debian/NEWS: Add link to my webpage to give more details.

[youtubedl] / youtube_dl / extractor / soundcloud.py

diff --git a/youtube_dl/extractor/soundcloud.py b/youtube_dl/extractor/soundcloud.py
index d47c49c03f6e5445e72255e79495baae3a16e115..393b5f17c53d5ed216b53ac0d1ff2941cb1d24f7 100644 (file)
--- a/youtube_dl/extractor/soundcloud.py
+++ b/youtube_dl/extractor/soundcloud.py
@@ -1,9 +1,15 @@
+# encoding: utf-8
+from __future__ import unicode_literals
+

 import json
 import re
 import json
 import re

+import itertools

 
 from .common import InfoExtractor
 from ..utils import (
     compat_str,
 
 from .common import InfoExtractor
 from ..utils import (
     compat_str,

+    compat_urlparse,
+    compat_urllib_parse,

 
     ExtractorError,
     unified_strdate,
 
     ExtractorError,
     unified_strdate,


@@ -19,143 +25,202 @@ class SoundcloudIE(InfoExtractor):
        of the stream token and uid
      """
 
        of the stream token and uid
      """
 

-    _VALID_URL = r'^(?:https?://)?(?:www\.)?soundcloud\.com/([\w\d-]+)/([\w\d-]+)(?:[?].*)?$'
-    IE_NAME = u'soundcloud'
-    _TEST = {
-        u'url': u'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
-        u'file': u'62986583.mp3',
-        u'md5': u'ebef0a451b909710ed1d7787dddbf0d7',
-        u'info_dict': {
-            u"upload_date": u"20121011", 
-            u"description": u"No Downloads untill we record the finished version this weekend, i was too pumped n i had to post it , earl is prolly gonna b hella p.o'd", 
-            u"uploader": u"E.T. ExTerrestrial Music", 
-            u"title": u"Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1"
-        }
-    }
+    _VALID_URL = r'''^(?:https?://)?
+                    (?:(?:(?:www\.|m\.)?soundcloud\.com/
+                            (?P<uploader>[\w\d-]+)/
+                            (?!sets/)(?P<title>[\w\d-]+)/?
+                            (?P<token>[^?]+?)?(?:[?].*)?$)
+                       |(?:api\.soundcloud\.com/tracks/(?P<track_id>\d+))
+                       |(?P<player>(?:w|player|p.)\.soundcloud\.com/player/?.*?url=.*)
+                    )
+                    '''
+    IE_NAME = 'soundcloud'
+    _TESTS = [
+        {
+            'url': 'http://soundcloud.com/ethmusic/lostin-powers-she-so-heavy',
+            'file': '62986583.mp3',
+            'md5': 'ebef0a451b909710ed1d7787dddbf0d7',
+            'info_dict': {
+                "upload_date": "20121011",
+                "description": "No Downloads untill we record the finished version this weekend, i was too pumped n i had to post it , earl is prolly gonna b hella p.o'd",
+                "uploader": "E.T. ExTerrestrial Music",
+                "title": "Lostin Powers - She so Heavy (SneakPreview) Adrian Ackers Blueprint 1"
+            }
+        },
+        # not streamable song
+        {
+            'url': 'https://soundcloud.com/the-concept-band/goldrushed-mastered?in=the-concept-band/sets/the-royal-concept-ep',
+            'info_dict': {
+                'id': '47127627',
+                'ext': 'mp3',
+                'title': 'Goldrushed',
+                'uploader': 'The Royal Concept',
+                'upload_date': '20120521',
+            },
+            'params': {
+                # rtmp
+                'skip_download': True,
+            },
+        },
+        # private link
+        {
+            'url': 'https://soundcloud.com/jaimemf/youtube-dl-test-video-a-y-baw/s-8Pjrp',
+            'md5': 'aa0dd32bfea9b0c5ef4f02aacd080604',
+            'info_dict': {
+                'id': '123998367',
+                'ext': 'mp3',
+                'title': 'Youtube - Dl Test Video \'\' Ä↭',
+                'uploader': 'jaimeMF',
+                'description': 'test chars:  \"\'/\\ä↭',
+                'upload_date': '20131209',
+            },
+        },
+        # downloadable song
+        {
+            'url': 'https://soundcloud.com/simgretina/just-your-problem-baby-1',
+            'md5': '56a8b69568acaa967b4c49f9d1d52d19',
+            'info_dict': {
+                'id': '105614606',
+                'ext': 'wav',
+                'title': 'Just Your Problem Baby (Acapella)',
+                'description': 'Vocals',
+                'uploader': 'Sim Gretina',
+                'upload_date': '20130815',
+            },
+        },
+    ]
+
+    _CLIENT_ID = 'b45b1aa10f1ac2941910a7f0d10f8e28'
+    _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
+
+    @classmethod
+    def suitable(cls, url):
+        return re.match(cls._VALID_URL, url, flags=re.VERBOSE) is not None

 
     def report_resolve(self, video_id):
         """Report information extraction."""
         self.to_screen(u'%s: Resolving id' % video_id)
 
 
     def report_resolve(self, video_id):
         """Report information extraction."""
         self.to_screen(u'%s: Resolving id' % video_id)
 

