411 lines
14 KiB
Python
411 lines
14 KiB
Python
import subprocess
|
|
import time
|
|
import markdown
|
|
from pathlib import Path
|
|
|
|
import psutil
|
|
from playwright.async_api import async_playwright
|
|
import os
|
|
import asyncio
|
|
import re
|
|
from loguru import logger
|
|
|
|
|
|
async def safe_close_browser(browser, timeout: float = 4.0) -> None:
|
|
if not browser:
|
|
return
|
|
|
|
for context in browser.contexts[:]:
|
|
for page in context.pages[:]:
|
|
try:
|
|
await asyncio.wait_for(page.close(), timeout=1.5)
|
|
except Exception:
|
|
pass
|
|
try:
|
|
await asyncio.wait_for(context.close(), timeout=timeout)
|
|
except Exception:
|
|
pass
|
|
|
|
try:
|
|
await asyncio.wait_for(browser.close(), timeout=timeout)
|
|
logger.debug("browser closed gracefully")
|
|
return
|
|
except (asyncio.TimeoutError, Exception) as e:
|
|
logger.warning(f"browser.close failed: {e}")
|
|
|
|
if browser.process and browser.process.pid:
|
|
try:
|
|
parent = psutil.Process(browser.process.pid)
|
|
children = parent.children(recursive=True)
|
|
for proc in children:
|
|
try:
|
|
proc.terminate()
|
|
except (psutil.NoSuchProcess, psutil.AccessDenied, psutil.ZombieProcess):
|
|
pass
|
|
try:
|
|
parent.terminate()
|
|
except (psutil.NoSuchProcess, psutil.AccessDenied, psutil.ZombieProcess):
|
|
pass
|
|
try:
|
|
gone, alive = psutil.wait_procs([parent] + children, timeout=2)
|
|
except Exception:
|
|
gone, alive = [], [parent] + children
|
|
for proc in alive:
|
|
try:
|
|
proc.kill()
|
|
except (psutil.NoSuchProcess, psutil.AccessDenied, psutil.ZombieProcess):
|
|
pass
|
|
try:
|
|
gone, alive = psutil.wait_procs([parent] + children, timeout=3)
|
|
except Exception:
|
|
alive = []
|
|
if alive:
|
|
logger.warning(f"process still alive after kill: {[p.pid for p in alive]}")
|
|
else:
|
|
logger.debug("process tree terminated")
|
|
except (psutil.NoSuchProcess, Exception) as e:
|
|
logger.warning(f"force kill failed: {e}")
|
|
|
|
|
|
def _split_hero(html_body: str):
|
|
title_match = re.search(r'<h1>(.*?)</h1>', html_body, re.S | re.I)
|
|
hero_title = title_match.group(1).strip() if title_match else "聊天总结"
|
|
remain = re.sub(r'<h1>.*?</h1>', '', html_body, count=1, flags=re.S | re.I).strip()
|
|
|
|
meta_match = re.match(r'^<p>(.*?)</p>', remain, re.S | re.I)
|
|
hero_meta = meta_match.group(1).strip() if meta_match else "群聊总结 / 自动生成"
|
|
if meta_match:
|
|
remain = re.sub(r'^<p>.*?</p>', '', remain, count=1, flags=re.S | re.I).strip()
|
|
|
|
return hero_title, hero_meta, remain
|
|
|
|
|
|
async def md_str_to_html_content(md_content):
|
|
"""将 Markdown 字符串转换为更美观的 HTML 内容。"""
|
|
html_body = markdown.markdown(md_content, extensions=['extra', 'codehilite'])
|
|
hero_title, hero_meta, remain_html = _split_hero(html_body)
|
|
|
|
css = """
|
|
<style>
|
|
:root {
|
|
