release 2014.08.24

We're seeing quite a number of people who do not put a video file in the og:video field, but the player URL. Try to detect some of these and filter them out.

youtube_dl/__init__.py

@@ -71,6 +71,7 @@ __authors__  = (
     'Sebastian Haas',
     'Alexander Kirk',
     'Erik Johnson',
+    'Keith Beckman',
 )
 
 __license__ = 'Public Domain'

youtube_dl/downloader/http.py

@@ -27,8 +27,16 @@ class HttpFD(FileDownloader):
             headers['Youtubedl-user-agent'] = info_dict['user_agent']
         if 'http_referer' in info_dict:
             headers['Referer'] = info_dict['http_referer']
-        basic_request = compat_urllib_request.Request(url, None, headers)
-        request = compat_urllib_request.Request(url, None, headers)
+        add_headers = info_dict.get('http_headers')
+        if add_headers:
+            headers.update(add_headers)
+        data = info_dict.get('http_post_data')
+        http_method = info_dict.get('http_method')
+        basic_request = compat_urllib_request.Request(url, data, headers)
+        request = compat_urllib_request.Request(url, data, headers)
+        if http_method is not None:
+            basic_request.get_method = lambda: http_method
+            request.get_method = lambda: http_method
 
         is_test = self.params.get('test', False)
 

youtube_dl/extractor/__init__.py

@@ -126,6 +126,7 @@ from .googleplus import GooglePlusIE
 from .googlesearch import GoogleSearchIE
 from .gorillavid import GorillaVidIE
 from .goshgay import GoshgayIE
+from .grooveshark import GroovesharkIE
 from .hark import HarkIE
 from .helsinki import HelsinkiIE
 from .hentaistigma import HentaiStigmaIE

youtube_dl/extractor/brightcove.py

@@ -154,12 +154,14 @@ class BrightcoveIE(InfoExtractor):
     def _extract_brightcove_urls(cls, webpage):
         """Return a list of all Brightcove URLs from the webpage """
 
-        url_m = re.search(r'<meta\s+property="og:video"\s+content="(http://c.brightcove.com/[^"]+)"', webpage)
+        url_m = re.search(
+            r'<meta\s+property="og:video"\s+content="(https?://(?:secure|c)\.brightcove.com/[^"]+)"',
+            webpage)
         if url_m:
             url = unescapeHTML(url_m.group(1))
             # Some sites don't add it, we can't download with this url, for example:
             # http://www.ktvu.com/videos/news/raw-video-caltrain-releases-video-of-man-almost/vCTZdY/
-            if 'playerKey' in url:
+            if 'playerKey' in url or 'videoId' in url:
                 return [url]
 
         matches = re.findall(
@@ -188,9 +190,13 @@ class BrightcoveIE(InfoExtractor):
             referer = smuggled_data.get('Referer', url)
             return self._get_video_info(
                 videoPlayer[0], query_str, query, referer=referer)
-        else:
+        elif 'playerKey' in query:
             player_key = query['playerKey']
             return self._get_playlist_info(player_key[0])
+        else:
+            raise ExtractorError(
+                'Cannot find playerKey= variable. Did you forget quotes in a shell invocation?',
+                expected=True)
 
     def _get_video_info(self, video_id, query_str, query, referer=None):
         request_url = self._FEDERATED_URL_TEMPLATE % query_str
@@ -202,6 +208,13 @@ class BrightcoveIE(InfoExtractor):
             req.add_header('Referer', referer)
         webpage = self._download_webpage(req, video_id)
 
+        error_msg = self._html_search_regex(
+            r"<h1>We're sorry.</h1>\s*<p>(.*?)</p>", webpage,
+            'error message', default=None)
+        if error_msg is not None:
+            raise ExtractorError(
+                'brightcove said: %s' % error_msg, expected=True)
+
         self.report_extraction(video_id)
         info = self._search_regex(r'var experienceJSON = ({.*});', webpage, 'json')
         info = json.loads(info)['data']