-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        if mobj is None:
-            raise ExtractorError(u'Invalid URL: %s' % url)
+    @classmethod
+    def _resolv_url(cls, url):
+        return 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=' + cls._CLIENT_ID

 
 

-        # extract uploader (which is in the url)
-        uploader = mobj.group(1)
-        # extract simple title (uploader + slug of song title)
-        slug_title =  mobj.group(2)
-        full_title = '%s/%s' % (uploader, slug_title)
+    def _extract_info_dict(self, info, full_title=None, quiet=False, secret_token=None):
+        track_id = compat_str(info['id'])
+        name = full_title or track_id
+        if quiet:
+            self.report_extraction(name)

 
 

-        self.report_resolve(full_title)
+        thumbnail = info['artwork_url']
+        if thumbnail is not None:
+            thumbnail = thumbnail.replace('-large', '-t500x500')
+        ext = 'mp3'
+        result = {
+            'id': track_id,
+            'uploader': info['user']['username'],
+            'upload_date': unified_strdate(info['created_at']),
+            'title': info['title'],
+            'description': info['description'],
+            'thumbnail': thumbnail,
+        }
+        if info.get('downloadable', False):
+            # We can build a direct link to the song
+            format_url = (
+                'https://api.soundcloud.com/tracks/{0}/download?client_id={1}'.format(
+                    track_id, self._CLIENT_ID))
+            result['formats'] = [{
+                'format_id': 'download',
+                'ext': info.get('original_format', 'mp3'),
+                'url': format_url,
+                'vcodec': 'none',
+            }]
+        else:
+            # We have to retrieve the url
+            streams_url = ('http://api.soundcloud.com/i1/tracks/{0}/streams?'
+                'client_id={1}&secret_token={2}'.format(track_id, self._IPHONE_CLIENT_ID, secret_token))
+            stream_json = self._download_webpage(
+                streams_url,
+                track_id, 'Downloading track url')

 
 

-        url = 'http://soundcloud.com/%s/%s' % (uploader, slug_title)
-        resolv_url = 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=b45b1aa10f1ac2941910a7f0d10f8e28'
-        info_json = self._download_webpage(resolv_url, full_title, u'Downloading info JSON')
+            formats = []
+            format_dict = json.loads(stream_json)
+            for key, stream_url in format_dict.items():
+                if key.startswith(u'http'):
+                    formats.append({
+                        'format_id': key,
+                        'ext': ext,
+                        'url': stream_url,
+                        'vcodec': 'none',
+                    })
+                elif key.startswith(u'rtmp'):
+                    # The url doesn't have an rtmp app, we have to extract the playpath
+                    url, path = stream_url.split('mp3:', 1)
+                    formats.append({
+                        'format_id': key,
+                        'url': url,
+                        'play_path': 'mp3:' + path,
+                        'ext': ext,
+                        'vcodec': 'none',
+                    })

 
 

-        info = json.loads(info_json)
-        video_id = info['id']
-        self.report_extraction(full_title)
+            if not formats:
+                # We fallback to the stream_url in the original info, this
+                # cannot be always used, sometimes it can give an HTTP 404 error
+                formats.append({
+                    'format_id': 'fallback',
+                    'url': info['stream_url'] + '?client_id=' + self._CLIENT_ID,
+                    'ext': ext,
+                    'vcodec': 'none',
+                })

 
 

-        streams_url = 'https://api.sndcdn.com/i1/tracks/' + str(video_id) + '/streams?client_id=b45b1aa10f1ac2941910a7f0d10f8e28'
-        stream_json = self._download_webpage(streams_url, full_title,
-                                             u'Downloading stream definitions',
-                                             u'unable to download stream definitions')
+            for f in formats:
+                if f['format_id'].startswith('http'):
+                    f['protocol'] = 'http'
+                if f['format_id'].startswith('rtmp'):
+                    f['protocol'] = 'rtmp'

