v0.2.6

YouTube created a new key 'commentCount' in addition to 'headerText'

README.md

@@ -153,7 +153,7 @@ For coding guidelines and an overview of the software architecture, see the [HAC
 
 yt-local is not made to work in public mode, however there is an instance of yt-local in public mode but with less features
 
-- <https://06ff-93-95-230-133.ngrok-free.app/https://www.youtube.com/>
+- <https://1cd1-93-95-230-133.ngrok-free.app/https://youtube.com>
 
 ## License
 

generate_release.py

@@ -18,7 +18,7 @@ if len(sys.argv) > 2:
 else:
     bitness = '64'
 
-if latest_version = 'oldwin':
+if latest_version == 'oldwin':
     bitness = '32'
     latest_version = '3.7.9'
     suffix = 'windows-vista-7-only'

youtube/__init__.py

@@ -115,7 +115,18 @@ def error_page(e):
                     error_message=exc_info()[1].error_message,
                     slim=slim
                 ), 502)
-    return flask.render_template('error.html', traceback=traceback.format_exc(), slim=slim), 500
+    elif (exc_info()[0] == util.FetchError
+        and exc_info()[1].code == '404'
+    ):
+        error_message = ('Error: The page you are looking for isn\'t here. ¯\_(ツ)_/¯')
+        return flask.render_template('error.html',
+                                     error_code=exc_info()[1].code,
+                                     error_message=error_message,
+                                     slim=slim), 404
+    return flask.render_template('error.html', traceback=traceback.format_exc(),
+                                 error_code=exc_info()[1].code,
+                                 slim=slim), 500
+    # return flask.render_template('error.html', traceback=traceback.format_exc(), slim=slim), 500
 
 
 font_choices = {

youtube/templates/error.html

@@ -1,4 +1,8 @@
-{% set page_title = 'Error' %}
+{% if error_code %}
+    {% set page_title = 'Error: ' ~ error_code %}
+{% else %}
+    {% set page_title = 'Error' %}
+{% endif %}
 
 {% if not slim %}
     {% extends "base.html" %}

youtube/version.py

@@ -1,3 +1,3 @@
 from __future__ import unicode_literals
 
-__version__ = '0.2.5'
+__version__ = '0.2.6'

youtube/watch.py

@@ -396,7 +396,7 @@ def fetch_player_response(client, video_id):
     data = {
         'videoId': video_id,
         'context': context,
-        'params': '8AEB',
+        'params': 'CgIQBg',
     }
     data = json.dumps(data)
     headers = (('Content-Type', 'application/json'),('User-Agent', user_agent))

youtube/yt_data_extract/common.py

@@ -109,7 +109,7 @@ def concat_or_none(*strings):
 def remove_redirect(url):
     if url is None:
         return None
-    if re.fullmatch(r'(((https?:)?//)?(www.)?youtube.com)?/redirect\?.*', url) is not None: # youtube puts these on external links to do tracking
+    if re.fullmatch(r'(((https?:)?//)?(www.)?youtube.com)?/redirect\?.*', url) is not None: # YouTube puts these on external links to do tracking
         query_string = url[url.find('?')+1: ]
         return urllib.parse.parse_qs(query_string)['q'][0]
     return url
@@ -133,11 +133,11 @@ def _recover_urls(runs):
     for run in runs:
         url = deep_get(run, 'navigationEndpoint', 'urlEndpoint', 'url')
         text = run.get('text', '')
-        # second condition is necessary because youtube makes other things into urls, such as hashtags, which we want to keep as text
+        # second condition is necessary because YouTube makes other things into urls, such as hashtags, which we want to keep as text
         if url is not None and (text.startswith('http://') or text.startswith('https://')):
             url = remove_redirect(url)
             run['url'] = url
-            run['text'] = url # youtube truncates the url text, use actual url instead
+            run['text'] = url # YouTube truncates the url text, use actual url instead
 
 def extract_str(node, default=None, recover_urls=False):
     '''default is the value returned if the extraction fails. If recover_urls is true, will attempt to fix YouTube's truncation of url text (most prominently seen in descriptions)'''
@@ -569,13 +569,13 @@ def extract_items(response, item_types=_item_types,
                     item_types=item_types)
                 if items:
                     break
-    elif ('onResponseReceivedEndpoints' in response
+    if ('onResponseReceivedEndpoints' in response
           or 'onResponseReceivedActions' in response):
         for endpoint in multi_get(response,
                                   'onResponseReceivedEndpoints',
                                   'onResponseReceivedActions',
                                   []):
-            items, ctoken = extract_items_from_renderer_list(
+            new_items, new_ctoken = extract_items_from_renderer_list(
                 multi_deep_get(
                     endpoint,
                     ['reloadContinuationItemsCommand', 'continuationItems'],
@@ -584,13 +584,17 @@ def extract_items(response, item_types=_item_types,
                 ),
                 item_types=item_types,
             )
-            if items:
-                break
-    elif 'contents' in response:
+            items += new_items
+            if (not ctoken) or (new_ctoken and new_items):
+                ctoken = new_ctoken
+    if 'contents' in response:
         renderer = get(response, 'contents', {})
-        items, ctoken = extract_items_from_renderer(
+        new_items, new_ctoken = extract_items_from_renderer(
             renderer,
             item_types=item_types)
+        items += new_items
+        if (not ctoken) or (new_ctoken and new_items):
+            ctoken = new_ctoken
 
     if search_engagement_panels and 'engagementPanels' in response:
         new_items, new_ctoken = extract_items_from_renderer_list(

youtube/yt_data_extract/watch_extraction.py

@@ -357,7 +357,8 @@ def _extract_watch_info_mobile(top_level):
         # https://www.androidpolice.com/2019/10/31/google-youtube-app-comment-section-below-videos/
         # https://www.youtube.com/watch?v=bR5Q-wD-6qo
         if header_type == 'commentsEntryPointHeaderRenderer':
-            comment_count_text = extract_str(comment_info.get('headerText'))
+            comment_count_text = extract_str(multi_get(
+                comment_info, 'commentCount', 'headerText'))
         else:
             comment_count_text = extract_str(deep_get(comment_info,
                 'header', 'commentSectionHeaderRenderer', 'countText'))
@@ -463,6 +464,13 @@ def _extract_formats(info, player_response):
     for yt_fmt in yt_formats:
         itag = yt_fmt.get('itag')
 
+        # Translated audio track
+        # Example: https://www.youtube.com/watch?v=gF9kkB0UWYQ
+        # Only get the original language for now so a foreign
+        # translation will not be picked just because it comes first
+        if deep_get(yt_fmt, 'audioTrack', 'audioIsDefault') is False:
+            continue
+
         fmt = {}
         fmt['itag'] = itag
         fmt['ext'] = None