nonebot-plugin-htmlrender 0.4.0__tar.gz → 0.5.0__tar.gz

{nonebot_plugin_htmlrender-0.4.0 → nonebot_plugin_htmlrender-0.5.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: nonebot-plugin-htmlrender
-Version: 0.4.0
+Version: 0.5.0
 Summary: 通过浏览器渲染图片
 Project-URL: Homepage, https://github.com/kexue-z/nonebot-plugin-htmlrender
 Project-URL: Bug Tracker, https://github.com/kexue-z/nonebot-plugin-htmlrender/issues
@@ -26,7 +26,6 @@ License: MIT License
         LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
         OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
         SOFTWARE.
-License-File: LICENSE
 Requires-Python: <4.0,>=3.9
 Requires-Dist: aiofiles>=0.8.0
 Requires-Dist: jinja2>=3.0.3

{nonebot_plugin_htmlrender-0.4.0 → nonebot_plugin_htmlrender-0.5.0}/nonebot_plugin_htmlrender/__init__.py

@@ -6,6 +6,7 @@ from .browser import (
     get_browser as get_browser,
     get_new_page as get_new_page,
     shutdown_browser as shutdown_browser,
+    init_browser as init_browser,
 )
 from .data_source import (
     capture_element as capture_element,
@@ -35,9 +36,8 @@ async def init(**kwargs):
     Returns:
         Browser: Browser
     """
-    browser = await get_browser(**kwargs)
-    logger.info("Browser Started.")
-    return browser
+    logger.info("Browser Starting.")
+    return await init_browser(**kwargs)
 
