]> Raphaël G. Git Repositories - youtubedl/blobdiff - youtube_dl/extractor/patreon.py
Rapsys Git / youtubedl / blobdiff
? search:
summary | shortlog | log | commit | commitdiff | tree
raw | inline | side by side
Merge pull request #3 from unit193/master
[youtubedl] / youtube_dl / extractor / patreon.py
diff --git a/youtube_dl/extractor/patreon.py b/youtube_dl/extractor/patreon.py
index 5429592a75a9f66fdc9f0e9fb908af9e67559aae..761a4b1de3d194686339d4d8eb15a5f1bb2b50a6 100644 (file)
--- a/youtube_dl/extractor/patreon.py
+++ b/youtube_dl/extractor/patreon.py
@@ -1,47 +1,75 @@
-# encoding: utf-8
+# coding: utf-8
 from __future__ import unicode_literals
 
 from __future__ import unicode_literals
 
-import json
-import re
-
 from .common import InfoExtractor
 from ..utils import (
 from .common import InfoExtractor
 from ..utils import (
-    js_to_json,
+    clean_html,
+    determine_ext,
+    int_or_none,
+    KNOWN_EXTENSIONS,
+    mimetype2ext,
+    parse_iso8601,
+    str_or_none,
+    try_get,
 )
 
 
 class PatreonIE(InfoExtractor):
 )
 
 
 class PatreonIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?patreon\.com/creation\?hid=(.+)'
-    _TESTS = [
-        {
-            'url': 'http://www.patreon.com/creation?hid=743933',
-            'md5': 'e25505eec1053a6e6813b8ed369875cc',
-            'info_dict': {
-                'id': '743933',
-                'ext': 'mp3',
-                'title': 'Episode 166: David Smalley of Dogma Debate',
-                'uploader': 'Cognitive Dissonance Podcast',
-                'thumbnail': 're:^https?://.*$',
-            },
+    _VALID_URL = r'https?://(?:www\.)?patreon\.com/(?:creation\?hid=|posts/(?:[\w-]+-)?)(?P<id>\d+)'
+    _TESTS = [{
+        'url': 'http://www.patreon.com/creation?hid=743933',
+        'md5': 'e25505eec1053a6e6813b8ed369875cc',
+        'info_dict': {
+            'id': '743933',
+            'ext': 'mp3',
+            'title': 'Episode 166: David Smalley of Dogma Debate',
+            'description': 'md5:713b08b772cd6271b9f3906683cfacdf',
+            'uploader': 'Cognitive Dissonance Podcast',
+            'thumbnail': 're:^https?://.*$',
+            'timestamp': 1406473987,
+            'upload_date': '20140727',
+            'uploader_id': '87145',
+        },
+    }, {
+        'url': 'http://www.patreon.com/creation?hid=754133',
+        'md5': '3eb09345bf44bf60451b8b0b81759d0a',
+        'info_dict': {
+            'id': '754133',
+            'ext': 'mp3',
+            'title': 'CD 167 Extra',
+            'uploader': 'Cognitive Dissonance Podcast',
+            'thumbnail': 're:^https?://.*$',
         },
         },
-        {
-            'url': 'http://www.patreon.com/creation?hid=754133',
-            'md5': '3eb09345bf44bf60451b8b0b81759d0a',
-            'info_dict': {
-                'id': '754133',
-                'ext': 'mp3',
-                'title': 'CD 167 Extra',
-                'uploader': 'Cognitive Dissonance Podcast',
-                'thumbnail': 're:^https?://.*$',
-            },
+        'skip': 'Patron-only content',
+    }, {
+        'url': 'https://www.patreon.com/creation?hid=1682498',
+        'info_dict': {
+            'id': 'SU4fj_aEMVw',
+            'ext': 'mp4',
+            'title': 'I\'m on Patreon!',
+            'uploader': 'TraciJHines',
+            'thumbnail': 're:^https?://.*$',
+            'upload_date': '20150211',
+            'description': 'md5:c5a706b1f687817a3de09db1eb93acd4',
+            'uploader_id': 'TraciJHines',
         },
         },
-    ]
+        'params': {
+            'noplaylist': True,
+            'skip_download': True,
+        }
+    }, {
+        'url': 'https://www.patreon.com/posts/episode-166-of-743933',
+        'only_matching': True,
+    }, {
+        'url': 'https://www.patreon.com/posts/743933',
+        'only_matching': True,
+    }]
 
     # Currently Patreon exposes download URL via hidden CSS, so login is not
     # needed. Keeping this commented for when this inevitably changes.
     '''
     def _login(self):
 
     # Currently Patreon exposes download URL via hidden CSS, so login is not
     # needed. Keeping this commented for when this inevitably changes.
     '''
     def _login(self):
-        (username, password) = self._get_login_info()
+        username, password = self._get_login_info()
         if username is None:
             return
 
         if username is None:
             return
 
@@ -51,11 +79,11 @@ class PatreonIE(InfoExtractor):
             'password': password,
         }
 
