X-Git-Url: http://git.bitcoin.ninja/index.cgi?a=blobdiff_plain;f=youtube_dl%2Fextractor%2Fgeneric.py;h=8b2d1d033ab2f260e49b18c8ecc7d1d3fc695eaf;hb=e5402ac1203c8fd0c438ad515bbd80eb288b3907;hp=f9b9d56d2efa917134071b5af1cd1415e70e507a;hpb=fd50bf623c3b3696ade7b8e1e5f4e0c283eabaaf;p=youtube-dl

diff --git a/youtube_dl/extractor/generic.py b/youtube_dl/extractor/generic.py
index f9b9d56d2..8b2d1d033 100644
--- a/youtube_dl/extractor/generic.py
+++ b/youtube_dl/extractor/generic.py
@@ -15,7 +15,9 @@ from ..utils import (
     compat_xml_parse_error,
 
     ExtractorError,
+    float_or_none,
     HEADRequest,
+    orderedSet,
     parse_xml,
     smuggle_url,
     unescapeHTML,
@@ -239,6 +241,96 @@ class GenericIE(InfoExtractor):
                 'uploader_id': 'rbctv_2012_4',
             },
         },
+        # Condé Nast embed
+        {
+            'url': 'http://www.wired.com/2014/04/honda-asimo/',
+            'md5': 'ba0dfe966fa007657bd1443ee672db0f',
+            'info_dict': {
+                'id': '53501be369702d3275860000',
+                'ext': 'mp4',
+                'title': 'Honda’s  New Asimo Robot Is More Human Than Ever',
+            }
+        },
+        # Dailymotion embed
+        {
+            'url': 'http://www.spi0n.com/zap-spi0n-com-n216/',
+            'md5': '441aeeb82eb72c422c7f14ec533999cd',
+            'info_dict': {
+                'id': 'k2mm4bCdJ6CQ2i7c8o2',
+                'ext': 'mp4',
+                'title': 'Le Zap de Spi0n n°216 - Zapping du Web',
+                'uploader': 'Spi0n',
+            },
+            'add_ie': ['Dailymotion'],
+        },
+        # YouTube embed
+        {
+            'url': 'http://www.badzine.de/ansicht/datum/2014/06/09/so-funktioniert-die-neue-englische-badminton-liga.html',
+            'info_dict': {
+                'id': 'FXRb4ykk4S0',
+                'ext': 'mp4',
+                'title': 'The NBL Auction 2014',
+                'uploader': 'BADMINTON England',
+                'uploader_id': 'BADMINTONEvents',
+                'upload_date': '20140603',
+                'description': 'md5:9ef128a69f1e262a700ed83edb163a73',
+            },
+            'add_ie': ['Youtube'],
+            'params': {
+                'skip_download': True,
+            }
+        },
+        # MTVSercices embed
+        {
+            'url': 'http://www.gametrailers.com/news-post/76093/north-america-europe-is-getting-that-mario-kart-8-mercedes-dlc-too',
+            'md5': '35727f82f58c76d996fc188f9755b0d5',
+            'info_dict': {
+                'id': '0306a69b-8adf-4fb5-aace-75f8e8cbfca9',
+                'ext': 'mp4',
+                'title': 'Review',
+                'description': 'Mario\'s life in the fast lane has never looked so good.',
+            },
+        },
+        # YouTube embed via <data-embed-url="">
+        {
+            'url': 'https://play.google.com/store/apps/details?id=com.gameloft.android.ANMP.GloftA8HM',
+            'info_dict': {
+                'id': 'jpSGZsgga_I',
+                'ext': 'mp4',
+                'title': 'Asphalt 8: Airborne - Launch Trailer',
+                'uploader': 'Gameloft',
+                'uploader_id': 'gameloft',
+                'upload_date': '20130821',
+                'description': 'md5:87bd95f13d8be3e7da87a5f2c443106a',
+            },
+            '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',
+            }
+        }
     ]
 
     def report_download_webpage(self, video_id):
@@ -322,22 +414,76 @@ 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 {
+                '_type': 'url',
+                'url': self.http_scheme() + url,
+            }
+
         parsed_url = compat_urlparse.urlparse(url)
         if not parsed_url.scheme:
             default_search = self._downloader.params.get('default_search')
             if default_search is None:
-                default_search = 'auto_warning'
+                default_search = 'fixup_error'
 