youtube_dl/extractor/common.py

@@ -84,6 +84,12 @@ class InfoExtractor(object):
                                  format, irrespective of the file format.
                                  -1 for default (order by other properties),
                                  -2 or smaller for less than default.
+                    * http_referer  HTTP Referer header value to set.
+                    * http_method  HTTP method to use for the download.
+                    * http_headers  A dictionary of additional HTTP headers
+                                 to add to the request.
+                    * http_post_data  Additional data to send with a POST
+                                 request.
     url:            Final video URL.
     ext:            Video filename extension.
     format:         The video format, defaults to ext (used for --get-format)

youtube_dl/extractor/generic.py

@@ -15,6 +15,7 @@ from ..utils import (
     compat_xml_parse_error,
 
     ExtractorError,
+    float_or_none,
     HEADRequest,
     orderedSet,
     parse_xml,
@@ -305,6 +306,30 @@ class GenericIE(InfoExtractor):
             'params': {
                 'skip_download': True,
             }
+        },
+        # Camtasia studio
+        {
+            'url': 'http://www.ll.mit.edu/workshops/education/videocourses/antennas/lecture1/video/',
+            'playlist': [{
+                'md5': '0c5e352edabf715d762b0ad4e6d9ee67',
+                'info_dict': {
+                    'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
+                    'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - video1',
+                    'ext': 'flv',
+                    'duration': 2235.90,
+                }
+            }, {
+                'md5': '10e4bb3aaca9fd630e273ff92d9f3c63',
+                'info_dict': {
+                    'id': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final_PIP',
+                    'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final - pip',
+                    'ext': 'flv',
+                    'duration': 2235.93,
+                }
+            }],
+            'info_dict': {
+                'title': 'Fenn-AA_PA_Radar_Course_Lecture_1c_Final',
+            }
         }
     ]
 
@@ -389,6 +414,43 @@ class GenericIE(InfoExtractor):
             'entries': entries,
         }
 
+    def _extract_camtasia(self, url, video_id, webpage):
+        """ Returns None if no camtasia video can be found. """
+
+        camtasia_cfg = self._search_regex(
+            r'fo\.addVariable\(\s*"csConfigFile",\s*"([^"]+)"\s*\);',
+            webpage, 'camtasia configuration file', default=None)
+        if camtasia_cfg is None:
+            return None
+
+        title = self._html_search_meta('DC.title', webpage, fatal=True)
+
+        camtasia_url = compat_urlparse.urljoin(url, camtasia_cfg)
+        camtasia_cfg = self._download_xml(
+            camtasia_url, video_id,
+            note='Downloading camtasia configuration',
+            errnote='Failed to download camtasia configuration')
+        fileset_node = camtasia_cfg.find('./playlist/array/fileset')
+
+        entries = []
+        for n in fileset_node.getchildren():
+            url_n = n.find('./uri')
+            if url_n is None:
+                continue
+
+            entries.append({
+                'id': os.path.splitext(url_n.text.rpartition('/')[2])[0],
+                'title': '%s - %s' % (title, n.tag),
+                'url': compat_urlparse.urljoin(url, url_n.text),
+                'duration': float_or_none(n.find('./duration').text),
+            })
+
+        return {
+            '_type': 'playlist',
+            'entries': entries,
+            'title': title,
+        }
+
     def _real_extract(self, url):
         if url.startswith('//'):
             return {
@@ -477,6 +539,11 @@ class GenericIE(InfoExtractor):
         except compat_xml_parse_error:
             pass
 
+        # Is it a Camtasia project?
+        camtasia_res = self._extract_camtasia(url, video_id, webpage)
+        if camtasia_res is not None:
+            return camtasia_res
+
         # Sometimes embedded video player is hidden behind percent encoding
         # (e.g. https://github.com/rg3/youtube-dl/issues/2448)
         # Unescaping the whole page allows to handle those cases in a generic way
@@ -764,7 +831,12 @@ class GenericIE(InfoExtractor):
             m_video_type = re.findall(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
             # We only look in og:video if the MIME type is a video, don't try if it's a Flash player:
             if m_video_type is not None:
-                found = re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
+                def check_video(vurl):
+                    vpath = compat_urlparse.urlparse(vurl).path
+                    return '.' in vpath and not vpath.endswith('.swf')
+                found = list(filter(
+                    check_video,
+                    re.findall(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)))
         if not found:
             # HTML5 video
             found = re.findall(r'(?s)<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage)