 
 @driver.on_shutdown

nonebot_plugin_htmlrender-0.5.0/nonebot_plugin_htmlrender/browser.py

@@ -0,0 +1,356 @@
+import asyncio
+import os
+import socket
+import sys
+from contextlib import asynccontextmanager, contextmanager
+from dataclasses import dataclass
+from enum import Enum
+from typing import Optional, Literal, AsyncIterator, TextIO
+from urllib.parse import urlparse
+
+from nonebot.log import logger
+from playwright.async_api import Browser, Error, Page, Playwright, async_playwright
+
+from .config import plugin_config
+
+_browser: Optional[Browser] = None
+_playwright: Optional[Playwright] = None
+BrowserType = Literal["chromium", "firefox", "webkit"]
+
+
+@contextmanager
+def _suppress_and_log():
+    """抑制并记录关闭期间的任何异常。"""
+    try:
+        yield
+    except Exception as e:
+        logger.opt(exception=e).warning("关闭 playwright 时发生错误。")
+
+
+async def init_browser(**kwargs) -> Browser:
+    """初始化全局 playwright 浏览器实例并返回浏览器实例。"""
+    try:
+        return await start_browser(**kwargs)
+    except Error:
+        await install_browser()
+        return await start_browser(**kwargs)
+
+
+async def start_browser(**kwargs) -> Browser:
+    """
+    启动 playwright 浏览器实例。
+    Args:
+        **kwargs (Any): 传递给 `playwright.launch` 的关键字参数。
+
+    Returns:
+        Browser: 浏览器实例。
+    """
+    global _browser, _playwright
+    browser_type = (
+        plugin_config.htmlrender_browser.lower()
+        if plugin_config.htmlrender_browser
+        else "chromium"
+    )
+
+    if plugin_config.htmlrender_browser_channel:
+        kwargs["channel"] = plugin_config.htmlrender_browser_channel
+
+    if plugin_config.htmlrender_proxy_host:
+        kwargs["proxy"] = {
+            "server": plugin_config.htmlrender_proxy_host,
+        }
+
+    _playwright = await async_playwright().start()
+
+    match browser_type:
+        case "firefox":
+            logger.info("使用 Firefox 启动中...")
+            _browser = await _playwright.firefox.launch(**kwargs)
+            logger.debug(f"Firefox 路径: {_playwright.firefox.executable_path}")
+
+        case "webkit":
+            logger.info("使用 WebKit 启动中...")
+            _browser = await _playwright.webkit.launch(**kwargs)
+            logger.debug(f"WebKit 路径: {_playwright.webkit.executable_path}")
+
+        case "chromium":
+            if plugin_config.htmlrender_connect_over_cdp:
+                logger.info("使用 Chromium CDP 连接中...")
+                return await _playwright.chromium.connect_over_cdp(**kwargs)
+
+            logger.info("使用 Chromium 启动中...")
+            _browser = await _playwright.chromium.launch(**kwargs)
+            logger.debug(f"Chromium 路径: {_playwright.chromium.executable_path}")
+
+        case _:
+            raise ValueError(f"不支持的浏览器类型: {browser_type}")
+
+    return _browser
+
+
+def get_browser() -> Browser:
+    """"""
+    if not _browser:
+        raise RuntimeError("playwright is not initalized")
+    return _browser
+
+
+@asynccontextmanager
+async def get_new_page(device_scale_factor: float = 2, **kwargs) -> AsyncIterator[Page]:
+    """
+    获取一个新的页面的上下文管理器。
+    Args:
+        device_scale_factor (float): 设备缩放因子。
+        **kwargs (Any): 传递给 `browser.new_context` 的关键字参数。
+
+    Returns:
+        AsyncIterator[Page]: 页面对象。
+    """
+    ctx = get_browser()
+    page = await ctx.new_page(device_scale_factor=device_scale_factor, **kwargs)
+    try:
+        yield page
+    finally:
+        await page.close()
+
+
+async def shutdown_browser() -> None:
+    """关闭浏览器实例。"""
+    if _browser:
+        with _suppress_and_log():
+            await _browser.close()
+    if _playwright:
+        with _suppress_and_log():
+            await _playwright.stop()
+
+
+@dataclass
+class MirrorSource:
+    name: str
+    url: str
+    priority: int
+
+
+async def check_mirror_connectivity(
+    mirrors: list[MirrorSource], timeout: int = 5
+) -> Optional[MirrorSource]:
+    """
+    检查镜像源的可用性并返回最佳镜像源。
+    Args:
+        mirrors (list[MirrorSource]): 镜像源列表。
+        timeout (int): 连接超时时间。
+
+    Returns:
+        Optional[MirrorSource]: 最佳镜像源。
+    """
+
+    async def _check_single_mirror(mirror: MirrorSource) -> tuple[MirrorSource, float]:
+        """
+        检查单个镜像源的可用性。
+        Args:
+            mirror (MirrorSource): 镜像源。
+
+        Returns:
+            tuple[MirrorSource, float]: 镜像源和连接耗时。
+        """
+        try:
+            parsed_url = urlparse(mirror.url)
+            host = parsed_url.hostname
+            port = parsed_url.port or (443 if parsed_url.scheme == "https" else 80)
+
+            start_time = asyncio.get_event_loop().time()
+
+            _, _ = await asyncio.wait_for(
+                asyncio.open_connection(host, port), timeout=timeout
+            )
+
+            elapsed = asyncio.get_event_loop().time() - start_time
+            return mirror, elapsed
+
+        except (asyncio.TimeoutError, ConnectionRefusedError, socket.gaierror) as e:
+            logger.debug(f"镜像源 {mirror.name} 连接失败: {str(e)}")
+        return mirror, float("inf")
+
+    tasks = [_check_single_mirror(mirror) for mirror in mirrors]
+    results = await asyncio.gather(*tasks)
+
