release 2014.03.24.5

Conflicts:
	youtube_dl/extractor/arte.py

test/test_utils.py

@@ -10,6 +10,7 @@ sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 # Various small unit tests
 import io
+import json
 import xml.etree.ElementTree
 
 #from youtube_dl.utils import htmlentity_transform
@@ -36,6 +37,7 @@ from youtube_dl.utils import (
     urlencode_postdata,
     xpath_with_ns,
     parse_iso8601,
+    strip_jsonp,
 )
 
 if sys.version_info < (3, 0):
@@ -272,5 +274,11 @@ class TestUtil(unittest.TestCase):
         self.assertEqual(parse_iso8601('2014-03-23T22:04:26+0000'), 1395612266)
         self.assertEqual(parse_iso8601('2014-03-23T22:04:26Z'), 1395612266)
 
+    def test_strip_jsonp(self):
+        stripped = strip_jsonp('cb ([ {"id":"532cb",\n\n\n"x":\n3}\n]\n);')
+        d = json.loads(stripped)
+        self.assertEqual(d, [{"id": "532cb", "x": 3}])
+
+
 if __name__ == '__main__':
     unittest.main()

youtube_dl/downloader/hls.py

@@ -13,8 +13,10 @@ class HlsFD(FileDownloader):
         self.report_destination(filename)
         tmpfilename = self.temp_name(filename)
 
-        args = ['-y', '-i', url, '-f', 'mp4', '-c', 'copy',
-            '-bsf:a', 'aac_adtstoasc', tmpfilename]
+        args = [
+            '-y', '-i', url, '-f', 'mp4', '-c', 'copy',
+            '-bsf:a', 'aac_adtstoasc',
+            encodeFilename(tmpfilename, for_subprocess=True)]
 
         for program in ['avconv', 'ffmpeg']:
             try:

youtube_dl/extractor/__init__.py

@@ -14,6 +14,7 @@ from .arte import (
     ArteTVConcertIE,
     ArteTVFutureIE,
     ArteTVDDCIE,
+    ArteTVEmbedIE,
 )
 from .auengine import AUEngineIE
 from .bambuser import BambuserIE, BambuserChannelIE
@@ -281,6 +282,7 @@ from .vine import VineIE
 from .viki import VikiIE
 from .vk import VKIE
 from .vube import VubeIE
+from .washingtonpost import WashingtonPostIE
 from .wat import WatIE
 from .wdr import WDRIE
 from .weibo import WeiboIE

youtube_dl/extractor/arte.py

@@ -2,7 +2,6 @@
 from __future__ import unicode_literals
 
 import re
-import json
 
 from .common import InfoExtractor
 from ..utils import (
@@ -19,114 +18,41 @@ from ..utils import (
 # is different for each one. The videos usually expire in 7 days, so we can't
 # add tests.
 
-class ArteTvIE(InfoExtractor):
-    _VIDEOS_URL = r'(?:http://)?videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html'
-    _LIVEWEB_URL = r'(?:http://)?liveweb\.arte\.tv/(?P<lang>fr|de)/(?P<subpage>.+?)/(?P<name>.+)'
-    _LIVE_URL = r'index-[0-9]+\.html$'
 
+class ArteTvIE(InfoExtractor):
+    _VALID_URL = r'http://videos\.arte\.tv/(?P<lang>fr|de)/.*-(?P<id>.*?)\.html'
     IE_NAME = 'arte.tv'
 