youtube_dl/extractor/grooveshark.py

@@ -0,0 +1,190 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import time
+import math
+import os.path
+import re
+
+
+from .common import InfoExtractor
+from ..utils import ExtractorError, compat_urllib_request, compat_html_parser
+
+from ..utils import (
+    compat_urllib_parse,
+    compat_urlparse,
+)
+
+
+class GroovesharkHtmlParser(compat_html_parser.HTMLParser):
+    def __init__(self):
+        self._current_object = None
+        self.objects = []
+        compat_html_parser.HTMLParser.__init__(self)
+
+    def handle_starttag(self, tag, attrs):
+        attrs = dict((k, v) for k, v in attrs)
+        if tag == 'object':
+            self._current_object = {'attrs': attrs, 'params': []}
+        elif tag == 'param':
+            self._current_object['params'].append(attrs)
+
+    def handle_endtag(self, tag):
+        if tag == 'object':
+            self.objects.append(self._current_object)
+            self._current_object = None
+
+    @classmethod
+    def extract_object_tags(cls, html):
+        p = cls()
+        p.feed(html)
+        p.close()
+        return p.objects
+
+
+class GroovesharkIE(InfoExtractor):
+    _VALID_URL = r'https?://(www\.)?grooveshark\.com/#!/s/([^/]+)/([^/]+)'
+    _TEST = {
+        'url': 'http://grooveshark.com/#!/s/Jolene+Tenth+Key+Remix+Ft+Will+Sessions/6SS1DW?src=5',
+        'md5': '7ecf8aefa59d6b2098517e1baa530023',
+        'info_dict': {
+            'id': '6SS1DW',
+            'title': 'Jolene (Tenth Key Remix ft. Will Sessions)',
+            'ext': 'mp3',
+            'duration': 227,
+        }
+    }
+
+    do_playerpage_request = True
+    do_bootstrap_request = True
+
+    def _parse_target(self, target):
+        uri = compat_urlparse.urlparse(target)
+        hash = uri.fragment[1:].split('?')[0]
+        token = os.path.basename(hash.rstrip('/'))
+        return (uri, hash, token)
+
+    def _build_bootstrap_url(self, target):
+        (uri, hash, token) = self._parse_target(target)
+        query = 'getCommunicationToken=1&hash=%s&%d' % (compat_urllib_parse.quote(hash, safe=''), self.ts)
+        return (compat_urlparse.urlunparse((uri.scheme, uri.netloc, '/preload.php', None, query, None)), token)
+
+    def _build_meta_url(self, target):
+        (uri, hash, token) = self._parse_target(target)
+        query = 'hash=%s&%d' % (compat_urllib_parse.quote(hash, safe=''), self.ts)
+        return (compat_urlparse.urlunparse((uri.scheme, uri.netloc, '/preload.php', None, query, None)), token)
+
+    def _build_stream_url(self, meta):
+        return compat_urlparse.urlunparse(('http', meta['streamKey']['ip'], '/stream.php', None, None, None))
+
+    def _build_swf_referer(self, target, obj):
+        (uri, _, _) = self._parse_target(target)
+        return compat_urlparse.urlunparse((uri.scheme, uri.netloc, obj['attrs']['data'], None, None, None))
+
+    def _transform_bootstrap(self, js):
+        return re.split('(?m)^\s*try\s*{', js)[0] \
+                 .split(' = ', 1)[1].strip().rstrip(';')
+
+    def _transform_meta(self, js):
+        return js.split('\n')[0].split('=')[1].rstrip(';')
+
+    def _get_meta(self, target):
+        (meta_url, token) = self._build_meta_url(target)
+        self.to_screen('Metadata URL: %s' % meta_url)
+
+        headers = {'Referer': compat_urlparse.urldefrag(target)[0]}
+        req = compat_urllib_request.Request(meta_url, headers=headers)
+        res = self._download_json(req, token,
+                                  transform_source=self._transform_meta)
+
+        if 'getStreamKeyWithSong' not in res:
+            raise ExtractorError(
+                'Metadata not found. URL may be malformed, or Grooveshark API may have changed.')
