release 2015.02.02.5

'' might actually be passed in, so check for None.

README.md

@@ -368,11 +368,11 @@ which means you can modify it, redistribute it or use it however you like.
     --add-metadata                   write metadata to the video file
     --xattrs                         write metadata to the video file's xattrs
                                      (using dublin core and xdg standards)
-    --fixup POLICY                   (experimental) Automatically correct known
-                                     faults of the file. One of never (do
-                                     nothing), warn (only emit a warning),
-                                     detect_or_warn(check whether we can do
-                                     anything about it, warn otherwise
+    --fixup POLICY                   Automatically correct known faults of the
+                                     file. One of never (do nothing), warn (only
+                                     emit a warning), detect_or_warn(the
+                                     default; fix file if we can, warn
+                                     otherwise)
     --prefer-avconv                  Prefer avconv over ffmpeg for running the
                                      postprocessors (default)
     --prefer-ffmpeg                  Prefer ffmpeg over avconv for running the

test/test_utils.py

@@ -238,6 +238,8 @@ class TestUtil(unittest.TestCase):
         self.assertEqual(parse_duration('5 s'), 5)
         self.assertEqual(parse_duration('3 min'), 180)
         self.assertEqual(parse_duration('2.5 hours'), 9000)
+        self.assertEqual(parse_duration('02:03:04'), 7384)
+        self.assertEqual(parse_duration('01:02:03:04'), 93784)
 
     def test_fix_xml_ampersands(self):
         self.assertEqual(
@@ -371,6 +373,16 @@ class TestUtil(unittest.TestCase):
         on = js_to_json('{"abc": true}')
         self.assertEqual(json.loads(on), {'abc': True})
 
+        # Ignore JavaScript code as well
+        on = js_to_json('''{
+            "x": 1,
+            y: "a",
+            z: some.code
+        }''')
+        d = json.loads(on)
+        self.assertEqual(d['x'], 1)
+        self.assertEqual(d['y'], 'a')
+
     def test_clean_html(self):
         self.assertEqual(clean_html('a:\nb'), 'a: b')
         self.assertEqual(clean_html('a:\n   "b"'), 'a:    "b"')

youtube_dl/downloader/external.py

@@ -45,6 +45,12 @@ class ExternalFD(FileDownloader):
     def supports(cls, info_dict):
         return info_dict['protocol'] in ('http', 'https', 'ftp', 'ftps')
 
+    def _source_address(self, command_option):
+        source_address = self.params.get('source_address')
+        if source_address is None:
+            return []
+        return [command_option, source_address]
+
     def _call_downloader(self, tmpfilename, info_dict):
         """ Either overwrite this or implement _make_cmd """
         cmd = self._make_cmd(tmpfilename, info_dict)
@@ -72,6 +78,7 @@ class CurlFD(ExternalFD):
         cmd = [self.exe, '-o', tmpfilename]
         for key, val in info_dict['http_headers'].items():
             cmd += ['--header', '%s: %s' % (key, val)]
+        cmd += self._source_address('--interface')
         cmd += ['--', info_dict['url']]
         return cmd
 
@@ -81,6 +88,7 @@ class WgetFD(ExternalFD):
         cmd = [self.exe, '-O', tmpfilename, '-nv', '--no-cookies']
         for key, val in info_dict['http_headers'].items():
             cmd += ['--header', '%s: %s' % (key, val)]
+        cmd += self._source_address('--bind-address')
         cmd += ['--', info_dict['url']]
         return cmd
 
