From 00370818f0caa7d87c7e596618e4bf39b04055ae Mon Sep 17 00:00:00 2001
From: CaliBrain <calibrain@l4n.xyz>
Date: Thu, 28 Aug 2025 17:42:50 -0400
Subject: [PATCH] Fix eager cloudflare check (#247)

---
 book_manager.py        |  3 ++-
 cloudflare_bypasser.py | 34 +++++++++++++++-------------------
 requirements-base.txt  |  1 +
 3 files changed, 18 insertions(+), 20 deletions(-)

diff --git a/book_manager.py b/book_manager.py
index aef28851..ea6baa37 100644
--- a/book_manager.py
+++ b/book_manager.py
@@ -199,7 +199,8 @@ def _parse_book_info_page(soup: BeautifulSoup, book_id: str) -> BookInfo:
             ):
                 libgen_url = url["href"]
                 # TODO : Temporary fix ? Maybe get URLs from https://open-slum.org/ ?
-                libgen_url = libgen_url = re.sub(r'libgen\.(\w+)', 'libgen.la', url["href"])
+                libgen_url = re.sub(r'libgen\.(lc|is|bz|st)', 'libgen.gl', url["href"])
+
                 external_urls_libgen.add(libgen_url)
             elif url.text.strip().lower().startswith("z-lib"):
                 if ".onion/" not in url["href"]:
diff --git a/cloudflare_bypasser.py b/cloudflare_bypasser.py
index 5f64aeac..47c5a836 100644
--- a/cloudflare_bypasser.py
+++ b/cloudflare_bypasser.py
@@ -45,7 +45,7 @@ def _reset_pyautogui_display_state():
     except Exception as e:
         logger.warning(f"Error resetting pyautogui display state: {e}")
 
-def _is_bypassed(sb) -> bool:
+def _is_bypassed(sb, escape_emojis : bool = True) -> bool:
     """Enhanced bypass detection with more comprehensive checks"""
     try:
         # Get page information with error handling
@@ -64,29 +64,25 @@ def _is_bypassed(sb) -> bool:
         except:
             current_url = ""
         
+        # Check if page is too long, if so we are probably bypassed
+        if len(body.strip()) > 100000:
+            logger.debug(f"Page content too long, we are probably bypassed len: {len(body.strip())}")
+            return True
+        
+        # Detect if there is an emoji in the page, any utf8 emoji, if so we are probably bypassed
+        if escape_emojis:
+            import emoji
+            emoji_list = emoji.emoji_list(text)
+            if len(emoji_list) >= 3:
+                logger.debug(f"Detected emoji in page, we are probably bypassed len: {len(emoji_list)}")
+                return True
+
         # Enhanced verification texts for newer Cloudflare versions
         verification_texts = [
             "just a moment",
             "verify you are human",
             "verifying you are human",
-            "needs to review the security of your connection before proceeding",
-            "checking your browser",
-            "checking connection",
-            "attention required",
-            "access denied",
-            "needs to review the security of your connection",
-            "checking the site connection security",
-            "enable javascript and cookies to continue",
-            "ray id",
-            "cloudflare",
-            "please wait",
-            "ddos protection",
-            "security check",
-            "moment please",
-            "hold on",
-            "loading",
-            "one more step",
-            "challenge"
+            "cloudflare.com/products/turnstile/?utm_source=turnstile"
         ]
         
         # Check for Cloudflare indicators
diff --git a/requirements-base.txt b/requirements-base.txt
index ecdacd26..242d4f06 100644
--- a/requirements-base.txt
+++ b/requirements-base.txt
@@ -6,3 +6,4 @@ dnspython
 gunicorn
 python-xlib
 psutil
+emoji