[HELP] Unable to share state between the main app and a plugin

[ProtocolNebula]

Hi,

I'm trying to do a plugin to do a "rotate proxy", but the next IP must be rotated by the main app, not the plugin itself.

I have the following code:

import proxy
from proxy.http.proxy import HttpProxyBasePlugin

class CustomProxyRedirection(HttpProxyBasePlugin):
  currentProxy = {
    "address": "0.0.0.0"
  }
  
  def __init__(self, *args, **kwargs):
    super().__init__(*args, **kwargs)
    
  def before_upstream_connection(self, request):
    print(f'Proxy host: {CustomProxyRedirection.currentProxy['address']}')
    proxy_host, proxy_port = CustomProxyRedirection.currentProxy['address'].split(':')
    # Set the proxy for the request
    request.host = proxy_host
    request.port = int(proxy_port)
    return request


class ProxyManager:
  proxy = None
  def __init__(self):
    self.proxy = proxy.Proxy(port=8899, plugins=[
      CustomProxyRedirection,
    ])
  def nextProxy(self):
    CustomProxyRedirection.currentProxy['address'] ="154.94.5.241:7001"
    # EXTREME CASE WHERE I INSTANTIATE THE WHOLE PROXY AGAIN
    self.proxy = proxy.Proxy(port=8899, plugins=[
      CustomProxyRedirection,
    ])




# Example to run the proxy
def main():
    proxyManager = ProxyManager()
    proxyManager.nextProxy()
    with proxyManager.proxy as p:
        logger.info('Application started')
        proxy.sleep_loop()

if __name__ == '__main__':
    main()

I've tried different solutions, even I tried to understand how to use the multiprocess manager, but I found no way to do it.

Someone can please help me with an example or an explanation of how can I do it? (I assume that I have to spawn by myself the proxy, but I'm too new to python and found no clues/information of how to manage this case).

Thanks.

[abhinavsingh]

1. If multiprocessing state sharing feels tricky, simply use an external storage like file on Disk, Sqlite, Memcache, Redis etc for sharing states.
2. You may use Python's inbuilt multiprocessing shared variables for the same too