]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/generic.py
debian/NEWS: Add link to my webpage to give more details.
[youtubedl] / youtube_dl / extractor / generic.py
index 7a877b3bcb4e51dfb68a3cbc2720ce52d17c4be9..5bcc78bf79734ddd47ee93b2215c31bdc28e59d3 100644 (file)
+# encoding: utf-8
+
+from __future__ import unicode_literals
+
 import os
 import re
 
 from .common import InfoExtractor
+from .youtube import YoutubeIE
 from ..utils import (
     compat_urllib_error,
     compat_urllib_parse,
     compat_urllib_request,
+    compat_urlparse,
 
     ExtractorError,
+    HEADRequest,
+    smuggle_url,
+    unescapeHTML,
+    unified_strdate,
+    url_basename,
 )
+from .brightcove import BrightcoveIE
+from .ooyala import OoyalaIE
 
-class GenericIE(InfoExtractor):
-    """Generic last-resort information extractor."""
 
+class GenericIE(InfoExtractor):
+    IE_DESC = 'Generic downloader that works on some sites'
     _VALID_URL = r'.*'
-    IE_NAME = u'generic'
+    IE_NAME = 'generic'
+    _TESTS = [
+        {
+            'url': 'http://www.hodiho.fr/2013/02/regis-plante-sa-jeep.html',
+            'file': '13601338388002.mp4',
+            'md5': '6e15c93721d7ec9e9ca3fdbf07982cfd',
+            'info_dict': {
+                'uploader': 'www.hodiho.fr',
+                'title': 'R\u00e9gis plante sa Jeep',
+            }
+        },
+        # bandcamp page with custom domain
+        {
+            'add_ie': ['Bandcamp'],
+            'url': 'http://bronyrock.com/track/the-pony-mash',
+            'file': '3235767654.mp3',
+            'info_dict': {
+                'title': 'The Pony Mash',
+                'uploader': 'M_Pallante',
+            },
+            'skip': 'There is a limit of 200 free downloads / month for the test song',
+        },
+        # embedded brightcove video
+        # it also tests brightcove videos that need to set the 'Referer' in the
+        # http requests
+        {
+            'add_ie': ['Brightcove'],
+            'url': 'http://www.bfmtv.com/video/bfmbusiness/cours-bourse/cours-bourse-l-analyse-technique-154522/',
+            'info_dict': {
+                'id': '2765128793001',
+                'ext': 'mp4',
+                'title': 'Le cours de bourse : l’analyse technique',
+                'description': 'md5:7e9ad046e968cb2d1114004aba466fd9',
+                'uploader': 'BFM BUSINESS',
+            },
+            'params': {
+                'skip_download': True,
+            },
+        },
+        {
+            # https://github.com/rg3/youtube-dl/issues/2253
+            'url': 'http://bcove.me/i6nfkrc3',
+            'file': '3101154703001.mp4',
+            'md5': '0ba9446db037002366bab3b3eb30c88c',
+            'info_dict': {
+                'title': 'Still no power',
+                'uploader': 'thestar.com',
+                'description': 'Mississauga resident David Farmer is still out of power as a result of the ice storm a month ago. To keep the house warm, Farmer cuts wood from his property for a wood burning stove downstairs.',
+            },
+            'add_ie': ['Brightcove'],
+        },
+        # Direct link to a video
+        {
+            'url': 'http://media.w3.org/2010/05/sintel/trailer.mp4',
+            'file': 'trailer.mp4',
+            'md5': '67d406c2bcb6af27fa886f31aa934bbe',
+            'info_dict': {
+                'id': 'trailer',
+                'title': 'trailer',
+                'upload_date': '20100513',
+            }
+        },
+        # ooyala video
+        {
+            'url': 'http://www.rollingstone.com/music/videos/norwegian-dj-cashmere-cat-goes-spartan-on-with-me-premiere-20131219',
+            'file': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ.mp4',
+            'md5': '5644c6ca5d5782c1d0d350dad9bd840c',
+            'info_dict': {
+                'id': 'BwY2RxaTrTkslxOfcan0UCf0YqyvWysJ',
+                'ext': 'mp4',
+                'title': '2cc213299525360.mov',  # that's what we get
+            },
+        },
+    ]
 
     def report_download_webpage(self, video_id):
         """Report webpage download."""
         if not self._downloader.params.get('test', False):
-            self._downloader.report_warning(u'Falling back on generic information extractor.')
+            self._downloader.report_warning('Falling back on generic information extractor.')
         super(GenericIE, self).report_download_webpage(video_id)
 
     def report_following_redirect(self, new_url):
         """Report information extraction."""
-        self._downloader.to_screen(u'[redirect] Following redirect to %s' % new_url)
+        self._downloader.to_screen('[redirect] Following redirect to %s' % new_url)
 