+
+        if res['getStreamKeyWithSong'] is None:
+            raise ExtractorError(
+                'Metadata download failed, probably due to Grooveshark anti-abuse throttling. Wait at least an hour before retrying from this IP.',
+                expected=True)
+
+        return res['getStreamKeyWithSong']
+
+    def _get_bootstrap(self, target):
+        (bootstrap_url, token) = self._build_bootstrap_url(target)
+
+        headers = {'Referer': compat_urlparse.urldefrag(target)[0]}
+        req = compat_urllib_request.Request(bootstrap_url, headers=headers)
+        res = self._download_json(req, token, fatal=False,
+                                  note='Downloading player bootstrap data',
+                                  errnote='Unable to download player bootstrap data',
+                                  transform_source=self._transform_bootstrap)
+        return res
+
+    def _get_playerpage(self, target):
+        (_, _, token) = self._parse_target(target)
+
+        webpage = self._download_webpage(
+            target, token,
+            note='Downloading player page',
+            errnote='Unable to download player page',
+            fatal=False)
+
+        if webpage is not None:
+            # Search (for example German) error message
+            error_msg = self._html_search_regex(
+                r'<div id="content">\s*<h2>(.*?)</h2>', webpage,
+                'error message', default=None)
+            if error_msg is not None:
+                error_msg = error_msg.replace('\n', ' ')
+                raise ExtractorError('Grooveshark said: %s' % error_msg)
+
+        if webpage is not None:
+            o = GroovesharkHtmlParser.extract_object_tags(webpage)
+            return (webpage, [x for x in o if x['attrs']['id'] == 'jsPlayerEmbed'])
+
+        return (webpage, None)
+
+    def _real_initialize(self):
+        self.ts = int(time.time() * 1000)  # timestamp in millis
+
+    def _real_extract(self, url):
+        (target_uri, _, token) = self._parse_target(url)
+
+        # 1. Fill cookiejar by making a request to the player page
+        swf_referer = None
+        if self.do_playerpage_request:
+            (_, player_objs) = self._get_playerpage(url)
+            if player_objs is not None:
+                swf_referer = self._build_swf_referer(url, player_objs[0])
+                self.to_screen('SWF Referer: %s' % swf_referer)
+
+        # 2. Ask preload.php for swf bootstrap data to better mimic webapp
+        if self.do_bootstrap_request:
+            bootstrap = self._get_bootstrap(url)
+            self.to_screen('CommunicationToken: %s' % bootstrap['getCommunicationToken'])
+
+        # 3. Ask preload.php for track metadata.
+        meta = self._get_meta(url)
+
+        # 4. Construct stream request for track.
+        stream_url = self._build_stream_url(meta)
+        duration = int(math.ceil(float(meta['streamKey']['uSecs']) / 1000000))
+        post_dict = {'streamKey': meta['streamKey']['streamKey']}
+        post_data = compat_urllib_parse.urlencode(post_dict).encode('utf-8')
+        headers = {
+            'Content-Length': len(post_data),
+            'Content-Type': 'application/x-www-form-urlencoded'
+        }
+        if swf_referer is not None:
+            headers['Referer'] = swf_referer
+
+        return {
+            'id': token,
+            'title': meta['song']['Name'],
+            'http_method': 'POST',
+            'url': stream_url,
+            'ext': 'mp3',
+            'format': 'mp3 audio',
+            'duration': duration,
+            'http_post_data': post_data,
+            'http_headers': headers,
+        }