-    @classmethod
-    def suitable(cls, url):
-        return any(re.match(regex, url) for regex in (cls._VIDEOS_URL, cls._LIVEWEB_URL))
-
-    # TODO implement Live Stream
-    # from ..utils import compat_urllib_parse
-    # def extractLiveStream(self, url):
-    #     video_lang = url.split('/')[-4]
-    #     info = self.grep_webpage(
-    #         url,
-    #         r'src="(.*?/videothek_js.*?\.js)',
-    #         0,
-    #         [
-    #             (1, 'url', 'Invalid URL: %s' % url)
-    #         ]
-    #     )
-    #     http_host = url.split('/')[2]
-    #     next_url = 'http://%s%s' % (http_host, compat_urllib_parse.unquote(info.get('url')))
-    #     info = self.grep_webpage(
-    #         next_url,
-    #         r'(s_artestras_scst_geoFRDE_' + video_lang + '.*?)\'.*?' +
-    #             '(http://.*?\.swf).*?' +
-    #             '(rtmp://.*?)\'',
-    #         re.DOTALL,
-    #         [
-    #             (1, 'path',   'could not extract video path: %s' % url),
-    #             (2, 'player', 'could not extract video player: %s' % url),
-    #             (3, 'url',    'could not extract video url: %s' % url)
-    #         ]
-    #     )
-    #     video_url = '%s/%s' % (info.get('url'), info.get('path'))
-
     def _real_extract(self, url):
-        mobj = re.match(self._VIDEOS_URL, url)
-        if mobj is not None:
-            id = mobj.group('id')
-            lang = mobj.group('lang')
-            return self._extract_video(url, id, lang)
+        mobj = re.match(self._VALID_URL, url)
+        lang = mobj.group('lang')
+        video_id = mobj.group('id')
 
-        mobj = re.match(self._LIVEWEB_URL, url)
-        if mobj is not None:
-            name = mobj.group('name')
-            lang = mobj.group('lang')
-            return self._extract_liveweb(url, name, lang)
-
-        if re.search(self._LIVE_URL, url) is not None:
-            raise ExtractorError('Arte live streams are not yet supported, sorry')
-            # self.extractLiveStream(url)
-            # return
-
-        raise ExtractorError('No video found')
-
-    def _extract_video(self, url, video_id, lang):
-        """Extract from videos.arte.tv"""
         ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
         ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
         ref_xml_doc = self._download_xml(
             ref_xml_url, video_id, note='Downloading metadata')
         config_node = find_xpath_attr(ref_xml_doc, './/video', 'lang', lang)
         config_xml_url = config_node.attrib['ref']
-        config_xml = self._download_webpage(
+        config = self._download_xml(
             config_xml_url, video_id, note='Downloading configuration')
 
-        video_urls = list(re.finditer(r'<url quality="(?P<quality>.*?)">(?P<url>.*?)</url>', config_xml))
-        def _key(m):
-            quality = m.group('quality')
-            if quality == 'hd':
-                return 2
-            else:
-                return 1
-        # We pick the best quality
-        video_urls = sorted(video_urls, key=_key)
-        video_url = list(video_urls)[-1].group('url')
-        
-        title = self._html_search_regex(r'<name>(.*?)</name>', config_xml, 'title')
-        thumbnail = self._html_search_regex(r'<firstThumbnailUrl>(.*?)</firstThumbnailUrl>',
-                                            config_xml, 'thumbnail')
-        return {'id': video_id,
-                'title': title,
-                'thumbnail': thumbnail,
-                'url': video_url,
-                'ext': 'flv',
-                }
+        formats = [{
+            'forma_id': q.attrib['quality'],
+            'url': q.text,
+            'ext': 'flv',
+            'quality': 2 if q.attrib['quality'] == 'hd' else 1,
+        } for q in config.findall('./urls/url')]
+        self._sort_formats(formats)
 