 
 

-        streams = json.loads(stream_json)
-        mediaURL = streams['http_mp3_128_url']
-        upload_date = unified_strdate(info['created_at'])
+            self._sort_formats(formats)
+            result['formats'] = formats

 
 

-        return [{
-            'id':       info['id'],
-            'url':      mediaURL,
-            'uploader': info['user']['username'],
-            'upload_date': upload_date,
-            'title':    info['title'],
-            'ext':      u'mp3',
-            'description': info['description'],
-        }]
+        return result

 
 

-class SoundcloudSetIE(InfoExtractor):
-    """Information extractor for soundcloud.com sets
-       To access the media, the uid of the song and a stream token
-       must be extracted from the page source and the script must make
-       a request to media.soundcloud.com/crossdomain.xml. Then
-       the media can be grabbed by requesting from an url composed
-       of the stream token and uid
-     """
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url, flags=re.VERBOSE)
+        if mobj is None:
+            raise ExtractorError(u'Invalid URL: %s' % url)

 
 

-    _VALID_URL = r'^(?:https?://)?(?:www\.)?soundcloud\.com/([\w\d-]+)/sets/([\w\d-]+)(?:[?].*)?$'
-    IE_NAME = u'soundcloud:set'
-    _TEST = {
-        u"url":"https://soundcloud.com/the-concept-band/sets/the-royal-concept-ep",
-        u"playlist": [
-            {
-                u"file":"30510138.mp3",
-                u"md5":"f9136bf103901728f29e419d2c70f55d",
-                u"info_dict": {
-                    u"upload_date": u"20111213",
-                    u"description": u"The Royal Concept from Stockholm\r\nFilip / Povel / David / Magnus\r\nwww.royalconceptband.com",
-                    u"uploader": u"The Royal Concept",
-                    u"title": u"D-D-Dance"
-                }
-            },
-            {
-                u"file":"47127625.mp3",
-                u"md5":"09b6758a018470570f8fd423c9453dd8",
-                u"info_dict": {
-                    u"upload_date": u"20120521",
-                    u"description": u"The Royal Concept from Stockholm\r\nFilip / Povel / David / Magnus\r\nwww.royalconceptband.com",
-                    u"uploader": u"The Royal Concept",
-                    u"title": u"The Royal Concept - Gimme Twice"
-                }
-            },
-            {
-                u"file":"47127627.mp3",
-                u"md5":"154abd4e418cea19c3b901f1e1306d9c",
-                u"info_dict": {
-                    u"upload_date": u"20120521",
-                    u"uploader": u"The Royal Concept",
-                    u"title": u"Goldrushed"
-                }
-            },
-            {
-                u"file":"47127629.mp3",
-                u"md5":"2f5471edc79ad3f33a683153e96a79c1",
-                u"info_dict": {
-                    u"upload_date": u"20120521",
-                    u"description": u"The Royal Concept from Stockholm\r\nFilip / Povel / David / Magnus\r\nwww.royalconceptband.com",
-                    u"uploader": u"The Royal Concept",
-                    u"title": u"In the End"
-                }
-            },
-            {
-                u"file":"47127631.mp3",
-                u"md5":"f9ba87aa940af7213f98949254f1c6e2",
-                u"info_dict": {
-                    u"upload_date": u"20120521",
-                    u"description": u"The Royal Concept from Stockholm\r\nFilip / David / Povel / Magnus\r\nwww.theroyalconceptband.com",
-                    u"uploader": u"The Royal Concept",
-                    u"title": u"Knocked Up"
-                }
-            },
-            {
-                u"file":"75206121.mp3",
-                u"md5":"f9d1fe9406717e302980c30de4af9353",
-                u"info_dict": {
-                    u"upload_date": u"20130116",
-                    u"description": u"The unreleased track World on Fire premiered on the CW's hit show Arrow (8pm/7pm central).  \r\nAs a gift to our fans we would like to offer you a free download of the track!  ",
-                    u"uploader": u"The Royal Concept",
-                    u"title": u"World On Fire"
-                }
-            }
-        ]
-    }
+        track_id = mobj.group('track_id')
+        token = None
+        if track_id is not None:
+            info_json_url = 'http://api.soundcloud.com/tracks/' + track_id + '.json?client_id=' + self._CLIENT_ID
+            full_title = track_id
+        elif mobj.group('player'):
+            query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
+            return self.url_result(query['url'][0], ie='Soundcloud')
+        else:
+            # extract uploader (which is in the url)
+            uploader = mobj.group('uploader')
+            # extract simple title (uploader + slug of song title)
+            slug_title =  mobj.group('title')
+            token = mobj.group('token')
+            full_title = resolve_title = '%s/%s' % (uploader, slug_title)
+            if token:
+                resolve_title += '/%s' % token
+    
+            self.report_resolve(full_title)
+    
+            url = 'http://soundcloud.com/%s' % resolve_title
+            info_json_url = self._resolv_url(url)
+        info_json = self._download_webpage(info_json_url, full_title, 'Downloading info JSON')