youtube_dl/extractor/wat.py

@@ -2,27 +2,30 @@
 from __future__ import unicode_literals
 
 import re
+import time
+import hashlib
 
 from .common import InfoExtractor
 from ..utils import (
+    ExtractorError,
     unified_strdate,
 )
 
 
 class WatIE(InfoExtractor):
-    _VALID_URL = r'http://www\.wat\.tv/.*-(?P<shortID>.*?)_.*?\.html'
+    _VALID_URL = r'http://www\.wat\.tv/video/(?P<display_id>.*)-(?P<short_id>.*?)_.*?\.html'
     IE_NAME = 'wat.tv'
     _TEST = {
-        'url': 'http://www.wat.tv/video/world-war-philadelphia-vost-6bv55_2fjr7_.html',
+        'url': 'http://www.wat.tv/video/soupe-figues-l-orange-aux-epices-6z1uz_2hvf7_.html',
+        'md5': 'ce70e9223945ed26a8056d413ca55dc9',
         'info_dict': {
-            'id': '10631273',
+            'id': '11713067',
+            'display_id': 'soupe-figues-l-orange-aux-epices',
             'ext': 'mp4',
-            'title': 'World War Z - Philadelphia VOST',
-            'description': 'La menace est partout. Que se passe-t-il à Philadelphia ?\r\nWORLD WAR Z, avec Brad Pitt, au cinéma le 3 juillet.\r\nhttp://www.worldwarz.fr',
-        },
-        'params': {
-            # Sometimes wat serves the whole file with the --test option
-            'skip_download': True,
+            'title': 'Soupe de figues à l\'orange et aux épices',
+            'description': 'Retrouvez l\'émission "Petits plats en équilibre", diffusée le 18 août 2014.',
+            'upload_date': '20140819',
+            'duration': 120,
         },
     }
 
@@ -36,13 +39,20 @@ class WatIE(InfoExtractor):
         def real_id_for_chapter(chapter):
             return chapter['tc_start'].split('-')[0]
         mobj = re.match(self._VALID_URL, url)
-        short_id = mobj.group('shortID')
-        webpage = self._download_webpage(url, short_id)
+        short_id = mobj.group('short_id')
+        display_id = mobj.group('display_id')
+        webpage = self._download_webpage(url, display_id or short_id)
         real_id = self._search_regex(r'xtpage = ".*-(.*?)";', webpage, 'real id')
 
         video_info = self.download_video_info(real_id)
+
+        if video_info.get('geolock'):
+            raise ExtractorError('This content is not available in your area', expected=True)
+
         chapters = video_info['chapters']
         first_chapter = chapters[0]
+        files = video_info['files']
+        first_file = files[0]
 
         if real_id_for_chapter(first_chapter) != real_id:
             self.to_screen('Multipart video detected')
@@ -61,12 +71,45 @@ class WatIE(InfoExtractor):
             upload_date = unified_strdate(first_chapter['date_diffusion'])
         # Otherwise we can continue and extract just one part, we have to use
         # the short id for getting the video url
+
+        formats = [{
+            'url': 'http://wat.tv/get/android5/%s.mp4' % real_id,
+            'format_id': 'Mobile',
+        }]
+
+        fmts = [('SD', 'web')]
+        if first_file.get('hasHD'):
+            fmts.append(('HD', 'webhd'))
+
+        def compute_token(param):
+            timestamp = '%08x' % int(time.time())
+            magic = '9b673b13fa4682ed14c3cfa5af5310274b514c4133e9b3a81e6e3aba009l2564'
+            return '%s/%s' % (hashlib.md5((magic + param + timestamp).encode('ascii')).hexdigest(), timestamp)
+
+        for fmt in fmts:
+            webid = '/%s/%s' % (fmt[1], real_id)
+            video_url = self._download_webpage(
+                'http://www.wat.tv/get%s?token=%s&getURL=1' % (webid, compute_token(webid)),
+                real_id,
+                'Downloding %s video URL' % fmt[0],
+                'Failed to download %s video URL' % fmt[0],
+                False)
+            if not video_url:
+                continue
+            formats.append({
+                'url': video_url,
+                'ext': 'mp4',
+                'format_id': fmt[0],
+            })
+
         return {
             'id': real_id,
-            'url': 'http://wat.tv/get/android5/%s.mp4' % real_id,
+            'display_id': display_id,
             'title': first_chapter['title'],
             'thumbnail': first_chapter['preview'],
             'description': first_chapter['description'],
             'view_count': video_info['views'],
             'upload_date': upload_date,
+            'duration': first_file['duration'],
+            'formats': formats,
         }

youtube_dl/version.py

@@ -1,2 +1,2 @@
 
-__version__ = '2014.08.23'
+__version__ = '2014.08.24'