-    def _test_redirect(self, url):
+    def _send_head(self, url):
         """Check if it is a redirect, like url shorteners, in case return the new url."""
-        class HeadRequest(compat_urllib_request.Request):
-            def get_method(self):
-                return "HEAD"
 
         class HEADRedirectHandler(compat_urllib_request.HTTPRedirectHandler):
             """
             Subclass the HTTPRedirectHandler to make it use our
-            HeadRequest also on the redirected URL
+            HEADRequest also on the redirected URL
             """
             def redirect_request(self, req, fp, code, msg, headers, newurl):
                 if code in (301, 302, 303, 307):
                     newurl = newurl.replace(' ', '%20')
                     newheaders = dict((k,v) for k,v in req.headers.items()
                                       if k.lower() not in ("content-length", "content-type"))
-                    return HeadRequest(newurl,
+                    return HEADRequest(newurl,
                                        headers=newheaders,
                                        origin_req_host=req.get_origin_req_host(),
                                        unverifiable=True)
@@ -71,38 +154,215 @@ class GenericIE(InfoExtractor):
                         compat_urllib_request.HTTPErrorProcessor, compat_urllib_request.HTTPSHandler]:
             opener.add_handler(handler())
 
-        response = opener.open(HeadRequest(url))
+        response = opener.open(HEADRequest(url))
         if response is None:
-            raise ExtractorError(u'Invalid URL protocol')
-        new_url = response.geturl()
+            raise ExtractorError('Invalid URL protocol')
+        return response
 
-        if url == new_url:
-            return False
+    def _real_extract(self, 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'
 
-        self.report_following_redirect(new_url)
-        return new_url
+            if default_search == 'auto':
+                if '/' in url:
+                    self._downloader.report_warning('The url doesn\'t specify the protocol, trying with http')
+                    return self.url_result('http://' + url)
+                else:
+                    return self.url_result('ytsearch:' + url)
+            else:
+                assert ':' in default_search
+                return self.url_result(default_search + url)
+        video_id = os.path.splitext(url.split('/')[-1])[0]
 
-    def _real_extract(self, url):
-        new_url = self._test_redirect(url)
-        if new_url: return [self.url_result(new_url)]
+        self.to_screen('%s: Requesting header' % video_id)
+
+        try:
+            response = self._send_head(url)
+
+            # Check for redirect
+            new_url = response.geturl()
+            if url != new_url:
+                self.report_following_redirect(new_url)
+                return self.url_result(new_url)
+
+            # Check for direct link to a video
+            content_type = response.headers.get('Content-Type', '')
+            m = re.match(r'^(?P<type>audio|video|application(?=/ogg$))/(?P<format_id>.+)$', content_type)
+            if m:
+                upload_date = response.headers.get('Last-Modified')
+                if upload_date:
+                    upload_date = unified_strdate(upload_date)
+                return {
+                    'id': video_id,
+                    'title': os.path.splitext(url_basename(url))[0],
+                    'formats': [{
+                        'format_id': m.group('format_id'),
+                        'url': url,
+                        'vcodec': 'none' if m.group('type') == 'audio' else None
+                    }],
+                    'upload_date': upload_date,
+                }
+
+        except compat_urllib_error.HTTPError:
+            # This may be a stupid server that doesn't like HEAD, our UA, or so
+            pass
 
-        video_id = url.split('/')[-1]
         try:
             webpage = self._download_webpage(url, video_id)
         except ValueError:
             # since this is the last-resort InfoExtractor, if
             # this error is thrown, it'll be thrown here
-            raise ExtractorError(u'Invalid URL: %s' % url)
+            raise ExtractorError('Failed to download URL: %s' % url)
 