+    available_mirrors = [(m, t) for m, t in results if t != float("inf")]
+    if not available_mirrors:
+        return None
+
+    return min(available_mirrors, key=lambda x: (x[1], -x[0].priority))[0]
+
+
+@asynccontextmanager
+async def download_host_context(mirrors: list[MirrorSource]) -> AsyncIterator[None]:
+    """
+    为下载设置镜像源上下文管理器。
+    Args:
+        mirrors (list[MirrorSource]): 镜像源列表。
+
+    Returns:
+        AsyncIterator[None]: 上下文管理器。
+    """
+    had_original = "PLAYWRIGHT_DOWNLOAD_HOST" in os.environ
+    original_host = os.environ.get("PLAYWRIGHT_DOWNLOAD_HOST")
+
+    try:
+        best_mirror = await check_mirror_connectivity(mirrors)
+
+        if plugin_config.htmlrender_download_host:
+            logger.info(f"使用配置的下载源: {plugin_config.htmlrender_download_host}")
+            os.environ["PLAYWRIGHT_DOWNLOAD_HOST"] = (
+                plugin_config.htmlrender_download_host
+            )
+        elif best_mirror:
+            logger.info(f"使用镜像源: {best_mirror.name} ({best_mirror.url})")
+            os.environ["PLAYWRIGHT_DOWNLOAD_HOST"] = best_mirror.url
+        else:
+            logger.info("无可用镜像源，使用默认源")
+
+        yield
+    finally:
+        if had_original and original_host is not None:
+            os.environ["PLAYWRIGHT_DOWNLOAD_HOST"] = original_host
+        elif "PLAYWRIGHT_DOWNLOAD_HOST" in os.environ:
+            del os.environ["PLAYWRIGHT_DOWNLOAD_HOST"]
+
+
+class MessageType(Enum):
+    PROGRESS = "progress"
+    DOWNLOAD = "download"
+    INFO = "info"
+
+
+async def read_stream(stream, out: TextIO = sys.stdout) -> None:
+    """实时读取并输出流内容
+
+    Args:
+        stream: 要读取的流
+        out: 输出流，默认为 sys.stdout
+    """
+    last_progress = ""
+
+    def write_progress(_: str) -> None:
+        """写入进度条"""
+        out.write(f"\r{_}")
+        out.flush()
+
+    def ensure_newline() -> None:
+        """确保进度条后换行"""
+        nonlocal last_progress
+        if last_progress:
+            out.write("\n")
+            out.flush()
+            last_progress = ""
+
+    def get_message_type(_: str) -> MessageType:
+        """获取消息类型"""
+        if "|" in _ and "%" in _ and ("MiB" in _ or "KiB" in _):
+            return MessageType.PROGRESS
+        elif _.startswith("Downloading "):
+            return MessageType.DOWNLOAD
+        return MessageType.INFO
+
+    while True:
+        line = await stream.readline()
+        if not line:
+            break
+
+        text = line.decode().strip()
+        if not text:
+            continue
+
+        msg_type = get_message_type(text)
+
+        match msg_type:
+            case MessageType.PROGRESS:
+                progress_text = text.split("|", 1)[1].strip()
+                write_progress(progress_text)
+                last_progress = text
+            case _:
+                ensure_newline()
+                logger.info(text)
+
+
+async def execute_install_command(
+    install_cmd: list[str], timeout: int
+) -> tuple[bool, str]:
+    """
+    执行安装命令并返回结果。
+    Args:
+        install_cmd (list[str]): 安装命令。
+        timeout (int): 等待超时时间。
+
+    Returns:
+        tuple[bool, str]: 是否安装成功和消息。
+    """
+    process = await asyncio.create_subprocess_exec(
+        *install_cmd,
+        stdout=asyncio.subprocess.PIPE,
+        stderr=asyncio.subprocess.PIPE,
+    )
+
+    try:
+        await asyncio.wait_for(
+            asyncio.gather(
+                read_stream(process.stdout),
+                read_stream(process.stderr),
+            ),
+            timeout=timeout,
+        )
+
+        await process.wait()
+        return (
+            process.returncode == 0,
+            "安装完成" if process.returncode == 0 else "安装失败",
+        )
+    except asyncio.TimeoutError:
+        return False, "安装超时"
+    finally:
+        if process and process.returncode is None:
+            process.kill()
+            await process.wait()
+
+
+async def install_browser(
+    browser_type: str = "chromium", timeout: int = 300, retries: int = 2
+) -> bool:
+    """
+    安装用于 HTML 渲染的浏览器。
+
+    Args:
+        browser_type (str): 要安装的浏览器类型 ("chromium"、"firefox" 或 "webkit")。
+        timeout (int): 安装过程中等待的最大秒数。
+        retries (int): 安装失败时的最大重试次数。
+
+    Returns:
+        bool: 如果安装成功返回 True，否则返回 False
+    """
+    mirrors = [
+        MirrorSource("淘宝", "https://npmmirror.com/mirrors/playwright/", 1),
+    ]
+    async with download_host_context(mirrors):
+        install_cmd = ["playwright", "install", "--with-deps", browser_type]
+
+        for attempt in range(1, retries + 1):
+            try:
+                logger.info(
+                    f"正在安装 {browser_type}（第 {attempt}/{retries} 次尝试）..."
+                )
+                success, message = await execute_install_command(install_cmd, timeout)
+
+                if success:
+                    logger.success(f"{browser_type} 安装成功")
+                    return True
+                else:
+                    logger.error(f"安装失败：{message}")
+
+            except Exception as e:
+                logger.error(f"安装过程中出现异常：{str(e)}")
+
+            if attempt < retries:
+                await asyncio.sleep(2)
+
+        logger.error(f"在 {retries} 次尝试后安装失败")
+        return False