-    def _extract_liveweb(self, url, name, lang):
-        """Extract form http://liveweb.arte.tv/"""
-        webpage = self._download_webpage(url, name)
-        video_id = self._search_regex(r'eventId=(\d+?)("|&)', webpage, 'event id')
-        config_doc = self._download_xml('http://download.liveweb.arte.tv/o21/liveweb/events/event-%s.xml' % video_id,
-                                            video_id, 'Downloading information')
-        event_doc = config_doc.find('event')
-        url_node = event_doc.find('video').find('urlHd')
-        if url_node is None:
-            url_node = event_doc.find('urlSd')
-
-        return {'id': video_id,
-                'title': event_doc.find('name%s' % lang.capitalize()).text,
-                'url': url_node.text.replace('MP4', 'mp4'),
-                'ext': 'flv',
-                'thumbnail': self._og_search_thumbnail(webpage),
-                }
+        title = config.find('.//name').text
+        thumbnail = config.find('.//firstThumbnailUrl').text
+        return {
+            'id': video_id,
+            'title': title,
+            'thumbnail': thumbnail,
+            'formats': formats,
+        }
 
 
 class ArteTVPlus7IE(InfoExtractor):
@@ -152,9 +78,7 @@ class ArteTVPlus7IE(InfoExtractor):
         return self._extract_from_json_url(json_url, video_id, lang)
 
     def _extract_from_json_url(self, json_url, video_id, lang):
-        json_info = self._download_webpage(json_url, video_id, 'Downloading info json')
-        self.report_extraction(video_id)
-        info = json.loads(json_info)
+        info = self._download_json(json_url, video_id)
         player_info = info['videoJsonPlayer']
 
         info_dict = {
@@ -176,6 +100,8 @@ class ArteTVPlus7IE(InfoExtractor):
                 l = 'F'
             elif lang == 'de':
                 l = 'A'
+            else:
+                l = lang
             regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l]
             return any(re.match(r, f['versionCode']) for r in regexes)
         # Some formats may not be in the same language as the url
@@ -305,3 +231,22 @@ class ArteTVConcertIE(ArteTVPlus7IE):
             'description': 'md5:486eb08f991552ade77439fe6d82c305',
         },
     }
+
+
+class ArteTVEmbedIE(ArteTVPlus7IE):
+    IE_NAME = 'arte.tv:embed'
+    _VALID_URL = r'''(?x)
+        http://www\.arte\.tv
+        /playerv2/embed\.php\?json_url=
+        (?P<json_url>
+            http://arte\.tv/papi/tvguide/videos/stream/player/
+            (?P<lang>[^/]+)/(?P<id>[^/]+)[^&]*
+        )
+    '''
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
+        lang = mobj.group('lang')
+        json_url = mobj.group('json_url')
+        return self._extract_from_json_url(json_url, video_id, lang)

youtube_dl/extractor/clipfish.py

@@ -1,22 +1,28 @@
+from __future__ import unicode_literals
+
 import re
 import time
 import xml.etree.ElementTree
 
 from .common import InfoExtractor
-from ..utils import ExtractorError
+from ..utils import (
+    ExtractorError,
+    parse_duration,
+)
 
 
 class ClipfishIE(InfoExtractor):
-    IE_NAME = u'clipfish'
+    IE_NAME = 'clipfish'
 
     _VALID_URL = r'^https?://(?:www\.)?clipfish\.de/.*?/video/(?P<id>[0-9]+)/'
     _TEST = {
-        u'url': u'http://www.clipfish.de/special/game-trailer/video/3966754/fifa-14-e3-2013-trailer/',
-        u'file': u'3966754.mp4',
-        u'md5': u'2521cd644e862936cf2e698206e47385',
-        u'info_dict': {
-            u'title': u'FIFA 14 - E3 2013 Trailer',
-            u'duration': 82,
+        'url': 'http://www.clipfish.de/special/game-trailer/video/3966754/fifa-14-e3-2013-trailer/',
+        'md5': '2521cd644e862936cf2e698206e47385',
+        'info_dict': {
+            'id': '3966754',
+            'ext': 'mp4',
+            'title': 'FIFA 14 - E3 2013 Trailer',
+            'duration': 82,
         },
         u'skip': 'Blocked in the US'
     }
@@ -33,21 +39,10 @@ class ClipfishIE(InfoExtractor):
         video_url = doc.find('filename').text
         if video_url is None:
             xml_bytes = xml.etree.ElementTree.tostring(doc)
-            raise ExtractorError(u'Cannot find video URL in document %r' %
+            raise ExtractorError('Cannot find video URL in document %r' %
                                  xml_bytes)
         thumbnail = doc.find('imageurl').text
-        duration_str = doc.find('duration').text
-        m = re.match(
-            r'^(?P<hours>[0-9]+):(?P<minutes>[0-9]{2}):(?P<seconds>[0-9]{2}):(?P<ms>[0-9]*)$',
-            duration_str)
-        if m:
-            duration = (
-                (int(m.group('hours')) * 60 * 60) +
-                (int(m.group('minutes')) * 60) +
-                (int(m.group('seconds')))
-            )
-        else:
-            duration = None
+        duration = parse_duration(doc.find('duration').text)
 
