[postprocessor,cleanup] Create `_download_json`

[yt-dlp.git] / yt_dlp / postprocessor / common.py

diff --git a/yt_dlp/postprocessor/common.py b/yt_dlp/postprocessor/common.py
index d761c9303b1646b2a06e52876b2d6ceaff27ea15..8420ee86417c23954639711570d28309130b4b9e 100644 (file)
--- a/yt_dlp/postprocessor/common.py
+++ b/yt_dlp/postprocessor/common.py
@@ -1,13 +1,18 @@
 from __future__ import unicode_literals
 
 import functools
+import itertools
+import json
 import os
+import time
+import urllib.error
 
-from ..compat import compat_str
 from ..utils import (
     _configuration_args,
     encodeFilename,
+    network_exceptions,
     PostProcessingError,
+    sanitized_Request,
     write_string,
 )
 
@@ -63,7 +68,7 @@ def __init__(self, downloader=None):
     @classmethod
     def pp_key(cls):
         name = cls.__name__[:-2]
-        return compat_str(name[6:]) if name[:6].lower() == 'ffmpeg' else name
+        return name[6:] if name[:6].lower() == 'ffmpeg' else name
 
     def to_screen(self, text, prefix=True, *args, **kwargs):
         tag = '[%s] ' % self.PP_NAME if prefix else ''
@@ -180,6 +185,28 @@ def report_progress(self, s):
             progress_template.get('postprocess-title') or 'yt-dlp %(progress._default_template)s',
             progress_dict))
 
+    def _download_json(self, url, *, expected_http_errors=(404,)):
+        # While this is not an extractor, it behaves similar to one and
+        # so obey extractor_retries and sleep_interval_requests
+        max_retries = self.get_param('extractor_retries', 3)
+        sleep_interval = self.get_param('sleep_interval_requests') or 0
+
+        self.write_debug(f'{self.PP_NAME} query: {url}')
+        for retries in itertools.count():
+            try:
+                rsp = self._downloader.urlopen(sanitized_Request(url))
+                return json.loads(rsp.read().decode(rsp.info().get_param('charset') or 'utf-8'))
+            except network_exceptions as e:
+                if isinstance(e, urllib.error.HTTPError) and e.code in expected_http_errors:
+                    return None
+                if retries < max_retries:
+                    self.report_warning(f'{e}. Retrying...')
+                    if sleep_interval > 0:
+                        self.to_screen(f'Sleeping {sleep_interval} seconds ...')
+                        time.sleep(sleep_interval)
+                    continue
+                raise PostProcessingError(f'Unable to communicate with {self.PP_NAME} API: {e}')
+
 
 class AudioConversionError(PostProcessingError):
     pass