first commit

This commit is contained in:
zmq
2026-05-26 14:16:16 +08:00
committed by Jerry Yan
commit 535d7f4ac5
40 changed files with 4571 additions and 0 deletions
+237
View File
@@ -0,0 +1,237 @@
import os
from pathlib import Path
from typing import Tuple
from playwright.async_api import async_playwright, Playwright, BrowserContext
from loguru import logger
from src.config.settings import get_settings
from src.config.accounts import AccountConfig
# 注入到所有页面的 JS 脚本,用于绕过 Cloudflare、Akamai 等自动化检测
STEALTH_JS = """
// 1. 隐藏 navigator.webdriver
Object.defineProperty(navigator, 'webdriver', {
get: () => undefined
});
// 2. 覆盖语言列表,优先日语(与目标地区一致)
Object.defineProperty(navigator, 'languages', {
get: () => ['ja-JP', 'ja', 'en-US', 'en']
});
// 3. 伪造 WebGL 渲染器信息,伪装为标准硬件
const getParameter = WebGLRenderingContext.prototype.getParameter;
WebGLRenderingContext.prototype.getParameter = function(parameter) {
// UNMASKED_VENDOR_WEBGL (37445)
if (parameter === 37445) {
return 'Intel Inc.';
}
// UNMASKED_RENDERER_WEBGL (37446)
if (parameter === 37446) {
return 'Intel(R) UHD Graphics';
}
return getParameter.apply(this, arguments);
};
// 4. 伪造设备规格
Object.defineProperty(navigator, 'hardwareConcurrency', { get: () => 8 });
Object.defineProperty(navigator, 'deviceMemory', { get: () => 8 });
// 5. 模拟完整插件列表(无头浏览器常见特征为空列表)
const mockPlugins = () => {
const makePlugin = (name, filename, description) => {
const p = Object.create(Plugin.prototype);
Object.defineProperties(p, {
name: { get: () => name },
filename: { get: () => filename },
description: { get: () => description },
length: { get: () => 0 }
});
return p;
};
const plugins = [
makePlugin("PDF Viewer", "internal-pdf-viewer", "Portable Document Format"),
makePlugin("Chrome PDF Viewer", "mhjfbgoacfdegokjbbpboihdaibdccjg", "Chrome PDF Viewer"),
makePlugin("Chromium PDF Viewer", "internal-pdf-viewer", "Chromium PDF Viewer")
];
const pList = Object.create(PluginArray.prototype);
Object.defineProperties(pList, {
length: { get: () => plugins.length },
item: { value: (index) => plugins[index] },
namedItem: { value: (name) => plugins.find(p => p.name === name) }
});
for (let i = 0; i < plugins.length; i++) {
Object.defineProperty(pList, i, { get: () => plugins[i] });
}
return pList;
};
Object.defineProperty(navigator, 'plugins', {
get: () => mockPlugins()
});
// 6. 模拟 window.chrome(无头浏览器通常没有,Cloudflare 会检测)
window.chrome = {
app: {
isInstalled: false,
InstallState: {
DISABLED: 'disabled',
INSTALLED: 'installed',
NOT_INSTALLED: 'not_installed'
},
RunningState: {
CANNOT_RUN: 'cannot_run',
READY_TO_RUN: 'ready_to_run',
RUNNING: 'running'
}
},
runtime: {
OnInstalledReason: {
CHROME_UPDATE: 'chrome_update',
INSTALL: 'install',
SHARED_MODULE_UPDATE: 'shared_module_update',
UPDATE: 'update'
},
OnRestartRequiredReason: {
APP_UPDATE: 'app_update',
OS_UPDATE: 'os_update',
PERIODIC: 'periodic'
},
PlatformArch: {
ARM: 'arm',
ARM64: 'arm64',
MIPS: 'mips',
MIPS64: 'mips64',
X86_32: 'x86-32',
X86_64: 'x86-64'
},
PlatformNaclArch: {
ARM: 'arm',
MIPS: 'mips',
MIPS64: 'mips64',
X86_32: 'x86-32',
X86_64: 'x86-64'
},
PlatformOs: {
ANDROID: 'android',
CROS: 'cros',
LINUX: 'linux',
MAC: 'mac',
OPENBSD: 'openbsd',
WIN: 'win'
},
RequestUpdateCheckStatus: {
NO_UPDATE: 'no_update',
THROTTLED: 'throttled',
UPDATE_AVAILABLE: 'update_available'
}
}
};
// 7. 模拟 Permissions 查询行为一致性
const originalQuery = window.navigator.permissions.query;
window.navigator.permissions.query = (parameters) => (
parameters.name === 'notifications' ?