             'password': password,
         }
 
-        request = compat_urllib_request.Request(
+        request = sanitized_Request(
             'https://www.patreon.com/processLogin',
             'https://www.patreon.com/processLogin',
-            compat_urllib_parse.urlencode(login_form).encode('utf-8')
+            compat_urllib_parse_urlencode(login_form).encode('utf-8')
         )
         )
-        login_page = self._download_webpage(request, None, note='Logging in as %s' % username)
+        login_page = self._download_webpage(request, None, note='Logging in')
 
         if re.search(r'onLoginFailed', login_page):
             raise ExtractorError('Unable to login, incorrect username and/or password', expected=True)
 
         if re.search(r'onLoginFailed', login_page):
             raise ExtractorError('Unable to login, incorrect username and/or password', expected=True)
@@ -65,36 +93,64 @@ class PatreonIE(InfoExtractor):
     '''
 
     def _real_extract(self, url):
     '''
 
     def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        video_id = mobj.group(1)
-
-        webpage = self._download_webpage(url, video_id)
-        title = self._og_search_title(webpage).strip()
-
-        attach_fn = self._html_search_regex(
-            r'<div class="attach"><a target="_blank" href="([^"]+)">',
-            webpage, 'attachment URL', default=None)
-        if attach_fn is not None:
-            video_url = 'http://www.patreon.com' + attach_fn
-            thumbnail = self._og_search_thumbnail(webpage)
-            uploader = self._html_search_regex(
-                r'<strong>(.*?)</strong> is creating', webpage, 'uploader')
-        else:
-            playlist_js = self._search_regex(
-                r'(?s)new\s+jPlayerPlaylist\(\s*\{\s*[^}]*},\s*(\[.*?,?\s*\])',
-                webpage, 'playlist JSON')
-            playlist_json = js_to_json(playlist_js)
-            playlist = json.loads(playlist_json)
-            data = playlist[0]
-            video_url = self._proto_relative_url(data['mp3'])
-            thumbnail = self._proto_relative_url(data.get('cover'))
-            uploader = data.get('artist')
-
-        return {
+        video_id = self._match_id(url)
+        post = self._download_json(
+            'https://www.patreon.com/api/posts/' + video_id, video_id, query={
+                'fields[media]': 'download_url,mimetype,size_bytes',
+                'fields[post]': 'comment_count,content,embed,image,like_count,post_file,published_at,title',
+                'fields[user]': 'full_name,url',
+                'json-api-use-default-includes': 'false',
+                'include': 'media,user',
+            })
+        attributes = post['data']['attributes']
+        title = attributes['title'].strip()
+        image = attributes.get('image') or {}
+        info = {
             'id': video_id,
             'id': video_id,
-            'url': video_url,
-            'ext': 'mp3',
             'title': title,
             'title': title,
-            'uploader': uploader,
-            'thumbnail': thumbnail,
+            'description': clean_html(attributes.get('content')),
+            'thumbnail': image.get('large_url') or image.get('url'),
+            'timestamp': parse_iso8601(attributes.get('published_at')),
+            'like_count': int_or_none(attributes.get('like_count')),
+            'comment_count': int_or_none(attributes.get('comment_count')),
         }
         }
+
+        for i in post.get('included', []):
+            i_type = i.get('type')
+            if i_type == 'media':
+                media_attributes = i.get('attributes') or {}
+                download_url = media_attributes.get('download_url')
+                ext = mimetype2ext(media_attributes.get('mimetype'))
+                if download_url and ext in KNOWN_EXTENSIONS:
+                    info.update({
+                        'ext': ext,
+                        'filesize': int_or_none(media_attributes.get('size_bytes')),
+                        'url': download_url,
+                    })
+            elif i_type == 'user':
+                user_attributes = i.get('attributes')
+                if user_attributes:
+                    info.update({
+                        'uploader': user_attributes.get('full_name'),
+                        'uploader_id': str_or_none(i.get('id')),
+                        'uploader_url': user_attributes.get('url'),
+                    })
+
+        if not info.get('url'):
+            embed_url = try_get(attributes, lambda x: x['embed']['url'])
+            if embed_url:
+                info.update({
+                    '_type': 'url',
+                    'url': embed_url,
+                })
+
+        if not info.get('url'):
+            post_file = attributes['post_file']
+            ext = determine_ext(post_file.get('name'))
+            if ext in KNOWN_EXTENSIONS:
+                info.update({
+                    'ext': ext,
+                    'url': post_file['url'],
+                })
+
+        return info
Backup clone of https://github.com/ytdl-org unavailable since RIAA fucking raid...