-            if default_search in ('auto', 'auto_warning'):
+            if default_search in ('auto', 'auto_warning', 'fixup_error'):
                 if '/' in url:
                     self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
                     return self.url_result('http://' + url)
-                else:
+                elif default_search != 'fixup_error':
                     if default_search == 'auto_warning':
-                        self._downloader.report_warning(
-                            'Falling back to youtube search for  %s . Set --default-search to "auto" to suppress this warning.' % url)
+                        if re.match(r'^(?:url|URL)$', url):
+                            raise ExtractorError(
+                                'Invalid URL:  %r . Call youtube-dl like this:  youtube-dl -v "https://www.youtube.com/watch?v=BaW_jenozKc"  ' % url,
+                                expected=True)
+                        else:
+                            self._downloader.report_warning(
+                                'Falling back to youtube search for  %s . Set --default-search "auto" to suppress this warning.' % url)
                     return self.url_result('ytsearch:' + url)
+
+            if default_search in ('error', 'fixup_error'):
+                raise ExtractorError(
+                    ('%r is not a valid URL. '
+                     'Set --default-search "ytsearch" (or run  youtube-dl "ytsearch:%s" ) to search YouTube'
+                    ) % (url, url), expected=True)
             else:
                 assert ':' in default_search
                 return self.url_result(default_search + url)
@@ -393,6 +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
@@ -412,6 +563,12 @@ class GenericIE(InfoExtractor):
         video_uploader = self._search_regex(
             r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
 
+        # Helper method
+        def _playlist_from_matches(matches, getter, ie=None):
+            urlrs = orderedSet(self.url_result(getter(m), ie) for m in matches)
+            return self.playlist_result(
+                urlrs, playlist_id=video_id, playlist_title=video_title)
+
         # Look for BrightCove:
         bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
         if bc_urls:
@@ -445,24 +602,26 @@ class GenericIE(InfoExtractor):
 
         # Look for embedded YouTube player
         matches = re.findall(r'''(?x)
-            (?:<iframe[^>]+?src=|embedSWF\(\s*)
-            (["\'])(?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
+            (?:
+                <iframe[^>]+?src=|
+                data-video-url=|
+                <embed[^>]+?src=|
+                embedSWF\(?:\s*
+            )
+            (["\'])
+                (?P<url>(?:https?:)?//(?:www\.)?youtube\.com/
                 (?:embed|v)/.+?)
             \1''', webpage)
         if matches:
-            urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Youtube')
-                     for tuppl in matches]
-            return self.playlist_result(
-                urlrs, playlist_id=video_id, playlist_title=video_title)
+            return _playlist_from_matches(
+                matches, lambda m: unescapeHTML(m[1]), ie='Youtube')
 
         # Look for embedded Dailymotion player
         matches = re.findall(
             r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?dailymotion\.com/embed/video/.+?)\1', webpage)
         if matches:
-            urlrs = [self.url_result(unescapeHTML(tuppl[1]), 'Dailymotion')
-                     for tuppl in matches]
-            return self.playlist_result(
-                urlrs, playlist_id=video_id, playlist_title=video_title)
+            return _playlist_from_matches(
+                matches, lambda m: unescapeHTML(m[1]))
 
         # Look for embedded Wistia player
         match = re.search(
@@ -481,10 +640,26 @@ class GenericIE(InfoExtractor):
         mobj = re.search(r'<meta\s[^>]*https?://api\.blip\.tv/\w+/redirect/\w+/(\d+)', webpage)
         if mobj:
             return self.url_result('http://blip.tv/a/a-'+mobj.group(1), 'BlipTV')
-        mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9]+)', webpage)
+        mobj = re.search(r'<(?:iframe|embed|object)\s[^>]*(https?://(?:\w+\.)?blip\.tv/(?:play/|api\.swf#)[a-zA-Z0-9_]+)', webpage)
         if mobj:
             return self.url_result(mobj.group(1), 'BlipTV')
 
