From e763efd36cd3496ecd5f110f37ad905e7063aa41 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Mike=20F=C3=A4hrmann?= <mike_faehrmann@web.de>
Date: Thu, 14 Nov 2024 23:06:26 +0100
Subject: [PATCH] [bilibili] add workarounds for getting rate-limited (#6443)

- set 3-6 second request_interval by default
- retry request after waiting 5 minutes
---
 docs/configuration.rst           |  6 +++++-
 gallery_dl/extractor/bilibili.py | 15 ++++++++++++---
 2 files changed, 17 insertions(+), 4 deletions(-)

diff --git a/docs/configuration.rst b/docs/configuration.rst
index f5ef9bdd..ebbd3a0c 100644
--- a/docs/configuration.rst
+++ b/docs/configuration.rst
@@ -394,7 +394,11 @@ Default
     * ``"2.0-4.0"``
         ``behance``, ``imagefap``, ``[Nijie]``
     * ``"3.0-6.0"``
-        ``exhentai``, ``idolcomplex``, ``[reactor]``, ``readcomiconline``
+        ``bilibili``,
+        ``exhentai``,
+        ``idolcomplex``,
+        ``[reactor]``,
+        ``readcomiconline``
     * ``"6.0-6.1"``
         ``twibooru``
     * ``"6.0-12.0"``
diff --git a/gallery_dl/extractor/bilibili.py b/gallery_dl/extractor/bilibili.py
index 718bbf37..d5c419eb 100644
--- a/gallery_dl/extractor/bilibili.py
+++ b/gallery_dl/extractor/bilibili.py
@@ -14,6 +14,7 @@ class BilibiliExtractor(Extractor):
     """Base class for bilibili extractors"""
     category = "bilibili"
     root = "https://www.bilibili.com"
+    request_interval = (3.0, 6.0)
 
     def _init(self):
         self.api = BilibiliAPI(self)
@@ -102,6 +103,14 @@ class BilibiliAPI():
 
     def article(self, article_id):
         url = "https://www.bilibili.com/opus/" + article_id
-        response = self.extractor.request(url)
-        return util.json_loads(text.extr(
-            response.text, "window.__INITIAL_STATE__=", "};") + "}")
+
+        while True:
+            page = self.extractor.request(url).text
+            try:
+                return util.json_loads(text.extr(
+                    page, "window.__INITIAL_STATE__=", "};") + "}")
+            except Exception:
+                if "window._riskdata_" not in page:
+                    raise exception.StopExtraction(
+                        "%s: Unable to extract INITIAL_STATE data", article_id)
+            self.extractor.wait(seconds=300)