@@ -96,6 +104,7 @@ class Aria2cFD(ExternalFD):
         cmd += ['--out', os.path.basename(tmpfilename)]
         for key, val in info_dict['http_headers'].items():
             cmd += ['--header', '%s: %s' % (key, val)]
+        cmd += self._source_address('--interface')
         cmd += ['--', info_dict['url']]
         return cmd
 

youtube_dl/downloader/http.py

@@ -3,6 +3,9 @@ from __future__ import unicode_literals
 import os
 import time
 
+from socket import error as SocketError
+import errno
+
 from .common import FileDownloader
 from ..compat import (
     compat_urllib_request,
@@ -99,6 +102,11 @@ class HttpFD(FileDownloader):
                             resume_len = 0
                             open_mode = 'wb'
                             break
+            except SocketError as e:
+                if e.errno != errno.ECONNRESET:
+                    # Connection reset is no problem, just retry
+                    raise
+
             # Retry
             count += 1
             if count <= retries:

youtube_dl/extractor/__init__.py

@@ -182,6 +182,7 @@ from .heise import HeiseIE
 from .hellporno import HellPornoIE
 from .helsinki import HelsinkiIE
 from .hentaistigma import HentaiStigmaIE
+from .historicfilms import HistoricFilmsIE
 from .hitbox import HitboxIE, HitboxLiveIE
 from .hornbunny import HornBunnyIE
 from .hostingbulk import HostingBulkIE
@@ -284,6 +285,7 @@ from .ndr import NDRIE
 from .ndtv import NDTVIE
 from .netzkino import NetzkinoIE
 from .nerdcubed import NerdCubedFeedIE
+from .nerdist import NerdistIE
 from .newgrounds import NewgroundsIE
 from .newstube import NewstubeIE
 from .nextmedia import (
@@ -316,7 +318,8 @@ from .nrk import (
     NRKIE,
     NRKTVIE,
 )
-from .ntv import NTVIE
+from .ntvde import NTVDeIE
+from .ntvru import NTVRuIE
 from .nytimes import NYTimesIE
 from .nuvid import NuvidIE
 from .oktoberfesttv import OktoberfestTVIE

youtube_dl/extractor/brightcove.py

@@ -108,7 +108,7 @@ class BrightcoveIE(InfoExtractor):
         """
 
         # Fix up some stupid HTML, see https://github.com/rg3/youtube-dl/issues/1553
-        object_str = re.sub(r'(<param name="[^"]+" value="[^"]+")>',
+        object_str = re.sub(r'(<param(?:\s+[a-zA-Z0-9_]+="[^"]*")*)>',
                             lambda m: m.group(1) + '/>', object_str)
         # Fix up some stupid XML, see https://github.com/rg3/youtube-dl/issues/1608
         object_str = object_str.replace('<--', '<!--')

youtube_dl/extractor/common.py

@@ -860,10 +860,13 @@ class InfoExtractor(object):
         return formats
 
     # TODO: improve extraction
-    def _extract_smil_formats(self, smil_url, video_id):
+    def _extract_smil_formats(self, smil_url, video_id, fatal=True):
         smil = self._download_xml(
             smil_url, video_id, 'Downloading SMIL file',
-            'Unable to download SMIL file')
+            'Unable to download SMIL file', fatal=fatal)
+        if smil is False:
+            assert not fatal
+            return []
 
         base = smil.find('./head/meta').get('base')
 

youtube_dl/extractor/drtv.py

@@ -25,9 +25,15 @@ class DRTVIE(SubtitlesInfoExtractor):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        programcard = self._download_json(
-            'http://www.dr.dk/mu/programcard/expanded/%s' % video_id, video_id, 'Downloading video JSON')
+        webpage = self._download_webpage(url, video_id)
 
+        video_id = self._search_regex(
+            r'data-(?:material-identifier|episode-slug)="([^"]+)"',
+            webpage, 'video id')
+
+        programcard = self._download_json(
+            'http://www.dr.dk/mu/programcard/expanded/%s' % video_id,
+            video_id, 'Downloading video JSON')
         data = programcard['Data'][0]
 
         title = data['Title']

