best
-        return video_url_list
-
     def _extract_from_m3u8(self, manifest_url, video_id):
         url_map = {}
         def _get_urls(_manifest):
@@ -1252,7 +1257,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
                 video_annotations = self._extract_annotations(video_id)
 
         # Decide which formats to download
-
         try:
             mobj = re.search(r';ytplayer.config = ({.*?});', video_webpage)
             if not mobj:
@@ -1277,9 +1281,26 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
         except ValueError:
             pass
 
+        def _map_to_format_list(urlmap):
+            formats = []
+            for itag, video_real_url in urlmap.items():
+                dct = {
+                    'format_id': itag,
+                    'url': video_real_url,
+                    'player_url': player_url,
+                }
+                dct.update(self._formats[itag])
+                formats.append(dct)
+            return formats
+
         if 'conn' in video_info and video_info['conn'][0].startswith('rtmp'):
             self.report_rtmp_download()
-            video_url_list = [('_rtmp', video_info['conn'][0])]
+            formats = [{
+                'format_id': '_rtmp',
+                'protocol': 'rtmp',
+                'url': video_info['conn'][0],
+                'player_url': player_url,
+            }]
         elif len(video_info.get('url_encoded_fmt_stream_map', [])) >= 1 or len(video_info.get('adaptive_fmts', [])) >= 1:
             encoded_url_map = video_info.get('url_encoded_fmt_stream_map', [''])[0] + ',' + video_info.get('adaptive_fmts',[''])[0]
             if 'rtmpe%3Dyes' in encoded_url_map:
@@ -1324,23 +1345,49 @@ class YoutubeIE(YoutubeBaseInfoExtractor, SubtitlesInfoExtractor):
                     if 'ratebypass' not in url:
                         url += '&ratebypass=yes'
                     url_map[url_data['itag'][0]] = url
-            video_url_list = self._get_video_url_list(url_map)
+            formats = _map_to_format_list(url_map)
         elif video_info.get('hlsvp'):
             manifest_url = video_info['hlsvp'][0]
             url_map = self._extract_from_m3u8(manifest_url, video_id)
-            video_url_list = self._get_video_url_list(url_map)
+            formats = _map_to_format_list(url_map)
         else:
             raise ExtractorError(u'no conn, hlsvp or url_encoded_fmt_stream_map information found in video info')
 
-        formats = []
-        for itag, video_real_url in video_url_list:
-            dct = {
-                'format_id': itag,
-                'url': video_real_url,
-                'player_url': player_url,
-            }
-            dct.update(self._formats[itag])
-            formats.append(dct)
+        # Look for the DASH manifest
+        dash_manifest_url_lst = video_info.get('dashmpd')
+        if dash_manifest_url_lst and dash_manifest_url_lst[0]:
+            try:
+                dash_doc = self._download_xml(
+                    dash_manifest_url_lst[0], video_id,
+                    note=u'Downloading DASH manifest',
+                    errnote=u'Could not download DASH manifest')
+                for r in dash_doc.findall(u'.//{urn:mpeg:DASH:schema:MPD:2011}Representation'):
+                    url_el = r.find('{urn:mpeg:DASH:schema:MPD:2011}BaseURL')
+                    if url_el is None:
+                        continue
+                    format_id = r.attrib['id']
+                    video_url = url_el.text
+                    filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength'))
+                    f = {
+                        'format_id': format_id,
+                        'url': video_url,
+                        'width': int_or_none(r.attrib.get('width')),
+                        'tbr': int_or_none(r.attrib.get('bandwidth'), 1000),
+                        'asr': int_or_none(r.attrib.get('audioSamplingRate')),
+                        'filesize': filesize,
+                    }
+                    try:
+                        existing_format = next(
+                            fo for fo in formats
+                            if fo['format_id'] == format_id)
+                    except StopIteration:
+                        f.update(self._formats.get(format_id, {}))
+                        formats.append(f)
+                    else:
+                        existing_format.update(f)
+
+            except (ExtractorError, KeyError) as e:
+                self.report_warning(u'Skipping DASH manifest: %s' % e, video_id)
 
         self._sort_formats(formats)
 
@@ -1580,44 +1627,35 @@ class YoutubeUserIE(InfoExtractor):
         # page by page until there are no video ids - it means we got
         # all of them.
 
