62 lines
1.9 KiB
Python
62 lines
1.9 KiB
Python
import sys
|
|
import os
|
|
sys.path.append(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
|
|
|
from core.crawler import BasePlugin
|
|
from core.log import logger
|
|
import json
|
|
import asyncio
|
|
|
|
class Fate0Plugin(BasePlugin):
|
|
def __init__(self):
|
|
super().__init__()
|
|
self.name = "Fate0聚合源"
|
|
# 这是一个持续更新的高质量代理聚合列表
|
|
self.urls = ["https://raw.githubusercontent.com/fate0/proxylist/master/proxy.list"]
|
|
|
|
async def parse(self, html):
|
|
if not html:
|
|
return
|
|
|
|
count = 0
|
|
# fate0 的数据格式是每行一个 JSON 对象
|
|
for line in html.split('\n'):
|
|
if not line.strip():
|
|
continue
|
|
try:
|
|
data = json.loads(line)
|
|
ip = data.get('host')
|
|
port = data.get('port')
|
|
protocol = data.get('type', 'http')
|
|
|
|
if ip and port:
|
|
yield ip, int(port), protocol
|
|
count += 1
|
|
except Exception:
|
|
continue
|
|
|
|
if count > 0:
|
|
logger.info(f"{self.name} 解析完成，获得 {count} 个潜在代理")
|
|
|
|
|
|
if __name__ == "__main__":
|
|
async def test_plugin():
|
|
plugin = Fate0Plugin()
|
|
print(f"========== 测试 {plugin.name} ==========")
|
|
print(f"目标URL数量: {len(plugin.urls)}")
|
|
print(f"开始抓取...\n")
|
|
|
|
proxies = await plugin.run()
|
|
|
|
print(f"\n========== 抓取结果 ==========")
|
|
print(f"总计获取 {len(proxies)} 个代理:")
|
|
print("-" * 60)
|
|
|
|
for idx, (ip, port, protocol) in enumerate(proxies, 1):
|
|
print(f"{idx:3d}. {ip:15s} : {str(port):5s} | {protocol}")
|
|
|
|
print("-" * 60)
|
|
print(f"完成！共 {len(proxies)} 个代理~")
|
|
|
|
asyncio.run(test_plugin())
|