{nonebot_plugin_htmlrender-0.4.0 → nonebot_plugin_htmlrender-0.5.0}/nonebot_plugin_htmlrender/config.py

@@ -1,6 +1,7 @@
 from typing import Optional
 
 from pydantic import BaseModel, Field
+from nonebot import get_driver, get_plugin_config
 
 
 class Config(BaseModel):
@@ -9,4 +10,8 @@ class Config(BaseModel):
     htmlrender_proxy_host: Optional[str] = Field(default=None)
     htmlrender_browser_channel: Optional[str] = Field(default=None)
     htmlrender_browser_executable_path: Optional[str] = Field(default=None)
-    htmlrender_connect_over_cdp: Optional[str] = Field(default=False)
+    htmlrender_connect_over_cdp: Optional[str] = Field(default=None)
+
+
+global_config = get_driver().config
+plugin_config = get_plugin_config(Config)

{nonebot_plugin_htmlrender-0.4.0 → nonebot_plugin_htmlrender-0.5.0}/pyproject.toml

@@ -1,6 +1,6 @@
 [project]
 name = "nonebot-plugin-htmlrender"
-version = "0.4.0"
+version = "0.5.0"
 description = "通过浏览器渲染图片"
 readme = "README.md"
 authors = [{ name = "kexue", email = "x@kexue-cloud.cn" }]
@@ -30,7 +30,7 @@ asyncio_mode = "auto"
 dev = [
     "nonebot-adapter-onebot>=2.4.6",
     "nonebot2[fastapi]>=2.3.3",
-    "nonebug>=0.4.1",
+    "nonebug>=0.4.2",
     "pillow>=11.0.0",
     "pytest-asyncio>=0.24.0",
 ]

nonebot_plugin_htmlrender-0.5.0/tests/conftest.py

@@ -0,0 +1,15 @@
+import nonebot
+import pytest
+from pytest_asyncio import is_async_test
+
+
+def pytest_collection_modifyitems(items: list[pytest.Item]):
+    pytest_asyncio_tests = (item for item in items if is_async_test(item))
+    session_scope_marker = pytest.mark.asyncio(loop_scope="session")
+    for async_test in pytest_asyncio_tests:
+        async_test.add_marker(session_scope_marker, append=False)
+
+
+@pytest.fixture(scope="session", autouse=True)
+async def after_nonebot_init(after_nonebot_init: None):
+    nonebot.require("nonebot_plugin_htmlrender")

{nonebot_plugin_htmlrender-0.4.0 → nonebot_plugin_htmlrender-0.5.0}/uv.lock

@@ -612,7 +612,7 @@ wheels = [
 
 [[package]]
 name = "nonebot-plugin-htmlrender"
-version = "0.4.0"
+version = "0.5.0"
 source = { editable = "." }
 dependencies = [
     { name = "aiofiles" },
@@ -650,7 +650,7 @@ requires-dist = [
 dev = [
     { name = "nonebot-adapter-onebot", specifier = ">=2.4.6" },
     { name = "nonebot2", extras = ["fastapi"], specifier = ">=2.3.3" },
-    { name = "nonebug", specifier = ">=0.4.1" },
+    { name = "nonebug", specifier = ">=0.4.2" },
     { name = "pillow", specifier = ">=11.0.0" },
     { name = "pytest-asyncio", specifier = ">=0.24.0" },
 ]
@@ -681,7 +681,7 @@ fastapi = [
 
 [[package]]
 name = "nonebug"
-version = "0.4.1"
+version = "0.4.2"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
     { name = "asgiref" },
@@ -690,9 +690,9 @@ dependencies = [
     { name = "pytest" },
     { name = "typing-extensions" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/d1/57/4300262bbf6f8b7b69a75e8d827a32b1503fe9322210440ad785fd6bf3d0/nonebug-0.4.1.tar.gz", hash = "sha256:1bb96d715ec415351f2c7b871a86c64fdba533fb1a105e9a01cc1e250d37d52b", size = 11789 }
+sdist = { url = "https://files.pythonhosted.org/packages/29/56/01eb465dc135787e463423bf4762b883159a639d02c3a577ab4efc8f9d56/nonebug-0.4.2.tar.gz", hash = "sha256:002dac00fed1fbff97574f3dffab2237fc41473b3bf3e73d2730ae34895bc137", size = 12118 }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/ce/e0/e8d6e0db6cbff3edbb034284bd9451e9e380dde6c580fd22f1d6c2a91687/nonebug-0.4.1-py3-none-any.whl", hash = "sha256:553bfc408337fb33a19fcc936e24f4cc43200e4550a828f21e7f4406e93b62db", size = 14659 },
+    { url = "https://files.pythonhosted.org/packages/a0/4a/7592e899cd5074a9b70df007afe9b6f8bbaa02b8e9c1ce7f5c5498634cb7/nonebug-0.4.2-py3-none-any.whl", hash = "sha256:83d9229ec899819e9f5786eabc0c27b9dd0cd5d6e2e7438f59923ab822d5c555", size = 15030 },
 ]
 