+        # Look for embedded condenast player
+        matches = re.findall(
+            r'<iframe\s+(?:[a-zA-Z-]+="[^"]+"\s+)*?src="(https?://player\.cnevids\.com/embed/[^"]+")',
+            webpage)
+        if matches:
+            return {
+                '_type': 'playlist',
+                'entries': [{
+                    '_type': 'url',
+                    'ie_key': 'CondeNast',
+                    'url': ma,
+                } for ma in matches],
+                'title': video_title,
+                'id': video_id,
+            }
+
         # Look for Bandcamp pages with custom domain
         mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
         if mobj is not None:
@@ -505,7 +680,7 @@ class GenericIE(InfoExtractor):
             return OoyalaIE._build_url_result(mobj.group('ec'))
 
         # Look for Aparat videos
-        mobj = re.search(r'<iframe src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
+        mobj = re.search(r'<iframe .*?src="(http://www\.aparat\.com/video/[^"]+)"', webpage)
         if mobj is not None:
             return self.url_result(mobj.group(1), 'Aparat')
 
@@ -516,7 +691,7 @@ class GenericIE(InfoExtractor):
 
         # Look for embedded NovaMov-based player
         mobj = re.search(
-            r'''(?x)<iframe[^>]+?src=(["\'])
+            r'''(?x)<(?:pagespeed_)?iframe[^>]+?src=(["\'])
                     (?P<url>http://(?:(?:embed|www)\.)?
                         (?:novamov\.com|
                            nowvideo\.(?:ch|sx|eu|at|ag|co)|
@@ -538,6 +713,11 @@ class GenericIE(InfoExtractor):
         if mobj is not None:
             return self.url_result(mobj.group('url'), 'VK')
 
+        # Look for embedded ivi player
+        mobj = re.search(r'<embed[^>]+?src=(["\'])(?P<url>https?://(?:www\.)?ivi\.ru/video/player.+?)\1', webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'), 'Ivi')
+
         # Look for embedded Huffington Post player
         mobj = re.search(
             r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
@@ -555,10 +735,8 @@ class GenericIE(InfoExtractor):
         # Look for funnyordie embed
         matches = re.findall(r'<iframe[^>]+?src="(https?://(?:www\.)?funnyordie\.com/embed/[^"]+)"', webpage)
         if matches:
-            urlrs = [self.url_result(unescapeHTML(eurl), 'FunnyOrDie')
-                     for eurl in matches]
-            return self.playlist_result(
-                urlrs, playlist_id=video_id, playlist_title=video_title)
+            return _playlist_from_matches(
+                matches, getter=unescapeHTML, ie='FunnyOrDie')
 
         # Look for embedded RUTV player
         rutv_url = RUTVIE._extract_url(webpage)
@@ -589,65 +767,121 @@ class GenericIE(InfoExtractor):
         if smotri_url:
             return self.url_result(smotri_url, 'Smotri')
 