youtube_dl/extractor/historicfilms.py

@@ -0,0 +1,46 @@
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import parse_duration
+
+
+class HistoricFilmsIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?historicfilms\.com/(?:tapes/|play)(?P<id>\d+)'
+    _TEST = {
+        'url': 'http://www.historicfilms.com/tapes/4728',
+        'md5': 'd4a437aec45d8d796a38a215db064e9a',
+        'info_dict': {
+            'id': '4728',
+            'ext': 'mov',
+            'title': 'Historic Films: GP-7',
+            'description': 'md5:1a86a0f3ac54024e419aba97210d959a',
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'duration': 2096,
+        },
+    }
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, video_id)
+
+        tape_id = self._search_regex(
+            r'class="tapeId">([^<]+)<', webpage, 'tape id')
+
+        title = self._og_search_title(webpage)
+        description = self._og_search_description(webpage)
+        thumbnail = self._html_search_meta(
+            'thumbnailUrl', webpage, 'thumbnails') or self._og_search_thumbnail(webpage)
+        duration = parse_duration(self._html_search_meta(
+            'duration', webpage, 'duration'))
+
+        video_url = 'http://www.historicfilms.com/video/%s_%s_web.mov' % (tape_id, video_id)
+
+        return {
+            'id': video_id,
+            'url': video_url,
+            'title': title,
+            'description': description,
+            'thumbnail': thumbnail,
+            'duration': duration,
+        }

youtube_dl/extractor/nerdist.py

@@ -0,0 +1,80 @@
+# encoding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+
+from ..utils import (
+    determine_ext,
+    parse_iso8601,
+    xpath_text,
+)
+
+
+class NerdistIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?nerdist\.com/vepisode/(?P<id>[^/?#]+)'
+    _TEST = {
+        'url': 'http://www.nerdist.com/vepisode/exclusive-which-dc-characters-w',
+        'md5': '3698ed582931b90d9e81e02e26e89f23',
+        'info_dict': {
+            'display_id': 'exclusive-which-dc-characters-w',
+            'id': 'RPHpvJyr',
+            'ext': 'mp4',
+            'title': 'Your TEEN TITANS Revealed! Who\'s on the show?',
+            'thumbnail': 're:^https?://.*/thumbs/.*\.jpg$',
+            'description': 'Exclusive: Find out which DC Comics superheroes will star in TEEN TITANS Live-Action TV Show on Nerdist News with Jessica Chobot!',
+            'uploader': 'Eric Diaz',
+            'upload_date': '20150202',
+            'timestamp': 1422892808,
+        }
+    }
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url)
+        webpage = self._download_webpage(url, display_id)
+
+        video_id = self._search_regex(
+            r'''(?x)<script\s+(?:type="text/javascript"\s+)?
+                src="https?://content\.nerdist\.com/players/([a-zA-Z0-9_]+)-''',
+            webpage, 'video ID')
+        timestamp = parse_iso8601(self._html_search_meta(
+            'shareaholic:article_published_time', webpage, 'upload date'))
+        uploader = self._html_search_meta(
+            'shareaholic:article_author_name', webpage, 'article author')
+
+        doc = self._download_xml(
+            'http://content.nerdist.com/jw6/%s.xml' % video_id, video_id)
+        video_info = doc.find('.//item')
+        title = xpath_text(video_info, './title', fatal=True)
+        description = xpath_text(video_info, './description')
+        thumbnail = xpath_text(
+            video_info, './{http://rss.jwpcdn.com/}image', 'thumbnail')
+
+        formats = []
+        for source in video_info.findall('./{http://rss.jwpcdn.com/}source'):
+            vurl = source.attrib['file']
+            ext = determine_ext(vurl)
+            if ext == 'm3u8':
+                formats.extend(self._extract_m3u8_formats(
+                    vurl, video_id, entry_protocol='m3u8_native', ext='mp4',
+                    preference=0))
+            elif ext == 'smil':
+                formats.extend(self._extract_smil_formats(
+                    vurl, video_id, fatal=False
+                ))
+            else:
+                formats.append({
+                    'format_id': ext,
+                    'url': vurl,
+                })
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'display_id': display_id,
+            'title': title,
+            'description': description,
+            'thumbnail': thumbnail,
+            'timestamp': timestamp,
+            'formats': formats,
+            'uploader': uploader,
+        }