         return {
             'id': video_id,

youtube_dl/extractor/generic.py

@@ -197,6 +197,21 @@ class GenericIE(InfoExtractor):
                 'description': 'No description',
             },
         },
+        # arte embed
+        {
+            'url': 'http://www.tv-replay.fr/redirection/20-03-14/x-enius-arte-10753389.html',
+            'md5': '7653032cbb25bf6c80d80f217055fa43',
+            'info_dict': {
+                'id': '048195-004_PLUS7-F',
+                'ext': 'flv',
+                'title': 'X:enius',
+                'description': 'md5:d5fdf32ef6613cdbfd516ae658abf168',
+                'upload_date': '20140320',
+            },
+            'params': {
+                'skip_download': 'Requires rtmpdump'
+            }
+        },
     ]
 
     def report_download_webpage(self, video_id):
@@ -525,6 +540,13 @@ class GenericIE(InfoExtractor):
         if mobj is not None:
             return self.url_result(mobj.group('url'), 'TED')
 
+        # Look for embedded arte.tv player
+        mobj = re.search(
+            r'<script [^>]*?src="(?P<url>http://www\.arte\.tv/playerv2/embed[^"]+)"',
+            webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'), 'ArteTVEmbed')
+
         # Start with something easy: JW Player in SWFObject
         mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
         if mobj is None:

youtube_dl/extractor/rts.py

@@ -28,6 +28,7 @@ class RTSIE(InfoExtractor):
             'uploader': 'Divers',
             'upload_date': '19680921',
             'timestamp': -40280400,
+            'thumbnail': 're:^https?://.*\.image'
         },
     }
 
@@ -58,4 +59,5 @@ class RTSIE(InfoExtractor):
             'duration': duration,
             'uploader': info.get('programName'),
             'timestamp': upload_timestamp,
+            'thumbnail': thumbnail,
         }