-        url_results = []
-
-        for pagenum in itertools.count(0):
+        def download_page(pagenum):
             start_index = pagenum * self._GDATA_PAGE_SIZE + 1
 
             gdata_url = self._GDATA_URL % (username, self._GDATA_PAGE_SIZE, start_index)
-            page = self._download_webpage(gdata_url, username,
-                                          u'Downloading video ids from %d to %d' % (start_index, start_index + self._GDATA_PAGE_SIZE))
+            page = self._download_webpage(
+                gdata_url, username,
+                u'Downloading video ids from %d to %d' % (
+                    start_index, start_index + self._GDATA_PAGE_SIZE))
 
             try:
                 response = json.loads(page)
             except ValueError as err:
                 raise ExtractorError(u'Invalid JSON in API response: ' + compat_str(err))
             if 'entry' not in response['feed']:
-                # Number of videos is a multiple of self._MAX_RESULTS
-                break
+                return
 
             # Extract video identifiers
             entries = response['feed']['entry']
             for entry in entries:
                 title = entry['title']['$t']
                 video_id = entry['id']['$t'].split('/')[-1]
-                url_results.append({
+                yield {
                     '_type': 'url',
                     'url': video_id,
                     'ie_key': 'Youtube',
                     'id': 'video_id',
                     'title': title,
-                })
-
-            # A little optimization - if current page is not
-            # "full", ie. does not contain PAGE_SIZE video ids then
-            # we can assume that this page is the last one - there
-            # are no more ids on further pages - no need to query
-            # again.
-
-            if len(entries) < self._GDATA_PAGE_SIZE:
-                break
+                }
+        url_results = PagedList(download_page, self._GDATA_PAGE_SIZE)
 
         return self.playlist_result(url_results, playlist_title=username)
 

youtube_dl/utils.py

@@ -6,6 +6,7 @@ import datetime
 import email.utils
 import errno
 import gzip
+import itertools
 import io
 import json
 import locale
@@ -1131,8 +1132,8 @@ class HEADRequest(compat_urllib_request.Request):
         return "HEAD"
 
 
-def int_or_none(v):
-    return v if v is None else int(v)
+def int_or_none(v, scale=1):
+    return v if v is None else (int(v) // scale)
 
 
 def parse_duration(s):
@@ -1164,3 +1165,46 @@ def check_executable(exe, args=[]):
     except OSError:
         return False
     return exe
+
+
+class PagedList(object):
+    def __init__(self, pagefunc, pagesize):
+        self._pagefunc = pagefunc
+        self._pagesize = pagesize
+
+    def getslice(self, start=0, end=None):
+        res = []
+        for pagenum in itertools.count(start // self._pagesize):
+            firstid = pagenum * self._pagesize
+            nextfirstid = pagenum * self._pagesize + self._pagesize
+            if start >= nextfirstid:
+                continue
+
+            page_results = list(self._pagefunc(pagenum))
+
+            startv = (
+                start % self._pagesize
+                if firstid <= start < nextfirstid
+                else 0)
+
+            endv = (
+                ((end - 1) % self._pagesize) + 1
+                if (end is not None and firstid <= end <= nextfirstid)
+                else None)
+
+            if startv != 0 or endv is not None:
+                page_results = page_results[startv:endv]
+            res.extend(page_results)
+
+            # A little optimization - if current page is not "full", ie. does
+            # not contain page_size videos then we can assume that this page
+            # is the last one - there are no more ids on further pages -
+            # i.e. no need to query again.
+            if len(page_results) + startv < self._pagesize:
+                break
+
+            # If we got the whole page, but the next page is not interesting,
+            # break out early as well
+            if end == nextfirstid:
+                break
+        return res

youtube_dl/version.py

@@ -1,2 +1,2 @@
 
-__version__ = '2014.01.22.2'
+__version__ = '2014.01.22.4'