From 1154728b616dbcf589c6f4eb1402fce869540644 Mon Sep 17 00:00:00 2001
From: jqtmviyu <jqtmviyu@gmail.com>
Date: Wed, 28 Aug 2024 23:05:47 +0800
Subject: [PATCH] feat: :sparkles: add china isp and update readme

---
 README.md            |  2 ++
 process_blacklist.py | 45 ++++++++++++++++++++++----------------------
 2 files changed, 25 insertions(+), 22 deletions(-)

diff --git a/README.md b/README.md
index 974e3d6..0e6107c 100644
--- a/README.md
+++ b/README.md
@@ -48,6 +48,8 @@ PATH=/usr/sbin:/usr/bin:/sbin:/bin
 
 修改自 [PBH-BTN/BTN-Collected-Rules](https://github.com/PBH-BTN/BTN-Collected-Rules) 的all.txt, github action 每天自动更新
 
+新增常见数据中心ip段: [isp/cn](https://github.com/zealic/autorosvpn/tree/master/isp/cn)
+
 设置transmission的URL阻止清单`https://github.com/jqtmviyu/Transmission-block-xunlei-pcdn/raw/main/transmission_blacklist.gz`
 
 ### 定时更新
diff --git a/process_blacklist.py b/process_blacklist.py
index d7bfd95..ada026d 100644
--- a/process_blacklist.py
+++ b/process_blacklist.py
@@ -4,39 +4,40 @@ import re
 
 import requests
 
+output_file_name = "transmission_blacklist"
 
-def process_blacklist(url, output_file_name):
+upstream_urls = [
+    "https://raw.githubusercontent.com/PBH-BTN/BTN-Collected-Rules/main/combine/all.txt",
+    "https://raw.githubusercontent.com/zealic/autorosvpn/master/isp/cn/route-isp-chinanet.txt",  # 中国电信
+    "https://raw.githubusercontent.com/zealic/autorosvpn/master/isp/cn/route-isp-cn-cmcc.txt",  # 中国移动
+    "https://raw.githubusercontent.com/zealic/autorosvpn/master/isp/cn/route-isp-cncgroup.txt",  # 中国联通
+    "https://raw.githubusercontent.com/zealic/autorosvpn/master/isp/cn/route-isp-cn-crtc.txt",  # 中国铁通
+]
+
+processed_lines = set()
+
+for url in upstream_urls:
     response = requests.get(url)
     lines = response.text.splitlines()
-
-    processed_lines = []
-
     for line in lines:
         line = line.strip()
         if not line or line.startswith("#"):
             continue
-        # 检查是否是合法的IP地址或CIDR
-        if re.match(r"(\d{1,3}\.){3}\d{1,3}(\/\d{1,2})?$", line) or re.match(
-            r"([a-fA-F0-9:]+)(\/\d{1,3})?$", line
-        ):
+        # 过滤掉注释
+        if re.match(r"^\d", line):
             if "/" in line:
                 network = ipaddress.ip_network(line, strict=False)
                 ip_range = f"{network.network_address}-{network.broadcast_address}"
+            elif "-" in line:
+                ip_range = line
             else:
                 ip_range = f"{line}-{line}"
-            processed_lines.append(f"btn:{ip_range}")
+            processed_lines.add(f"btn:{ip_range}")
 
-    with open(f"{output_file_name}.txt", "w", encoding="utf-8") as file:
-        for line in processed_lines:
-            file.write(line + "\n")
+with open(f"{output_file_name}.txt", "w", encoding="utf-8") as file:
+    for line in processed_lines:
+        file.write(line + "\n")
 
-    with gzip.open(f"{output_file_name}.gz", "wt", encoding="utf-8") as file:
-        for line in processed_lines:
-            file.write(line + "\n")
-
-
-# 调用函数处理远程文件并压缩为.gz格式
-process_blacklist(
-    "https://raw.githubusercontent.com/PBH-BTN/BTN-Collected-Rules/main/combine/all.txt",
-    "transmission_blacklist",
-)
+with gzip.open(f"{output_file_name}.gz", "wt", encoding="utf-8") as file:
+    for line in processed_lines:
+        file.write(line + "\n")