youtube_dl/extractor/washingtonpost.py

@@ -0,0 +1,103 @@
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    int_or_none,
+    strip_jsonp,
+)
+
+
+class WashingtonPostIE(InfoExtractor):
+    _VALID_URL = r'^https?://(?:www\.)?washingtonpost\.com/.*?/(?P<id>[^/]+)/(?:$|[?#])'
+    _TEST = {
+        'url': 'http://www.washingtonpost.com/sf/national/2014/03/22/sinkhole-of-bureaucracy/',
+        'playlist': [{
+            'md5': 'c3f4b4922ffa259243f68e928db2db8c',
+            'info_dict': {
+                'id': 'fc433c38-b146-11e3-b8b3-44b1d1cd4c1f',
+                'ext': 'mp4',
+                'title': 'Breaking Points: The Paper Mine',
+                'duration': 1287,
+                'description': 'Overly complicated paper pushing is nothing new to government bureaucracy. But the way federal retirement applications are filed may be the most outdated. David Fahrenthold explains.',
+                'uploader': 'The Washington Post',
+                'timestamp': 1395527908,
+                'upload_date': '20140322',
+            },
+        }, {
+            'md5': 'f645a07652c2950cd9134bb852c5f5eb',
+            'info_dict': {
+                'id': '41255e28-b14a-11e3-b8b3-44b1d1cd4c1f',
+                'ext': 'mp4',
+                'title': 'The town bureaucracy sustains',
+                'description': 'Underneath the friendly town of Boyers is a sea of government paperwork. In a disused limestone mine, hundreds of locals now track, file and process retirement applications for the federal government. We set out to find out what it\'s like to do paperwork 230 feet underground.',
+                'duration': 2217,
+                'timestamp': 1395528005,
+                'upload_date': '20140322',
+                'uploader': 'The Washington Post',
+            },
+        }]
+    }
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        page_id = mobj.group('id')
+
+        webpage = self._download_webpage(url, page_id)
+        title = self._og_search_title(webpage)
+        uuids = re.findall(r'data-video-uuid="([^"]+)"', webpage)
+        entries = []
+        for i, uuid in enumerate(uuids, start=1):
+            vinfo_all = self._download_json(
+                'http://www.washingtonpost.com/posttv/c/videojson/%s?resType=jsonp' % uuid,
+                page_id,
+                transform_source=strip_jsonp,
+                note='Downloading information of video %d/%d' % (i, len(uuids))
+            )
+            vinfo = vinfo_all[0]['contentConfig']
+            uploader = vinfo.get('credits', {}).get('source')
+            timestamp = int_or_none(
+                vinfo.get('dateConfig', {}).get('dateFirstPublished'), 1000)
+
+            formats = [{
+                'format_id': (
+                    '%s-%s-%s' % (s.get('type'), s.get('width'), s.get('bitrate'))
+                    if s.get('width')
+                    else s.get('type')),
+                'vbr': s.get('bitrate') if s.get('width') != 0 else None,
+                'width': s.get('width'),
+                'height': s.get('height'),
+                'acodec': s.get('audioCodec'),
+                'vcodec': s.get('videoCodec') if s.get('width') != 0 else 'none',
+                'filesize': s.get('fileSize'),
+                'url': s.get('url'),
+                'ext': 'mp4',
+                'protocol': {
+                    'MP4': 'http',
+                    'F4F': 'f4m',
+                }.get(s.get('type'))
+            } for s in vinfo.get('streams', [])]
+            source_media_url = vinfo.get('sourceMediaURL')
+            if source_media_url:
+                formats.append({
+                    'format_id': 'source_media',
+                    'url': source_media_url,
+                })
+            self._sort_formats(formats)
+            entries.append({
+                'id': uuid,
+                'title': vinfo['title'],
+                'description': vinfo.get('blurb'),
+                'uploader': uploader,
+                'formats': formats,
+                'duration': int_or_none(vinfo.get('videoDuration'), 100),
+                'timestamp': timestamp,
+            })
+
+        return {
+            '_type': 'playlist',
+            'entries': entries,
+            'id': page_id,
+            'title': title,
+        }

youtube_dl/utils.py

@@ -1186,7 +1186,7 @@ def parse_duration(s):
         return None
 
     m = re.match(
-        r'(?:(?:(?P<hours>[0-9]+)[:h])?(?P<mins>[0-9]+)[:m])?(?P<secs>[0-9]+)s?$', s)
+        r'(?:(?:(?P<hours>[0-9]+)[:h])?(?P<mins>[0-9]+)[:m])?(?P<secs>[0-9]+)s?(?::[0-9]+)?$', s)
     if not m:
         return None
     res = int(m.group('secs'))
@@ -1328,3 +1328,7 @@ US_RATINGS = {
     'R': 16,
     'NC': 18,
 }
+
+
+def strip_jsonp(code):
+    return re.sub(r'(?s)^[a-zA-Z_]+\s*\(\s*(.*)\);\s*?\s*$', r'\1', code)

youtube_dl/version.py

@@ -1,2 +1,2 @@
 
-__version__ = '2014.03.24.3'
+__version__ = '2014.03.24.5'