youtube_dl/extractor/nfl.py

@@ -46,7 +46,18 @@ class NFLIE(InfoExtractor):
                 'timestamp': 1388354455,
                 'thumbnail': 're:^https?://.*\.jpg$',
             }
-        }
+        },
+        {
+            'url': 'http://www.nfl.com/news/story/0ap3000000467586/article/patriots-seahawks-involved-in-lategame-skirmish',
+            'info_dict': {
+                'id': '0ap3000000467607',
+                'ext': 'mp4',
+                'title': 'Frustrations flare on the field',
+                'description': 'Emotions ran high at the end of the Super Bowl on both sides of the ball after a dramatic finish.',
+                'timestamp': 1422850320,
+                'upload_date': '20150202',
+            },
+        },
     ]
 
     @staticmethod
@@ -80,7 +91,11 @@ class NFLIE(InfoExtractor):
         webpage = self._download_webpage(url, video_id)
 
         config_url = NFLIE.prepend_host(host, self._search_regex(
-            r'(?:config|configURL)\s*:\s*"([^"]+)"', webpage, 'config URL'))
+            r'(?:config|configURL)\s*:\s*"([^"]+)"', webpage, 'config URL',
+            default='static/content/static/config/video/config.json'))
+        # For articles, the id in the url is not the video id
+        video_id = self._search_regex(
+            r'contentId\s*:\s*"([^"]+)"', webpage, 'video id', default=video_id)
         config = self._download_json(config_url, video_id,
                                      note='Downloading player config')
         url_template = NFLIE.prepend_host(

youtube_dl/extractor/normalboots.py

@@ -1,8 +1,6 @@
 # encoding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
 
 from ..utils import (
@@ -11,7 +9,7 @@ from ..utils import (
 
 
 class NormalbootsIE(InfoExtractor):
-    _VALID_URL = r'http://(?:www\.)?normalboots\.com/video/(?P<videoid>[0-9a-z-]*)/?$'
+    _VALID_URL = r'http://(?:www\.)?normalboots\.com/video/(?P<id>[0-9a-z-]*)/?$'
     _TEST = {
         'url': 'http://normalboots.com/video/home-alone-games-jontron/',
         'md5': '8bf6de238915dd501105b44ef5f1e0f6',
@@ -30,19 +28,22 @@ class NormalbootsIE(InfoExtractor):
     }
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('videoid')
-
+        video_id = self._match_id(url)
         webpage = self._download_webpage(url, video_id)
-        video_uploader = self._html_search_regex(r'Posted\sby\s<a\shref="[A-Za-z0-9/]*">(?P<uploader>[A-Za-z]*)\s</a>',
-                                                 webpage, 'uploader')
-        raw_upload_date = self._html_search_regex('<span style="text-transform:uppercase; font-size:inherit;">[A-Za-z]+, (?P<date>.*)</span>',
-                                                  webpage, 'date')
-        video_upload_date = unified_strdate(raw_upload_date)
 
-        player_url = self._html_search_regex(r'<iframe\swidth="[0-9]+"\sheight="[0-9]+"\ssrc="(?P<url>[\S]+)"', webpage, 'url')
+        video_uploader = self._html_search_regex(
+            r'Posted\sby\s<a\shref="[A-Za-z0-9/]*">(?P<uploader>[A-Za-z]*)\s</a>',
+            webpage, 'uploader', fatal=False)
+        video_upload_date = unified_strdate(self._html_search_regex(
+            r'<span style="text-transform:uppercase; font-size:inherit;">[A-Za-z]+, (?P<date>.*)</span>',
+            webpage, 'date', fatal=False))
+
+        player_url = self._html_search_regex(
+            r'<iframe\swidth="[0-9]+"\sheight="[0-9]+"\ssrc="(?P<url>[\S]+)"',
+            webpage, 'player url')
         player_page = self._download_webpage(player_url, video_id)
-        video_url = self._html_search_regex(r"file:\s'(?P<file>[^']+\.mp4)'", player_page, 'file')
+        video_url = self._html_search_regex(
+            r"file:\s'(?P<file>[^']+\.mp4)'", player_page, 'file')
 