 
 

-    def report_resolve(self, video_id):
-        """Report information extraction."""
-        self.to_screen(u'%s: Resolving id' % video_id)
+        info = json.loads(info_json)
+        return self._extract_info_dict(info, full_title, secret_token=token)
+
+class SoundcloudSetIE(SoundcloudIE):
+    _VALID_URL = r'^(?:https?://)?(?:www\.)?soundcloud\.com/([\w\d-]+)/sets/([\w\d-]+)(?:[?].*)?$'
+    IE_NAME = 'soundcloud:set'
+    # it's in tests/test_playlists.py
+    _TESTS = []

 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)
 
     def _real_extract(self, url):
         mobj = re.match(self._VALID_URL, url)


@@ -171,10 +236,9 @@ class SoundcloudSetIE(InfoExtractor):
         self.report_resolve(full_title)
 
         url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title)
         self.report_resolve(full_title)
 
         url = 'http://soundcloud.com/%s/sets/%s' % (uploader, slug_title)

-        resolv_url = 'http://api.soundcloud.com/resolve.json?url=' + url + '&client_id=b45b1aa10f1ac2941910a7f0d10f8e28'
+        resolv_url = self._resolv_url(url)

         info_json = self._download_webpage(resolv_url, full_title)
 
         info_json = self._download_webpage(resolv_url, full_title)
 

-        videos = []

         info = json.loads(info_json)
         if 'errors' in info:
             for err in info['errors']:
         info = json.loads(info_json)
         if 'errors' in info:
             for err in info['errors']:


@@ -182,23 +246,46 @@ class SoundcloudSetIE(InfoExtractor):
             return
 
         self.report_extraction(full_title)
             return
 
         self.report_extraction(full_title)

-        for track in info['tracks']:
-            video_id = track['id']
-
-            streams_url = 'https://api.sndcdn.com/i1/tracks/' + str(video_id) + '/streams?client_id=b45b1aa10f1ac2941910a7f0d10f8e28'
-            stream_json = self._download_webpage(streams_url, video_id, u'Downloading track info JSON')
-
-            self.report_extraction(video_id)
-            streams = json.loads(stream_json)
-            mediaURL = streams['http_mp3_128_url']
-
-            videos.append({
-                'id':       video_id,
-                'url':      mediaURL,
-                'uploader': track['user']['username'],
-                'upload_date':  unified_strdate(track['created_at']),
-                'title':    track['title'],
-                'ext':      u'mp3',
-                'description': track['description'],
-            })
-        return videos
+        return {'_type': 'playlist',
+                'entries': [self._extract_info_dict(track) for track in info['tracks']],
+                'id': info['id'],
+                'title': info['title'],
+                }
+
+
+class SoundcloudUserIE(SoundcloudIE):
+    _VALID_URL = r'https?://(www\.)?soundcloud\.com/(?P<user>[^/]+)(/?(tracks/)?)?(\?.*)?$'
+    IE_NAME = 'soundcloud:user'
+
+    # it's in tests/test_playlists.py
+    _TESTS = []
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        uploader = mobj.group('user')
+
+        url = 'http://soundcloud.com/%s/' % uploader
+        resolv_url = self._resolv_url(url)
+        user_json = self._download_webpage(resolv_url, uploader,
+            'Downloading user info')
+        user = json.loads(user_json)
+
+        tracks = []
+        for i in itertools.count():
+            data = compat_urllib_parse.urlencode({'offset': i*50,
+                                                  'client_id': self._CLIENT_ID,
+                                                  })
+            tracks_url = 'http://api.soundcloud.com/users/%s/tracks.json?' % user['id'] + data
+            response = self._download_webpage(tracks_url, uploader, 
+                'Downloading tracks page %s' % (i+1))
+            new_tracks = json.loads(response)
+            tracks.extend(self._extract_info_dict(track, quiet=True) for track in new_tracks)
+            if len(new_tracks) < 50:
+                break
+
+        return {
+            '_type': 'playlist',
+            'id': compat_str(user['id']),
+            'title': user['username'],
+            'entries': tracks,
+        }