--bg1: #f7f4ff;
|
|
--bg2: #eef8ff;
|
|
--paper: #ffffff;
|
|
--text: #233044;
|
|
--muted: #718198;
|
|
--primary: #6d5efc;
|
|
--primary-soft: rgba(109,94,252,0.10);
|
|
--accent: #14b8a6;
|
|
--line: rgba(148,163,184,0.18);
|
|
--code-bg: #0f172a;
|
|
--code-fg: #e2e8f0;
|
|
--quote-bg: rgba(20,184,166,0.08);
|
|
--shadow: 0 20px 45px rgba(80, 84, 125, 0.10);
|
|
}
|
|
* { box-sizing: border-box; }
|
|
html, body { margin: 0; padding: 0; }
|
|
body {
|
|
font-family: -apple-system, BlinkMacSystemFont, 'Segoe UI', Roboto, Oxygen, Ubuntu, Cantarell, 'Open Sans', 'Helvetica Neue', 'Noto Sans CJK SC', 'Microsoft YaHei', sans-serif;
|
|
color: var(--text);
|
|
font-size: 16px;
|
|
line-height: 1.78;
|
|
background:
|
|
radial-gradient(circle at top left, #ffffff 0%, var(--bg1) 38%, transparent 70%),
|
|
radial-gradient(circle at top right, #ffffff 0%, var(--bg2) 34%, transparent 70%),
|
|
linear-gradient(180deg, #f8f9fd 0%, #eef4fb 100%);
|
|
padding: 26px;
|
|
}
|
|
.wrap {
|
|
max-width: 820px;
|
|
margin: 0 auto;
|
|
background: rgba(255,255,255,0.96);
|
|
border: 1px solid rgba(255,255,255,0.7);
|
|
border-radius: 28px;
|
|
box-shadow: var(--shadow);
|
|
overflow: hidden;
|
|
}
|
|
.hero {
|
|
padding: 30px 34px 22px;
|
|
background:
|
|
radial-gradient(circle at 85% 20%, rgba(255,255,255,0.42), transparent 24%),
|
|
linear-gradient(135deg, rgba(109,94,252,0.18), rgba(20,184,166,0.10));
|
|
border-bottom: 1px solid var(--line);
|
|
}
|
|
.hero-badge {
|
|
display: inline-block;
|
|
padding: 6px 12px;
|
|
border-radius: 999px;
|
|
font-size: 12px;
|
|
color: #5747ef;
|
|
background: rgba(255,255,255,0.70);
|
|
border: 1px solid rgba(109,94,252,0.12);
|
|
margin-bottom: 14px;
|
|
letter-spacing: .04em;
|
|
}
|
|
.hero-title {
|
|
margin: 0;
|
|
font-size: 2.05em;
|
|
line-height: 1.28;
|
|
font-weight: 800;
|
|
color: #1f2557;
|
|
text-align: center;
|
|
letter-spacing: -0.02em;
|
|
}
|
|
.hero-meta {
|
|
margin: 10px auto 0;
|
|
max-width: 620px;
|
|
text-align: center;
|
|
color: var(--muted);
|
|
font-size: 0.92em;
|
|
line-height: 1.7;
|
|
}
|
|
.content {
|
|
padding: 24px 34px 34px;
|
|
}
|
|
h1, h2, h3, h4, h5, h6 {
|
|
color: var(--text);
|
|
margin-top: 24px;
|
|
margin-bottom: 14px;
|
|
font-weight: 700;
|
|
line-height: 1.35;
|
|
letter-spacing: -0.01em;
|
|
}
|
|
h1 { display: none; }
|
|
h2 {
|
|
font-size: 1.42em;
|
|
margin-top: 30px;
|
|
padding: 10px 14px;
|
|
background: linear-gradient(90deg, var(--primary-soft), rgba(255,255,255,0));
|
|
border-left: 4px solid var(--primary);
|
|
border-radius: 12px;
|
|
}
|
|
h3 {
|
|
font-size: 1.15em;
|
|
margin-top: 24px;
|
|
color: #30435f;
|
|
padding-left: 12px;
|
|
border-left: 3px solid rgba(20,184,166,0.55);
|
|
}
|
|
p {
|
|
margin: 14px 0;
|
|
color: #334155;
|
|
line-height: 1.88;
|
|
}
|
|
ul, ol {
|
|
padding-left: 26px;
