[instagram] Add base extractor for playlists and tag extractor
authorjhwgh1968 <jhwgh1968@users.noreply.github.com>
Sun, 20 Jan 2019 09:10:46 +0000 (09:10 +0000)
committerSergey M <dstftw@gmail.com>
Sun, 20 Jan 2019 09:10:46 +0000 (16:10 +0700)
youtube_dl/extractor/extractors.py
youtube_dl/extractor/instagram.py

index de38c6641acb6730c73d3d3975ed2faca1e70914..24361def403dbeb96ea7fd9fcf9f4df7b9c471e8 100644 (file)
@@ -494,7 +494,11 @@ from .ina import InaIE
 from .inc import IncIE
 from .indavideo import IndavideoEmbedIE
 from .infoq import InfoQIE
-from .instagram import InstagramIE, InstagramUserIE
+from .instagram import (
+    InstagramIE,
+    InstagramUserIE,
+    InstagramTagIE,
+)
 from .internazionale import InternazionaleIE
 from .internetvideoarchive import InternetVideoArchiveIE
 from .iprima import IPrimaIE
index 7e0e838f05a5e4a527cd9cf89e84f884ac1b2498..ffd87b55f6d8bc78ab7853892ce92454b89bfe8e 100644 (file)
@@ -227,44 +227,37 @@ class InstagramIE(InfoExtractor):
         }
 
 
-class InstagramUserIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<id>[^/]{2,})/?(?:$|[?#])'
-    IE_DESC = 'Instagram user profile'
-    IE_NAME = 'instagram:user'
-    _TEST = {
-        'url': 'https://instagram.com/porsche',
-        'info_dict': {
-            'id': 'porsche',
-            'title': 'porsche',
-        },
-        'playlist_count': 5,
-        'params': {
-            'extract_flat': True,
-            'skip_download': True,
-            'playlistend': 5,
-        }
-    }
+class InstagramPlaylistIE(InfoExtractor):
+    # A superclass for handling any kind of query based on GraphQL which
+    # results in a playlist.
+
+    _gis_tmpl = None  # used to cache GIS request type
 
-    _gis_tmpl = None
+    def _parse_graphql(self, webpage, item_id):
+        # Reads a webpage and returns its GraphQL data.
+        return self._parse_json(
+            self._search_regex(
+                r'sharedData\s*=\s*({.+?})\s*;\s*[<\n]', webpage, 'data'),
+            item_id)
 
-    def _entries(self, data):
+    def _extract_graphql(self, data, url):
+        # Parses GraphQL queries containing videos and generates a playlist.
         def get_count(suffix):
             return int_or_none(try_get(
                 node, lambda x: x['edge_media_' + suffix]['count']))
 
-        uploader_id = data['entry_data']['ProfilePage'][0]['graphql']['user']['id']
+        uploader_id = self._match_id(url)
         csrf_token = data['config']['csrf_token']
         rhx_gis = data.get('rhx_gis') or '3c7ca9dcefcf966d11dacf1f151335e8'
 
-        self._set_cookie('instagram.com', 'ig_pr', '1')
-
         cursor = ''
         for page_num in itertools.count(1):
-            variables = json.dumps({
-                'id': uploader_id,
+            variables = {
                 'first': 12,
                 'after': cursor,
-            })
+            }
+            variables.update(self._query_vars_for(data))
+            variables = json.dumps(variables)
 
             if self._gis_tmpl:
                 gis_tmpls = [self._gis_tmpl]
@@ -276,21 +269,26 @@ class InstagramUserIE(InfoExtractor):
                     '%s:%s:%s' % (rhx_gis, csrf_token, std_headers['User-Agent']),
                 ]
 