         return {
             'id': video_id,

youtube_dl/extractor/ntvde.py

@@ -0,0 +1,68 @@
+# encoding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    int_or_none,
+    js_to_json,
+    parse_duration,
+)
+
+
+class NTVDeIE(InfoExtractor):
+    IE_NAME = 'n-tv.de'
+    _VALID_URL = r'https?://(?:www\.)?n-tv\.de/mediathek/videos/[^/?#]+/[^/?#]+-article(?P<id>.+)\.html'
+
+    _TESTS = [{
+        'url': 'http://www.n-tv.de/mediathek/videos/panorama/Schnee-und-Glaette-fuehren-zu-zahlreichen-Unfaellen-und-Staus-article14438086.html',
+        'md5': '6ef2514d4b1e8e03ca24b49e2f167153',
+        'info_dict': {
+            'id': '14438086',
+            'ext': 'mp4',
+            'thumbnail': 're:^https?://.*\.jpg$',
+            'title': 'Schnee und Glätte führen zu zahlreichen Unfällen und Staus',
+            'alt_title': 'Winterchaos auf deutschen Straßen',
+            'description': 'Schnee und Glätte sorgen deutschlandweit für einen chaotischen Start in die Woche: Auf den Straßen kommt es zu kilometerlangen Staus und Dutzenden Glätteunfällen. In Düsseldorf und München wirbelt der Schnee zudem den Flugplan durcheinander. Dutzende Flüge landen zu spät, einige fallen ganz aus.',
+            'duration': 4020,
+            'timestamp': 1422892797,
+            'upload_date': '20150202',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        info = self._parse_json(self._search_regex(
+            r'(?s)ntv.pageInfo.article =\s(\{.*?\});', webpage, 'info'),
+            video_id, transform_source=js_to_json)
+        timestamp = int_or_none(info.get('publishedDateAsUnixTimeStamp'))
+        vdata = self._parse_json(self._search_regex(
+            r'(?s)\$\(\s*"\#player"\s*\)\s*\.data\(\s*"player",\s*(\{.*?\})\);',
+            webpage, 'player data'),
+            video_id, transform_source=js_to_json)
+        duration = parse_duration(vdata.get('duration'))
+        formats = [{
+            'format_id': 'flash',
+            'url': 'rtmp://fms.n-tv.de/' + vdata['video'],
+        }, {
+            'format_id': 'mobile',
+            'url': 'http://video.n-tv.de' + vdata['videoMp4'],
+            'tbr': 400,  # estimation
+        }]
+        m3u8_url = 'http://video.n-tv.de' + vdata['videoM3u8']
+        formats.extend(self._extract_m3u8_formats(
+            m3u8_url, video_id, ext='mp4',
+            entry_protocol='m3u8_native', preference=0))
+        self._sort_formats(formats)
+
+        return {
+            'id': video_id,
+            'title': info['headline'],
+            'description': info.get('intro'),
+            'alt_title': info.get('kicker'),
+            'timestamp': timestamp,
+            'thumbnail': vdata.get('html5VideoPoster'),
+            'duration': duration,
+            'formats': formats,
+        }

youtube_dl/extractor/ntvru.py

@@ -1,15 +1,14 @@
 # encoding: utf-8
 from __future__ import unicode_literals
 
-import re
-
 from .common import InfoExtractor
 from ..utils import (
     unescapeHTML
 )
 
 
-class NTVIE(InfoExtractor):
+class NTVRuIE(InfoExtractor):
+    IE_NAME = 'ntv.ru'
     _VALID_URL = r'http://(?:www\.)?ntv\.ru/(?P<id>.+)'
 
     _TESTS = [
@@ -92,9 +91,7 @@ class NTVIE(InfoExtractor):
     ]
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group('id')
-
+        video_id = self._match_id(url)
         page = self._download_webpage(url, video_id)
 
         video_id = self._html_search_regex(self._VIDEO_ID_REGEXES, page, 'video id')

youtube_dl/extractor/vevo.py

@@ -9,6 +9,7 @@ from ..compat import (
 )
 from ..utils import (
     ExtractorError,
+    int_or_none,
 )
 
