[anvato] Improve extraction (closes #12913)

* Promote to regular shortcut based extractor
* Add mcp to access key mapping table
* Add support for embeds extraction
* Add support for anvato embeds in generic extractor

youtube_dl/extractor/anvato.py

@@ -5,6 +5,7 @@
 import hashlib
 import json
 import random
+import re
 import time
 
 from .common import InfoExtractor
@@ -16,6 +17,7 @@
     intlist_to_bytes,
     int_or_none,
     strip_jsonp,
+    unescapeHTML,
 )
 
 
@@ -26,6 +28,8 @@ def md5_text(s):
 
 
 class AnvatoIE(InfoExtractor):
+    _VALID_URL = r'anvato:(?P<access_key_or_mcp>[^:]+):(?P<id>\d+)'
+    
     # Copied from anvplayer.min.js
     _ANVACK_TABLE = {
         'nbcu_nbcd_desktop_web_prod_93d8ead38ce2024f8f544b78306fbd15895ae5e6': 'NNemUkySjxLyPTKvZRiGntBIjEyK8uqicjMakIaQ',
@@ -114,6 +118,22 @@ class AnvatoIE(InfoExtractor):
         'nbcu_nbcd_desktop_web_prod_93d8ead38ce2024f8f544b78306fbd15895ae5e6_secure': 'NNemUkySjxLyPTKvZRiGntBIjEyK8uqicjMakIaQ'
     }
 
+    _MCP_TO_ACCESS_KEY_TABLE = {
+        'qa': 'anvato_mcpqa_demo_web_stage_18b55e00db5a13faa8d03ae6e41f6f5bcb15b922',
+        'lin': 'anvato_mcp_lin_web_prod_4c36fbfd4d8d8ecae6488656e21ac6d1ac972749',
+        'univison': 'anvato_mcp_univision_web_prod_37fe34850c99a3b5cdb71dab10a417dd5cdecafa',
+        'uni': 'anvato_mcp_univision_web_prod_37fe34850c99a3b5cdb71dab10a417dd5cdecafa',
+        'dev': 'anvato_mcp_fs2go_web_prod_c7b90a93e171469cdca00a931211a2f556370d0a',
+        'sps': 'anvato_mcp_sps_web_prod_54bdc90dd6ba21710e9f7074338365bba28da336',
+        'spsstg': 'anvato_mcp_sps_web_prod_54bdc90dd6ba21710e9f7074338365bba28da336',
+        'anv': 'anvato_mcp_anv_web_prod_791407490f4c1ef2a4bcb21103e0cb1bcb3352b3',
+        'gray': 'anvato_mcp_gray_web_prod_4c10f067c393ed8fc453d3930f8ab2b159973900',
+        'hearst': 'anvato_mcp_hearst_web_prod_5356c3de0fc7c90a3727b4863ca7fec3a4524a99',
+        'cbs': 'anvato_mcp_cbs_web_prod_02f26581ff80e5bda7aad28226a8d369037f2cbe',
+        'telemundo': 'anvato_mcp_telemundo_web_prod_c5278d51ad46fda4b6ca3d0ea44a7846a054f582'
+    }
+
+    _ANVP_RE = r'<script[^>]+\bdata-anvp\s*=\s*(["\'])(?P<anvp>(?:(?!\1).)+)\1'
     _AUTH_KEY = b'\x31\xc2\x42\x84\x9e\x73\xa0\xce'
 
     def __init__(self, *args, **kwargs):
@@ -217,9 +237,42 @@ def _get_anvato_videos(self, access_key, video_id):
             'subtitles': subtitles,
         }
 
+    @staticmethod
+    def _extract_urls(ie, webpage, video_id):
+        entries = []
+        for mobj in re.finditer(AnvatoIE._ANVP_RE, webpage):
+            anvplayer_data = ie._parse_json(
+                mobj.group('anvp'), video_id, transform_source=unescapeHTML,
+                fatal=False)
+            if not anvplayer_data:
+                continue
+            video = anvplayer_data.get('video')
+            if not isinstance(video, compat_str) or not video.isdigit():
+                continue
+            access_key = anvplayer_data.get('accessKey')
+            if not access_key:
+                mcp = anvplayer_data.get('mcp')
+                if mcp:
+                    access_key = AnvatoIE._MCP_TO_ACCESS_KEY_TABLE.get(
+                        mcp.lower())
+            if not access_key:
+                continue
+            entries.append(ie.url_result(
+                'anvato:%s:%s' % (access_key, video), ie=AnvatoIE.ie_key(),
+                video_id=video))
+        return entries
+
     def _extract_anvato_videos(self, webpage, video_id):
-        anvplayer_data = self._parse_json(self._html_search_regex(
-            r'<script[^>]+data-anvp=\'([^\']+)\'', webpage,
-            'Anvato player data'), video_id)
+        anvplayer_data = self._parse_json(
+            self._html_search_regex(
+                self._ANVP_RE, webpage, 'Anvato player data', group='anvp'),
+            video_id)
         return self._get_anvato_videos(
             anvplayer_data['accessKey'], anvplayer_data['video'])
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        access_key, video_id = mobj.group('access_key_or_mcp', 'id')
+        if access_key not in self._ANVACK_TABLE:
+            access_key = self._MCP_TO_ACCESS_KEY_TABLE[access_key]
+        return self._get_anvato_videos(access_key, video_id)

youtube_dl/extractor/extractors.py

@@ -41,6 +41,7 @@
 from .amcnetworks import AMCNetworksIE
 from .animeondemand import AnimeOnDemandIE
 from .anitube import AnitubeIE
+from .anvato import AnvatoIE
 from .anysex import AnySexIE
 from .aol import AolIE
 from .allocine import AllocineIE

youtube_dl/extractor/generic.py

@@ -86,6 +86,7 @@
 from .videopress import VideoPressIE
 from .rutube import RutubeIE
 from .limelight import LimelightBaseIE
+from .anvato import AnvatoIE
 
 
 class GenericIE(InfoExtractor):
@@ -1677,6 +1678,15 @@ class GenericIE(InfoExtractor):
             },
             'playlist_mincount': 5,
         },
+        {
+            'url': 'http://kron4.com/2017/04/28/standoff-with-walnut-creek-murder-suspect-ends-with-arrest/',
+            'info_dict': {
+                'id': 'standoff-with-walnut-creek-murder-suspect-ends-with-arrest',
+                'title': 'Standoff with Walnut Creek murder suspect ends',
+                'description': 'md5:3ccc48a60fc9441eeccfc9c469ebf788',
+            },
+            'playlist_mincount': 4,
+        },
         # {
         #     # TODO: find another test
         #     # http://schema.org/VideoObject
@@ -2537,6 +2547,12 @@ def _real_extract(self, url):
                 'limelight:media:%s' % mobj.group('id'),
                 {'source_url': url}), 'LimelightMedia', mobj.group('id'))
 
+        # Look for Anvato embeds
+        anvato_urls = AnvatoIE._extract_urls(self, webpage, video_id)
+        if anvato_urls:
+            return self.playlist_result(
+                anvato_urls, video_id, video_title, video_description)
+
         # Look for AdobeTVVideo embeds
         mobj = re.search(
             r'<iframe[^>]+src=[\'"]((?:https?:)?//video\.tv\.adobe\.com/v/\d+[^"]+)[\'"]',