v2.0 - Global browser proxy feature added

Author: odell0111

.gitignore

@@ -1,2 +1,3 @@
 build/
 site-keys.txt
+*.env

README.md

@@ -5,6 +5,8 @@ Python server to automatically solve Cloudflare Turnstile CAPTCHA with an averag
 PD: This repository was initially created for personal use. I've adjusted it for sharing, but it might still be slightly disorganized. Feel free to contribute, open issues, and request new features.
 
 ## Screenshots
+
+TODO: Update
 ![Help Menu](images/help_menu.png)
 
 ![Server Console](images/server_console.png)
@@ -30,6 +32,16 @@ solver
 ```bash
 solver --port 8088 --secret jWRN7DH6 --browser-position --max-attempts 3  --captcha-timeout 30 --page-load-timeout 30 --reload-on-overrun
 ```
+#### Use global browser proxy
+```bash
+solver --proxy-server http://myproxy.com:3128 --proxy-username user --proxy-password pass
+```
+##### Load proxy parameters from environment variables (all caps)
+```bash
+solver --proxy-server MY_PROXY_SERVER --proxy-username MY_PROXY_USERNAME --proxy-password MY_PROXY_PASSWORD
+```
+
+TODO: Implement proxy rotation system
 
 ### Get token
 

pyproject.toml

@@ -6,7 +6,7 @@ build-backend = "setuptools.build_meta"
 #dynamic = ["version"]
 dynamic = ["dependencies"]
 name = "turnstile_solver"
-version = "1.21"
+version = "2.0"
 description = "Python server to automatically solve Cloudflare Turnstile CAPTCHA with an average solving time of two seconds"
 readme = "README.md"
 authors = [{ name = "OGM" }]
@@ -19,7 +19,7 @@ classifiers = [
 ]
 # keywords = [ ]
 
-requires-python = ">=3.11"
+requires-python = ">=3.10"
 
 [project.urls]
 Repository = "https://github.com/odell0111/turnstile_solver"

src/turnstile_solver/main.py

@@ -4,6 +4,7 @@
 import argparse
 import random
 import time
+import dotenv
 from pathlib import Path
 from threading import Thread
 
@@ -28,7 +29,7 @@
 __pname__ = "Turnstile Solver"
 
 # mdata = metadata.metadata(__pname__)
-__version__ = "1.21"  # mdata['Version']
+__version__ = "2.0"  # mdata['Version']
 __homepage__ = "https://github.com/odell0111/turnstile_solver"  # mdata['Home-page']
 __author__ = "OGM"  # mdata['Author']
 __summary__ = "Automatically solve Cloudflare Turnstile captcha"  # mdata['Summary']
@@ -100,9 +101,13 @@ def positive_integer_exclusive(value):
   parser.add_argument("--headless", action="store_true", help=f"Open browser in headless mode. WARNING: This feature has never worked so far, captcha always fail! It's here only in case it works on future version of Playwright.")
   parser.add_argument("-bep", "--browser-executable-path", help=f"Chromium-based browser executable path. If not specified, Patchright (Playwright) will attempt to use its bundled version. Ensure you are using a Chromium-based browser installed with the command `patchright install chromium`. Other browsers may be detected by Cloudflare, which could result in the CAPTCHA not being solved.")
   parser.add_argument("-bp", "--browser-position", type=int, nargs='*', metavar="x|y", default=c.BROWSER_POSITION, help=f"Browser position x, y. Default: {c.BROWSER_POSITION}. If the browser window is positioned beyond the screen's resolution, it will be inaccessible, behaving similar to headless mode.")
+  parser.add_argument("-ps", "--proxy-server", help=f"Global browser proxy server in the format SCHEME://IP|ADDRESS:PORT. Ex: http://myproxy.com:3128")
+  parser.add_argument("-pun", "--proxy-username", help=f"Global browser proxy username: Use all caps to load from environment variables.")
+  parser.add_argument("-pp", "--proxy-password", help=f"Global browser proxy password: Use all caps to load from environment variables.")
+
   parser.add_argument("-nfl", "--no-file-logs", action="store_true", help=f"Do not log to file '$HOME.turnstile_solver/logs.log'.")
-  #
-  # # Solver
+
+  # Solver
   solver = parser.add_argument_group("Solver")
   solver.add_argument("-ma", "--max-attempts", type=positive_integer, metavar="N", default=c.MAX_ATTEMPTS_TO_SOLVE_CAPTCHA, help=f"Max attempts to perform to solve captcha. Default: {c.MAX_ATTEMPTS_TO_SOLVE_CAPTCHA}.")
   solver.add_argument("-cto", "--captcha-timeout", type=positive_float, metavar="N.", default=c.CAPTCHA_ATTEMPT_TIMEOUT, help=f"Max time to wait for captcha to solve before reloading page. Default: {c.CAPTCHA_ATTEMPT_TIMEOUT} seconds.")
@@ -211,7 +216,9 @@ async def run_server(
     attempt_timeout: int = c.CAPTCHA_ATTEMPT_TIMEOUT,
     headless: bool = False,
     solver_log_level: int | str = logging.INFO,
-
+    proxy_server: str | None = None,
+    proxy_username: str | None = None,
+    proxy_password: str | None = None,
 ):
   server = TurnstileSolverServer(
     host=host,
@@ -236,6 +243,9 @@ async def run_server(
     headless=headless,
     console=console,
     log_level=solver_log_level,
+    proxy_server=proxy_server,
+    proxy_username=proxy_username,
+    proxy_password=proxy_password,
   )
   server.solver = solver
   await solver.server.create_page_pool()
@@ -259,6 +269,8 @@ async def run_server(
 
 async def main():
 