 
@@ -192,9 +193,29 @@ class VevoIE(InfoExtractor):
         # Download via HLS API
         formats.extend(self._download_api_formats(video_id))
 
+        # Download SMIL
+        smil_blocks = sorted((
+            f for f in video_info['videoVersions']
+            if f['sourceType'] == 13),
+            key=lambda f: f['version'])
+        smil_url = '%s/Video/V2/VFILE/%s/%sr.smil' % (
+            self._SMIL_BASE_URL, video_id, video_id.lower())
+        if smil_blocks:
+            smil_url_m = self._search_regex(
+                r'url="([^"]+)"', smil_blocks[-1]['data'], 'SMIL URL',
+                default=None)
+            if smil_url_m is not None:
+                smil_url = smil_url_m
+        if smil_url:
+            smil_xml = self._download_webpage(
+                smil_url, video_id, 'Downloading SMIL info', fatal=False)
+            if smil_xml:
+                formats.extend(self._formats_from_smil(smil_xml))
+
         self._sort_formats(formats)
-        timestamp_ms = int(self._search_regex(
-            r'/Date\((\d+)\)/', video_info['launchDate'], 'launch date'))
+        timestamp_ms = int_or_none(self._search_regex(
+            r'/Date\((\d+)\)/',
+            video_info['launchDate'], 'launch date', fatal=False))
 
         return {
             'id': video_id,

youtube_dl/options.py

@@ -698,10 +698,9 @@ def parseOpts(overrideArguments=None):
     postproc.add_option(
         '--fixup',
         metavar='POLICY', dest='fixup', default='detect_or_warn',
-        help='(experimental) Automatically correct known faults of the file. '
+        help='Automatically correct known faults of the file. '
              'One of never (do nothing), warn (only emit a warning), '
-             'detect_or_warn(check whether we can do anything about it, warn '
-             'otherwise')
+             'detect_or_warn(the default; fix file if we can, warn otherwise)')
     postproc.add_option(
         '--prefer-avconv',
         action='store_false', dest='prefer_ffmpeg',

youtube_dl/utils.py

@@ -1275,7 +1275,10 @@ def parse_duration(s):
             (?P<only_hours>[0-9.]+)\s*(?:hours?)|
 
             (?:
-                (?:(?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*)?
+                (?:
+                    (?:(?P<days>[0-9]+)\s*(?:[:d]|days?)\s*)?
+                    (?P<hours>[0-9]+)\s*(?:[:h]|hours?)\s*
+                )?
                 (?P<mins>[0-9]+)\s*(?:[:m]|mins?|minutes?)\s*
             )?
             (?P<secs>[0-9]+)(?P<ms>\.[0-9]+)?\s*(?:s|secs?|seconds?)?
@@ -1293,6 +1296,8 @@ def parse_duration(s):
         res += int(m.group('mins')) * 60
     if m.group('hours'):
         res += int(m.group('hours')) * 60 * 60
+    if m.group('days'):
+        res += int(m.group('days')) * 24 * 60 * 60
     if m.group('ms'):
         res += float(m.group('ms'))
     return res
@@ -1543,7 +1548,7 @@ def js_to_json(code):
     res = re.sub(r'''(?x)
         "(?:[^"\\]*(?:\\\\|\\")?)*"|
         '(?:[^'\\]*(?:\\\\|\\')?)*'|
-        [a-zA-Z_][a-zA-Z_0-9]*
+        [a-zA-Z_][.a-zA-Z_0-9]*
         ''', fix_kv, code)
     res = re.sub(r',(\s*\])', lambda m: m.group(1), res)
     return res

youtube_dl/version.py

@@ -1,3 +1,3 @@
 from __future__ import unicode_literals
 
-__version__ = '2015.02.02'
+__version__ = '2015.02.02.5'