+        # Look for embeded soundcloud player
+        mobj = re.search(
+            r'<iframe src="(?P<url>https?://(?:w\.)?soundcloud\.com/player[^"]+)"',
+            webpage)
+        if mobj is not None:
+            url = unescapeHTML(mobj.group('url'))
+            return self.url_result(url)
+
+        # Look for embedded vulture.com player
+        mobj = re.search(
+            r'<iframe src="(?P<url>https?://video\.vulture\.com/[^"]+)"',
+            webpage)
+        if mobj is not None:
+            url = unescapeHTML(mobj.group('url'))
+            return self.url_result(url, ie='Vulture')
+
+        # Look for embedded mtvservices player
+        mobj = re.search(
+            r'<iframe src="(?P<url>https?://media\.mtvnservices\.com/embed/[^"]+)"',
+            webpage)
+        if mobj is not None:
+            url = unescapeHTML(mobj.group('url'))
+            return self.url_result(url, ie='MTVServicesEmbedded')
+
+        # Look for embedded yahoo player
+        mobj = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:screen|movies)\.yahoo\.com/.+?\.html\?format=embed)\1',
+            webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'), 'Yahoo')
+
+        # Look for embedded sbs.com.au player
+        mobj = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>https?://(?:www\.)sbs\.com\.au/ondemand/video/single/.+?)\1',
+            webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'), 'SBS')
+
         # Start with something easy: JW Player in SWFObject
-        mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
-        if mobj is None:
+        found = re.findall(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
+        if not found:
             # Look for gorilla-vid style embedding
-            mobj = re.search(r'(?s)(?:jw_plugins|JWPlayerOptions).*?file\s*:\s*["\'](.*?)["\']', webpage)
-        if mobj is None:
+            found = re.findall(r'''(?sx)
+                (?:
+                    jw_plugins|
+                    JWPlayerOptions|
+                    jwplayer\s*\(\s*["'][^'"]+["']\s*\)\s*\.setup
+                )
+                .*?file\s*:\s*["\'](.*?)["\']''', webpage)
+        if not found:
             # Broaden the search a little bit
-            mobj = re.search(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
-        if mobj is None:
-            # Broaden the search a little bit: JWPlayer JS loader
-            mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
-
-        if mobj is None:
+            found = re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage)
+        if not found:
+            # Broaden the findall a little bit: JWPlayer JS loader
+            found = re.findall(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
+        if not found:
             # Try to find twitter cards info
-            mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
-        if mobj is None:
+            found = re.findall(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
+        if not found:
             # We look for Open Graph info:
             # We have to match any number spaces between elements, some sites try to align them (eg.: statigr.am)
-            m_video_type = re.search(r'<meta.*?property="og:video:type".*?content="video/(.*?)"', webpage)
+            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:
-                mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
-        if mobj is None:
+                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
-            mobj = re.search(r'<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
-        if mobj is None:
-            mobj = re.search(
+            found = re.findall(r'(?s)<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage)
+        if not found:
+            found = re.search(
                 r'(?i)<meta\s+(?=(?:[a-z-]+="[^"]+"\s+)*http-equiv="refresh")'
                 r'(?:[a-z-]+="[^"]+"\s+)*?content="[0-9]{,2};url=\'([^\']+)\'"',
                 webpage)
-            if mobj:
-                new_url = mobj.group(1)
+            if found:
+                new_url = found.group(1)
                 self.report_following_redirect(new_url)
                 return {
                     '_type': 'url',
                     'url': new_url,
                 }
-        if mobj is None:
+        if not found:
             raise ExtractorError('Unsupported URL: %s' % url)
 
-        # It's possible that one of the regexes
-        # matched, but returned an empty group:
-        if mobj.group(1) is None:
-            raise ExtractorError('Did not find a valid video URL at %s' % url)
+        entries = []
+        for video_url in found:
+            video_url = compat_urlparse.urljoin(url, video_url)
+            video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
 
-        video_url = mobj.group(1)
-        video_url = compat_urlparse.urljoin(url, video_url)
-        video_id = compat_urllib_parse.unquote(os.path.basename(video_url))
+            # Sometimes, jwplayer extraction will result in a YouTube URL
+            if YoutubeIE.suitable(video_url):
+                entries.append(self.url_result(video_url, 'Youtube'))
+                continue
 
-        # Sometimes, jwplayer extraction will result in a YouTube URL
-        if YoutubeIE.suitable(video_url):
-            return self.url_result(video_url, 'Youtube')
+            # here's a fun little line of code for you:
+            video_id = os.path.splitext(video_id)[0]
 
-        # here's a fun little line of code for you:
-        video_id = os.path.splitext(video_id)[0]
+            entries.append({
+                'id': video_id,
+                'url': video_url,
+                'uploader': video_uploader,
+                'title': video_title,
+            })
+
+        if len(entries) == 1:
+            return entries[0]
+        else:
+            for num, e in enumerate(entries, start=1):
+                e['title'] = '%s (%d)' % (e['title'], num)
+            return {
+                '_type': 'playlist',
+                'entries': entries,
+            }
 
-        return {
-            'id': video_id,
-            'url': video_url,
-            'uploader': video_uploader,
-            'title': video_title,
-        }