+  dotenv.load_dotenv()
+
   from rich import traceback
   traceback.install(
     show_locals=True,
@@ -315,6 +327,9 @@ async def main():
     attempt_timeout=args.captcha_timeout,
     headless=args.headless,
     solver_log_level=args.solver_log_level,
+    proxy_server=args.proxy_server,
+    proxy_username=args.proxy_username,
+    proxy_password=args.proxy_password,
   )
 
 

src/turnstile_solver/solver.py

@@ -1,9 +1,12 @@
 import datetime
 import logging
+import os
+import re
 import time
 from pathlib import Path
 from typing import Callable, Awaitable
 from patchright.async_api import async_playwright, Page, BrowserContext, Browser, Playwright
+from turnstile_solver.utils import is_all_caps
 
 from . import constants as c
 from .enums import CaptchaApiMessageEvent
@@ -80,6 +83,9 @@ def __init__(self,
                headless: bool = False,
                console: SolverConsole = SolverConsole(),
                log_level: int | str = logging.INFO,
+               proxy_server: str | None = None,
+               proxy_username: str | None = None,
+               proxy_password: str | None = None,
                ):
 
     logger.setLevel(log_level)
@@ -98,6 +104,13 @@ def __init__(self,
     self.max_attempts = max_attempts
     self.attempt_timeout = attempt_timeout
 
+    if bool(proxy_username) ^ bool(proxy_password):
+      raise ValueError("Proxy username and password must both be specified")
+
+    self.proxy_server = self._load_proxy_param(proxy_server)
+    self.proxy_username = self._load_proxy_param(proxy_username)
+    self.proxy_password = self._load_proxy_param(proxy_password)
+
   @property
   def _server_down(self) -> bool:
     if self.server.down:
@@ -267,10 +280,38 @@ async def _setup_page(
 
   async def _get_browser_context(self) -> tuple[BrowserContext, Playwright]:
     playwright = await async_playwright().start()
+
+    if self.proxy_server:
+      logger.debug(f"Using proxy: '{self.proxy_server}'")
+      if not re.search(r':\d+$', self.proxy_server):
+        logger.warning("No proxy port specified")
+      proxy = {
+        'server': self.proxy_server,
+      }
+      if self.proxy_username:
+        logger.debug("Using proxy credentials")
+        # logger.debug(f"Proxy Username: {self.proxy_username}")
+        # logger.debug(f"Proxy Password: {self.proxy_password}")
+        proxy['username'] = self.proxy_username
+        proxy['password'] = self.proxy_password
+    else:
+      proxy = None
+
     browser: Browser | None = await playwright.chromium.launch(
       executable_path=self.browser_executable_path,
       # channel=channel,
       args=self.browser_args,
       headless=self.headless,
+      proxy=proxy,
     )
     return await browser.new_context(), playwright
+
+  @staticmethod
+  def _load_proxy_param(param) -> str | None:
+    if is_all_caps(param):
+      if p := os.environ.get(param):
+        logger.debug("Proxy parameter loaded from environment variables")
+        return p
+      else:
+        logger.warning("Proxy parameter intended to be loaded from environment variables was not found")
+    return param

src/turnstile_solver/utils.py

@@ -55,3 +55,12 @@ def get_file_handler(
   formatter = logging.Formatter('%(asctime)s::%(levelname)s::%(name)s %(message)s, line %(lineno)d')
   handler.setFormatter(formatter)
   return handler
+
+
+def is_all_caps(word: str) -> bool:
+  if not word:
+    return False
+  filtered = [c.isupper() for c in word if c.isalpha()]
+  return filtered and all(filtered)
+
+

tests/conftest.py

@@ -4,6 +4,7 @@
 
 import logging
 import pytest
+import dotenv
 
 from turnstile_solver.solver_console import SolverConsole
 from turnstile_solver.utils import init_logger, get_file_handler
@@ -30,6 +31,8 @@ def console() -> SolverConsole:
 def pytest_configure(config: pytest.Config):
   print()
 
+  dotenv.load_dotenv()
+
   if not PROJECT_HOME_DIR.exists():
     PROJECT_HOME_DIR.mkdir(parents=True)
 

tests/test_solver.py

@@ -1,5 +1,7 @@
 import logging
 import asyncio
+import os
+
 import pytest
 import requests
 
@@ -37,6 +39,10 @@ def server(console: SolverConsole) -> TurnstileSolverServer:
 def solver(server: TurnstileSolverServer) -> TurnstileSolver:
   EXECUTABLE_PATH = Path.home() / "AppData/Local/ms-playwright/chromium-1155/chrome-win/chrome.exe"
 
+  proxyServer = os.environ.get('PROXY_SERVER')
+  proxyUsername = os.environ.get('PROXY_USERNAME')
+  proxyPassword = os.environ.get('PROXY_PASSWORD')
+
   s = TurnstileSolver(
     console=server.console,
     log_level=logging.DEBUG,
@@ -46,6 +52,9 @@ def solver(server: TurnstileSolverServer) -> TurnstileSolver:
     browser_position=None,
     browser_executable_path=EXECUTABLE_PATH,
     headless=False,
+    proxy_server=proxyServer,
+    proxy_username=proxyUsername,
+    proxy_password=proxyPassword,
   )
   server.solver = s
   return s