         self.report_extraction(video_id)
+
+        # it's tempting to parse this further, but you would
+        # have to take into account all the variations like
+        #   Video Title - Site Name
+        #   Site Name | Video Title
+        #   Video Title - Tagline | Site Name
+        # and so on and so forth; it's just not practical
+        video_title = self._html_search_regex(
+            r'(?s)<title>(.*?)</title>', webpage, 'video title',
+            default='video')
+
+        # video uploader is domain name
+        video_uploader = self._search_regex(
+            r'^(?:https?://)?([^/]*)/.*', url, 'video uploader')
+
+        # Look for BrightCove:
+        bc_urls = BrightcoveIE._extract_brightcove_urls(webpage)
+        if bc_urls:
+            self.to_screen('Brightcove video detected.')
+            entries = [{
+                '_type': 'url',
+                'url': smuggle_url(bc_url, {'Referer': url}),
+                'ie_key': 'Brightcove'
+            } for bc_url in bc_urls]
+
+            return {
+                '_type': 'playlist',
+                'title': video_title,
+                'id': video_id,
+                'entries': entries,
+            }
+
+        # Look for embedded (iframe) Vimeo player
+        mobj = re.search(
+            r'<iframe[^>]+?src="((?:https?:)?//player\.vimeo\.com/video/.+?)"', webpage)
+        if mobj:
+            player_url = unescapeHTML(mobj.group(1))
+            surl = smuggle_url(player_url, {'Referer': url})
+            return self.url_result(surl, 'Vimeo')
+
+        # Look for embedded (swf embed) Vimeo player
+        mobj = re.search(
+            r'<embed[^>]+?src="(https?://(?:www\.)?vimeo\.com/moogaloop\.swf.+?)"', webpage)
+        if mobj:
+            return self.url_result(mobj.group(1), 'Vimeo')
+
+        # Look for embedded YouTube player
+        matches = re.findall(r'''(?x)
+            (?:<iframe[^>]+?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)
+
+        # 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)
+
+        # Look for embedded Wistia player
+        match = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:fast\.)?wistia\.net/embed/iframe/.+?)\1', webpage)
+        if match:
+            return {
+                '_type': 'url_transparent',
+                'url': unescapeHTML(match.group('url')),
+                'ie_key': 'Wistia',
+                'uploader': video_uploader,
+                'title': video_title,
+                'id': video_id,
+            }
+
+        # Look for embedded blip.tv player
+        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)
+        if mobj:
+            return self.url_result(mobj.group(1), 'BlipTV')
+
+        # Look for Bandcamp pages with custom domain
+        mobj = re.search(r'<meta property="og:url"[^>]*?content="(.*?bandcamp\.com.*?)"', webpage)
+        if mobj is not None:
+            burl = unescapeHTML(mobj.group(1))
+            # Don't set the extractor because it can be a track url or an album
+            return self.url_result(burl)
+
+        # Look for embedded Vevo player
+        mobj = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>(?:https?:)?//(?:cache\.)?vevo\.com/.+?)\1', webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'))
+
+        # Look for Ooyala videos
+        mobj = re.search(r'player.ooyala.com/[^"?]+\?[^"]*?(?:embedCode|ec)=([^"&]+)', webpage)
+        if mobj is not None:
+            return OoyalaIE._build_url_result(mobj.group(1))
+
+        # Look for Aparat videos
+        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')
+
+        # Look for MPORA videos
+        mobj = re.search(r'<iframe .*?src="(http://mpora\.(?:com|de)/videos/[^"]+)"', webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group(1), 'Mpora')
+
+        # Look for embedded Novamov player
+        mobj = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>http://(?:(?:embed|www)\.)?novamov\.com/embed\.php.+?)\1', webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'), 'Novamov')
+
+        # Look for embedded Facebook player
+        mobj = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>https://www\.facebook\.com/video/embed.+?)\1', webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'), 'Facebook')
+
+        # Look for embedded Huffington Post player
+        mobj = re.search(
+            r'<iframe[^>]+?src=(["\'])(?P<url>https?://embed\.live\.huffingtonpost\.com/.+?)\1', webpage)
+        if mobj is not None:
+            return self.url_result(mobj.group('url'), 'HuffPost')
+
         # Start with something easy: JW Player in SWFObject
         mobj = re.search(r'flashvars: [\'"](?:.*&)?file=(http[^\'"&]*)', webpage)
+        if mobj is None:
+            # Look for gorilla-vid style embedding
+            mobj = re.search(r'(?s)(?:jw_plugins|JWPlayerOptions).*?file\s*:\s*["\'](.*?)["\']', webpage)
         if mobj is None:
             # 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[^\'"&]*)', webpage)
+            mobj = re.search(r'[^A-Za-z0-9]?file["\']?:\s*["\'](http(?![^\'"]+\.[0-9]+[\'"])[^\'"]+)["\']', webpage)
         if mobj is None:
             # Try to find twitter cards info
             mobj = re.search(r'<meta (?:property|name)="twitter:player:stream" (?:content|value)="(.+?)"', webpage)
@@ -114,38 +374,30 @@ class GenericIE(InfoExtractor):
             if m_video_type is not None:
                 mobj = re.search(r'<meta.*?property="og:video".*?content="(.*?)"', webpage)
         if mobj is None:
-            raise ExtractorError(u'Invalid URL: %s' % url)
+            # HTML5 video
+            mobj = re.search(r'<video[^<]*(?:>.*?<source.*?)? src="([^"]+)"', webpage, flags=re.DOTALL)
+        if mobj is None:
+            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(u'Invalid URL: %s' % url)
+            raise ExtractorError('Did not find a valid video URL at %s' % 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))
 
-        video_url = compat_urllib_parse.unquote(mobj.group(1))
-        video_id = os.path.basename(video_url)
+        # 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_extension = os.path.splitext(video_id)[1][1:]
         video_id = os.path.splitext(video_id)[0]
 
-        # it's tempting to parse this further, but you would
-        # have to take into account all the variations like
-        #   Video Title - Site Name
-        #   Site Name | Video Title
-        #   Video Title - Tagline | Site Name
-        # and so on and so forth; it's just not practical
-        video_title = self._html_search_regex(r'<title>(.*)</title>',
-            webpage, u'video title')
-
-        # video uploader is domain name
-        video_uploader = self._search_regex(r'(?:https?://)?([^/]*)/.*',
-            url, u'video uploader')
-
-        return [{
-            'id':       video_id,
-            'url':      video_url,
+        return {
+            'id': video_id,
+            'url': video_url,
             'uploader': video_uploader,
-            'upload_date':  None,
-            'title':    video_title,
-            'ext':      video_extension,
-        }]
+            'title': video_title,
+        }