+            # try all of the ways to generate a GIS query, and not only use the
+            # first one that works, but cache it for future requests
             for gis_tmpl in gis_tmpls:
                 try:
-                    media = self._download_json(
+                    json_data = self._download_json(
                         'https://www.instagram.com/graphql/query/', uploader_id,
                         'Downloading JSON page %d' % page_num, headers={
                             'X-Requested-With': 'XMLHttpRequest',
                             'X-Instagram-GIS': hashlib.md5(
                                 ('%s:%s' % (gis_tmpl, variables)).encode('utf-8')).hexdigest(),
                         }, query={
-                            'query_hash': '42323d64886122307be10013ad2dcc44',
+                            'query_hash': self._QUERY_HASH,
                             'variables': variables,
-                        })['data']['user']['edge_owner_to_timeline_media']
+                        })
+                    media = self._parse_timeline_from(json_data)
                     self._gis_tmpl = gis_tmpl
                     break
                 except ExtractorError as e:
+                    # if it's an error caused by a bad query, and there are
+                    # more GIS templates to try, ignore it and keep trying
                     if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
                         if gis_tmpl != gis_tmpls[-1]:
                             continue
@@ -348,14 +346,80 @@ class InstagramUserIE(InfoExtractor):
                 break
 
     def _real_extract(self, url):
-        username = self._match_id(url)
-
-        webpage = self._download_webpage(url, username)
+        user_or_tag = self._match_id(url)
+        webpage = self._download_webpage(url, user_or_tag)
+        data = self._parse_graphql(webpage, user_or_tag)
 
-        data = self._parse_json(
-            self._search_regex(
-                r'sharedData\s*=\s*({.+?})\s*;\s*[<\n]', webpage, 'data'),
-            username)
+        self._set_cookie('instagram.com', 'ig_pr', '1')
 
         return self.playlist_result(
-            self._entries(data), username, username)
+            self._extract_graphql(data, url), user_or_tag, user_or_tag)
+
+
+class InstagramUserIE(InstagramPlaylistIE):
+    _VALID_URL = r'https?://(?:www\.)?instagram\.com/(?P<id>[^/]{2,})/?(?:$|[?#])'
+    IE_DESC = 'Instagram user profile'
+    IE_NAME = 'instagram:user'
+    _TEST = {
+        'url': 'https://instagram.com/porsche',
+        'info_dict': {
+            'id': 'porsche',
+            'title': 'porsche',
+        },
+        'playlist_count': 5,
+        'params': {
+            'extract_flat': True,
+            'skip_download': True,
+            'playlistend': 5,
+        }
+    }
+
+    _QUERY_HASH = '42323d64886122307be10013ad2dcc44',
+
+    @staticmethod
+    def _parse_timeline_from(data):
+        # extracts the media timeline data from a GraphQL result
+        return data['data']['user']['edge_owner_to_timeline_media']
+
+    @staticmethod
+    def _query_vars_for(data):
+        # returns a dictionary of variables to add to the timeline query based
+        # on the GraphQL of the original page
+        return {
+            'id': data['entry_data']['ProfilePage'][0]['graphql']['user']['id']
+        }
+
+
+class InstagramTagIE(InstagramPlaylistIE):
+    _VALID_URL = r'https?://(?:www\.)?instagram\.com/explore/tags/(?P<id>[^/]+)'
+    IE_DESC = 'Instagram hashtag search'
+    IE_NAME = 'instagram:tag'
+    _TEST = {
+        'url': 'https://instagram.com/explore/tags/lolcats',
+        'info_dict': {
+            'id': 'lolcats',
+            'title': 'lolcats',
+        },
+        'playlist_count': 50,
+        'params': {
+            'extract_flat': True,
+            'skip_download': True,
+            'playlistend': 50,
+        }
+    }
+
+    _QUERY_HASH = 'f92f56d47dc7a55b606908374b43a314',
+
+    @staticmethod
+    def _parse_timeline_from(data):
+        # extracts the media timeline data from a GraphQL result
+        return data['data']['hashtag']['edge_hashtag_to_media']
+
+    @staticmethod
+    def _query_vars_for(data):
+        # returns a dictionary of variables to add to the timeline query based
+        # on the GraphQL of the original page
+        return {
+            'tag_name':
+                data['entry_data']['TagPage'][0]['graphql']['hashtag']['name']
+        }