X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/e76f531201cd41dfc0ce00be28bcc5c575c7acc5..c64c42cb1f9487398f8b91a51266ff6e79115b90:/youtube_dl/extractor/soundcloud.py

diff --git a/youtube_dl/extractor/soundcloud.py b/youtube_dl/extractor/soundcloud.py
index 2b60d35..aeae931 100644
--- a/youtube_dl/extractor/soundcloud.py
+++ b/youtube_dl/extractor/soundcloud.py
@@ -4,11 +4,14 @@ from __future__ import unicode_literals
 import re
 import itertools
 
-from .common import InfoExtractor
+from .common import (
+    InfoExtractor,
+    SearchInfoExtractor
+)
 from ..compat import (
     compat_str,
     compat_urlparse,
-    compat_urllib_parse,
+    compat_urllib_parse_urlencode,
 )
 from ..utils import (
     ExtractorError,
@@ -116,6 +119,12 @@ class SoundcloudIE(InfoExtractor):
     _CLIENT_ID = '02gUJC0hH2ct1EGOcYXQIzRFU91c72Ea'
     _IPHONE_CLIENT_ID = '376f225bf427445fc4bfb6b99b72e0bf'
 
+    @staticmethod
+    def _extract_urls(webpage):
+        return [m.group('url') for m in re.finditer(
+            r'<iframe[^>]+src=(["\'])(?P<url>(?:https?://)?(?:w\.)?soundcloud\.com/player.+?)\1',
+            webpage)]
+
     def report_resolve(self, video_id):
         """Report information extraction."""
         self.to_screen('%s: Resolving id' % video_id)
@@ -218,7 +227,7 @@ class SoundcloudIE(InfoExtractor):
             full_title = track_id
             token = mobj.group('secret_token')
             if token:
-                info_json_url += "&secret_token=" + token
+                info_json_url += '&secret_token=' + token
         elif mobj.group('player'):
             query = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
             real_url = query['url'][0]
@@ -380,27 +389,24 @@ class SoundcloudUserIE(SoundcloudIE):
         resource = mobj.group('rsrc') or 'all'
         base_url = self._BASE_URL_MAP[resource] % user['id']
 
-        next_href = None
+        COMMON_QUERY = {
+            'limit': 50,
+            'client_id': self._CLIENT_ID,
+            'linked_partitioning': '1',
+        }
+
+        query = COMMON_QUERY.copy()
+        query['offset'] = 0
+
+        next_href = base_url + '?' + compat_urllib_parse_urlencode(query)
 
         entries = []
         for i in itertools.count():
-            if not next_href:
-                data = compat_urllib_parse.urlencode({
-                    'offset': i * 50,
-                    'limit': 50,
-                    'client_id': self._CLIENT_ID,
-                    'linked_partitioning': '1',
-                    'representation': 'speedy',
-                })
-                next_href = base_url + '?' + data
-
             response = self._download_json(
                 next_href, uploader, 'Downloading track page %s' % (i + 1))
 
             collection = response['collection']
-
             if not collection:
-                self.to_screen('%s: End page received' % uploader)
                 break
 
             def resolve_permalink_url(candidates):
@@ -415,12 +421,15 @@ class SoundcloudUserIE(SoundcloudIE):
                 if permalink_url:
                     entries.append(self.url_result(permalink_url))
 
-            if 'next_href' in response:
-                next_href = response['next_href']
-                if not next_href:
-                    break
-            else:
-                next_href = None
+            next_href = response.get('next_href')
+            if not next_href:
+                break
+
+            parsed_next_href = compat_urlparse.urlparse(response['next_href'])
+            qs = compat_urlparse.parse_qs(parsed_next_href.query)
+            qs.update(COMMON_QUERY)
+            next_href = compat_urlparse.urlunparse(
+                parsed_next_href._replace(query=compat_urllib_parse_urlencode(qs, True)))
 
         return {
             '_type': 'playlist',
@@ -456,7 +465,7 @@ class SoundcloudPlaylistIE(SoundcloudIE):
         if token:
             data_dict['secret_token'] = token
 
-        data = compat_urllib_parse.urlencode(data_dict)
+        data = compat_urllib_parse_urlencode(data_dict)
         data = self._download_json(
             base_url + data, playlist_id, 'Downloading playlist')
 
@@ -469,3 +478,60 @@ class SoundcloudPlaylistIE(SoundcloudIE):
             'description': data.get('description'),
             'entries': entries,
         }
+
+
+class SoundcloudSearchIE(SearchInfoExtractor, SoundcloudIE):
+    IE_NAME = 'soundcloud:search'
+    IE_DESC = 'Soundcloud search'
+    _MAX_RESULTS = float('inf')
+    _TESTS = [{
+        'url': 'scsearch15:post-avant jazzcore',
+        'info_dict': {
+            'title': 'post-avant jazzcore',
+        },
+        'playlist_count': 15,
+    }]
+
+    _SEARCH_KEY = 'scsearch'
+    _MAX_RESULTS_PER_PAGE = 200
+    _DEFAULT_RESULTS_PER_PAGE = 50
+    _API_V2_BASE = 'https://api-v2.soundcloud.com'
+
+    def _get_collection(self, endpoint, collection_id, **query):
+        limit = min(
+            query.get('limit', self._DEFAULT_RESULTS_PER_PAGE),
+            self._MAX_RESULTS_PER_PAGE)
+        query['limit'] = limit
+        query['client_id'] = self._CLIENT_ID
+        query['linked_partitioning'] = '1'
+        query['offset'] = 0
+        data = compat_urllib_parse_urlencode(query)
+        next_url = '{0}{1}?{2}'.format(self._API_V2_BASE, endpoint, data)
+
+        collected_results = 0
+
+        for i in itertools.count(1):
+            response = self._download_json(
+                next_url, collection_id, 'Downloading page {0}'.format(i),
+                'Unable to download API page')
+
+            collection = response.get('collection', [])
+            if not collection:
+                break
+
+            collection = list(filter(bool, collection))
+            collected_results += len(collection)
+
+            for item in collection:
+                yield self.url_result(item['uri'], SoundcloudIE.ie_key())
+
+            if not collection or collected_results >= limit:
+                break
+
+            next_url = response.get('next_href')
+            if not next_url:
+                break
+
+    def _get_n_results(self, query, n):
+        tracks = self._get_collection('/search/tracks', query, limit=n, q=query)
+        return self.playlist_result(tracks, playlist_title=query)