 [[package]]

nonebot_plugin_htmlrender-0.4.0/nonebot_plugin_htmlrender/browser.py

@@ -1,131 +0,0 @@
-#!/usr/bin/env python3
-# -*- coding: utf-8 -*-
-"""
-@Author         : yanyongyu
-@Date           : 2021-03-12 13:42:43
-@LastEditors    : yanyongyu
-@LastEditTime   : 2021-11-01 14:05:41
-@Description    : None
-@GitHub         : https://github.com/yanyongyu
-"""
-
-__author__ = "yanyongyu"
-
-from contextlib import asynccontextmanager
-from typing import AsyncIterator, Optional
-
-from nonebot import get_plugin_config
-from nonebot.log import logger
-from playwright.async_api import Browser, Error, Page, Playwright, async_playwright
-
-from .config import Config
-
-config = get_plugin_config(Config)
-
-_browser: Optional[Browser] = None
-_playwright: Optional[Playwright] = None
-
-
-async def init(**kwargs) -> Browser:
-    global _browser
-    global _playwright
-    _playwright = await async_playwright().start()
-    try:
-        _browser = await launch_browser(**kwargs)
-    except Error:
-        await install_browser()
-        _browser = await launch_browser(**kwargs)
-    return _browser
-
-
-async def launch_browser(**kwargs) -> Browser:
-    assert _playwright is not None, "Playwright 未安装"
-
-    if config.htmlrender_browser_channel:
-        kwargs["channel"] = config.htmlrender_browser_channel
-
-    if config.htmlrender_proxy_host:
-        kwargs["proxy"] = {
-            "server": config.htmlrender_proxy_host,
-        }
-
-    if config.htmlrender_browser_executable_path:
-        kwargs["executable_path"] = config.htmlrender_browser_executable_path
-    
-    if config.htmlrender_browser:
-        if config.htmlrender_browser.lower() == "firefox":
-            logger.info("使用 Firefox 启动")
-            return await _playwright.firefox.launch(**kwargs)
-        elif config.htmlrender_browser.lower() == "webkit":
-            logger.info("使用 Webkit 启动")
-            return await _playwright.webkit.launch(**kwargs)
-        else:
-            logger.warning("浏览器参数错误，无 {} 浏览器核心可用，使用 Chromium 启动")
-
-    if config.htmlrender_connect_over_cdp:
-        logger.info("使用 chromium connect_over_cdp 启动")
-        return await _playwright.chromium.connect_over_cdp(**kwargs)
-
-    # 默认使用 chromium
-    logger.info("使用 Chromium 启动")
-    return await _playwright.chromium.launch(**kwargs)
-
-
-async def get_browser(**kwargs) -> Browser:
-    return _browser if _browser and _browser.is_connected() else await init(**kwargs)
-
-
-@asynccontextmanager
-async def get_new_page(device_scale_factor: float = 2, **kwargs) -> AsyncIterator[Page]:
-    browser = await get_browser()
-    page = await browser.new_page(device_scale_factor=device_scale_factor, **kwargs)
-    try:
-        yield page
-    finally:
-        await page.close()
-
-
-async def shutdown_browser():
-    global _browser
-    global _playwright
-    if _browser:
-        if _browser.is_connected():
-            await _browser.close()
-        _browser = None
-    if _playwright:
-        # await _playwright.stop()
-        _playwright = None
-
-
-async def install_browser():
-    import os
-    import sys
-
-    from playwright.__main__ import main
-
-    if host := config.htmlrender_download_host:
-        logger.info("使用配置源进行下载")
-        os.environ["PLAYWRIGHT_DOWNLOAD_HOST"] = host
-    else:
-        logger.info("使用镜像源进行下载")
-        os.environ["PLAYWRIGHT_DOWNLOAD_HOST"] = (
-            "https://npmmirror.com/mirrors/playwright/"
-        )
-    success = False
-
-    if config.htmlrender_browser == "firefox":
-        logger.info("正在安装 firefox")
-        sys.argv = ["", "install", "firefox"]
-    else:
-        # 默认使用 chromium
-        logger.info("正在安装 chromium")
-        sys.argv = ["", "install", "chromium"]
-    try:
-        logger.info("正在安装依赖")
-        os.system("playwright install-deps")  # noqa: ASYNC102, S605, S607
-        main()
-    except SystemExit as e:
-        if e.code == 0:
-            success = True
-    if not success:
-        logger.error("浏览器更新失败, 请检查网络连通性")

nonebot_plugin_htmlrender-0.4.0/tests/conftest.py

@@ -1,10 +0,0 @@
-import nonebot
-import pytest
-from nonebug import App
-
-
-@pytest.fixture
-def app():
-    nonebot.require("nonebot_plugin_htmlrender")
-
-    yield App()