✨ feat(http_utils): 重构网络请求工具链,增强稳定性与易用性 (#1951)

* ✨ feat(http_utils): 重构网络请求工具链,增强稳定性与易用性

🔧 HTTP工具优化:
  • 全局httpx.AsyncClient管理,提升连接复用效率
  • AsyncHttpx类重构,支持临时客户端和配置覆盖
  • 新增JSON请求方法(get_json/post_json),内置重试机制
  • 兼容httpx>=0.28.0版本

🔄 重试机制升级:
  • Retry装饰器重构,提供simple/api/download预设
  • 支持指数退避、条件重试和自定义失败处理
  • 扩展异常覆盖范围,提升网络容错能力

🏗️ 架构改进:
  • 新增AllURIsFailedError统一异常处理
  • 浏览器工具模块化,提升代码组织性

* 🚨 auto fix by pre-commit hooks

* 🎨 代码格式化

* 🐛 测试修复

---------

Co-authored-by: webjoin111 <455457521@qq.com>
Co-authored-by: pre-commit-ci[bot] <66853113+pre-commit-ci[bot]@users.noreply.github.com>
Co-authored-by: HibiKier <45528451+HibiKier@users.noreply.github.com>
Co-authored-by: HibiKier <775757368@qq.com>
This commit is contained in:
Rumio
2025-07-03 17:39:13 +08:00
committed by GitHub
co-authored by webjoin111 pre-commit-ci[bot] HibiKier HibiKier
parent e0773174d0
commit 1c5f66beee
5 changed files with 771 additions and 331 deletions
+84 -81
View File
@@ -1,91 +1,94 @@
import os
import sys
from collections.abc import AsyncGenerator
from contextlib import asynccontextmanager
from pathlib import Path
from typing import Any, Literal
from nonebot import get_driver
from playwright.__main__ import main
from playwright.async_api import Browser, Playwright, async_playwright
from nonebot_plugin_alconna import UniMessage
from nonebot_plugin_htmlrender import get_browser
from playwright.async_api import Page
from zhenxun.configs.config import BotConfig
from zhenxun.services.log import logger
driver = get_driver()
_playwright: Playwright | None = None
_browser: Browser | None = None
from zhenxun.utils.message import MessageUtils
# @driver.on_startup
# async def start_browser():
# global _playwright
# global _browser
# install()
# await check_playwright_env()
# _playwright = await async_playwright().start()
# _browser = await _playwright.chromium.launch()
class BrowserIsNone(Exception):
pass
# @driver.on_shutdown
# async def shutdown_browser():
# if _browser:
# await _browser.close()
# if _playwright:
# await _playwright.stop() # type: ignore
class AsyncPlaywright:
@classmethod
@asynccontextmanager
async def new_page(
cls, cookies: list[dict[str, Any]] | dict[str, Any] | None = None, **kwargs
) -> AsyncGenerator[Page, None]:
"""获取一个新页面
# def get_browser() -> Browser:
# if not _browser:
# raise RuntimeError("playwright is not initalized")
# return _browser
def install():
"""自动安装、更新 Chromium"""
def set_env_variables():
os.environ["PLAYWRIGHT_DOWNLOAD_HOST"] = (
"https://npmmirror.com/mirrors/playwright/"
)
if BotConfig.system_proxy:
os.environ["HTTPS_PROXY"] = BotConfig.system_proxy
def restore_env_variables():
os.environ.pop("PLAYWRIGHT_DOWNLOAD_HOST", None)
if BotConfig.system_proxy:
os.environ.pop("HTTPS_PROXY", None)
if original_proxy is not None:
os.environ["HTTPS_PROXY"] = original_proxy
def try_install_chromium():
参数:
cookies: cookies
"""
browser = await get_browser()
ctx = await browser.new_context(**kwargs)
if cookies:
if isinstance(cookies, dict):
cookies = [cookies]
await ctx.add_cookies(cookies) # type: ignore
page = await ctx.new_page()
try:
sys.argv = ["", "install", "chromium"]
main()
except SystemExit as e:
return e.code == 0
return False
yield page
finally:
await page.close()
await ctx.close()
logger.info("检查 Chromium 更新")
@classmethod
async def screenshot(
cls,
url: str,
path: Path | str,
element: str | list[str],
*,
wait_time: int | None = None,
viewport_size: dict[str, int] | None = None,
wait_until: (
Literal["domcontentloaded", "load", "networkidle"] | None
) = "networkidle",
timeout: float | None = None,
type_: Literal["jpeg", "png"] | None = None,
user_agent: str | None = None,
cookies: list[dict[str, Any]] | dict[str, Any] | None = None,
**kwargs,
) -> UniMessage | None:
"""截图,该方法仅用于简单快捷截图,复杂截图请操作 page
original_proxy = os.environ.get("HTTPS_PROXY")
set_env_variables()
success = try_install_chromium()
if not success:
logger.info("Chromium 更新失败,尝试从原始仓库下载,速度较慢")
os.environ["PLAYWRIGHT_DOWNLOAD_HOST"] = ""
success = try_install_chromium()
restore_env_variables()
if not success:
raise RuntimeError("未知错误,Chromium 下载失败")
async def check_playwright_env():
"""检查 Playwright 依赖"""
logger.info("检查 Playwright 依赖")
try:
async with async_playwright() as p:
await p.chromium.launch()
except Exception as e:
raise ImportError("加载失败,Playwright 依赖不全,") from e
参数:
url: 网址
path: 存储路径
element: 元素选择
wait_time: 等待截取超时时间
viewport_size: 窗口大小
wait_until: 等待类型
timeout: 超时限制
type_: 保存类型
user_agent: user_agent
cookies: cookies
"""
if viewport_size is None:
viewport_size = {"width": 2560, "height": 1080}
if isinstance(path, str):
path = Path(path)
wait_time = wait_time * 1000 if wait_time else None
element_list = [element] if isinstance(element, str) else element
async with cls.new_page(
cookies,
viewport=viewport_size,
user_agent=user_agent,
**kwargs,
) as page:
await page.goto(url, timeout=timeout, wait_until=wait_until)
card = page
for e in element_list:
if not card:
return None
card = await card.wait_for_selector(e, timeout=wait_time)
if card:
await card.screenshot(path=path, timeout=timeout, type=type_)
return MessageUtils.build_message(path)
return None