Promise.resolve({ state: Notification.permission }) :
originalQuery(parameters)
);
"""
class BrowserManager:
"""
浏览器管理器:负责初始化 Playwright 上下文。
按账号隔离数据环境(user_data_dir),并注入反爬脚本以规避 Cloudflare、Akamai 等检测。
"""
def __init__(self) -> None:
self.settings = get_settings()
async def get_context(self, account: AccountConfig) -> Tuple[Playwright, BrowserContext]:
"""
为指定账号启动持久化浏览器上下文。
返回 (Playwright, BrowserContext) 元组。
调用方必须在 try-finally 中确保关闭 context 与 playwright 实例。
Args:
account: 包含 account_id、user_data_dir 及可选 proxy 的账号配置。
Returns:
Tuple[Playwright, BrowserContext]: 已启动的驱动与隔离浏览器上下文实例。
"""
# 1. 解析并确保本地持久化目录存在
data_dir = Path(account.user_data_dir)
try:
data_dir.mkdir(parents=True, exist_ok=True)
# logger.debug(f"持久化用户数据目录: {data_dir.absolute()}")
except Exception as e:
logger.error(f"创建账号 '{account.account_id}' 的持久化目录失败: {e}")
raise e
# 2. 配置隐身 Chromium 启动参数
chrome_args = [
"--disable-blink-features=AutomationControlled", # 浏览器层面移除 navigator.webdriver
"--disable-infobars", # 隐藏「Chrome 正受到自动测试软件控制」提示条
"--no-sandbox", # Docker/服务器环境兼容
"--disable-setuid-sandbox",
"--disable-dev-shm-usage", # 内存受限环境避免资源崩溃
"--start-maximized", # 启动时全屏(最大化)
"--disable-gpu", # 服务器无头模式稳定渲染
"--use-mock-keychain", # 避免本地密钥环弹窗(macOS/Linux)
"--password-store=basic",
]
executable_path_raw = self.settings.browser_executable_path
executable_path = None
if executable_path_raw:
candidate_path = Path(executable_path_raw).expanduser()
if candidate_path.is_file():
executable_path = str(candidate_path)
logger.info(f"使用自定义浏览器可执行文件: {executable_path}")
else:
logger.warning(
f"配置的浏览器可执行文件无效,将使用 Playwright 默认浏览器: {executable_path_raw}"
)
headless_mode = self.settings.browser_headless
# 使用较新的 Windows Chrome User-Agent
user_agent = (
"Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 "
"(KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36"
)
logger.info(f"正在为账号 {account.account_id} 启动 Playwright | 无头模式: {headless_mode}")
# 3. 按账号配置可选代理(服务器 IP 段绕过 Cloudflare 封锁时常用)
proxy_config = None
if account.proxy:
proxy_config = {"server": account.proxy}
logger.info(f"账号 '{account.account_id}' 已启用代理: {account.proxy}")
playwright_driver = await async_playwright().start()
try:
# 4. 使用隔离目录创建持久化上下文
launch_options = {
"user_data_dir": str(data_dir.absolute()),
"headless": headless_mode,
"args": chrome_args,
"user_agent": user_agent,
"no_viewport": True, # 配合 --start-maximized 使得网页视口也铺满全屏
"locale": "ja-JP", # 区域语言一致
"timezone_id": "Asia/Tokyo", # 日本本地时区
"accept_downloads": True, # 允许自动下载(如需要)
"proxy": proxy_config # 住宅代理等
}
if executable_path:
launch_options["executable_path"] = executable_path
context = await playwright_driver.chromium.launch_persistent_context(**launch_options)
# 5. 注入高级反指纹隐身脚本
await context.add_init_script(STEALTH_JS)
logger.success(f"账号 '{account.account_id}' 的 Playwright 上下文启动成功。")
return playwright_driver, context
except Exception as e:
logger.error(f"账号 '{account.account_id}' 启动持久化上下文失败: {e}")
# 启动失败时清理驱动进程
await playwright_driver.stop()
raise e