|
|
margin: 14px 0 18px;
|
|
}
|
|
li {
|
|
margin: 8px 0;
|
|
color: #334155;
|
|
}
|
|
li::marker { color: var(--primary); }
|
|
strong {
|
|
color: #1e293b;
|
|
font-weight: 700;
|
|
}
|
|
em { color: #5b6b84; }
|
|
code {
|
|
background: rgba(109,94,252,0.08);
|
|
color: #5b3df5;
|
|
padding: 2px 8px;
|
|
border-radius: 8px;
|
|
font-size: 0.92em;
|
|
border: 1px solid rgba(109,94,252,0.10);
|
|
}
|
|
pre {
|
|
background: var(--code-bg);
|
|
color: var(--code-fg);
|
|
padding: 16px 18px;
|
|
border-radius: 16px;
|
|
overflow-x: auto;
|
|
border: 1px solid rgba(255,255,255,0.06);
|
|
box-shadow: inset 0 1px 0 rgba(255,255,255,0.03);
|
|
}
|
|
pre code {
|
|
background: transparent;
|
|
color: inherit;
|
|
border: none;
|
|
padding: 0;
|
|
}
|
|
table {
|
|
border-collapse: separate;
|
|
border-spacing: 0;
|
|
width: 100%;
|
|
margin: 20px 0;
|
|
background: rgba(255,255,255,0.96);
|
|
border: 1px solid rgba(148,163,184,0.16);
|
|
border-radius: 16px;
|
|
overflow: hidden;
|
|
box-shadow: 0 8px 24px rgba(15,23,42,0.05);
|
|
}
|
|
th, td {
|
|
padding: 12px 14px;
|
|
text-align: left;
|
|
border-bottom: 1px solid rgba(148,163,184,0.12);
|
|
}
|
|
tr:last-child td { border-bottom: none; }
|
|
th {
|
|
background: linear-gradient(180deg, rgba(109,94,252,0.10), rgba(109,94,252,0.04));
|
|
color: #334155;
|
|
font-weight: 700;
|
|
}
|
|
blockquote {
|
|
margin: 18px 0;
|
|
padding: 14px 18px;
|
|
background: var(--quote-bg);
|
|
border: 1px solid rgba(20,184,166,0.16);
|
|
border-left: 5px solid var(--accent);
|
|
border-radius: 14px;
|
|
color: #355468;
|
|
}
|
|
hr {
|
|
border: none;
|
|
height: 1px;
|
|
background: linear-gradient(90deg, transparent, rgba(148,163,184,0.35), transparent);
|
|
margin: 26px 0;
|
|
}
|
|
a {
|
|
color: var(--primary);
|
|
text-decoration: none;
|
|
border-bottom: 1px dashed rgba(109,94,252,0.35);
|
|
}
|
|
.signature {
|
|
margin-top: 34px;
|
|
text-align: right;
|
|
color: var(--muted);
|
|
font-size: 0.95em;
|
|
font-style: italic;
|
|
}
|
|
</style>
|
|
"""
|
|
|
|
full_html = f'''<html>
|
|
<head>
|
|
<meta charset="UTF-8">
|
|
<meta name="viewport" content="width=device-width, initial-scale=1.0">
|
|
{css}
|
|
</head>
|
|
<body>
|
|
<div class="wrap">
|
|
<div class="hero">
|
|
<div class="hero-badge">AI 群聊总结</div>
|
|
<h1 class="hero-title">{hero_title}</h1>
|
|
<div class="hero-meta">{hero_meta}</div>
|
|
</div>
|
|
<div class="content">
|
|
{remain_html}
|
|
</div>
|
|
</div>
|
|
</body>
|
|
</html>'''
|
|
return full_html
|
|
|
|
|
|
def check_chromium_installed(path):
|
|
return os.path.isfile(path) and os.access(path, os.X_OK)
|
|
|
|
|
|
async def html_to_image(html_content, output_image):
|
|
async with async_playwright() as p:
|
|
browser_path = None
|
|
if os.name == 'nt':
|
|
possible_chrome_paths = [
|
|
r"C:\Users\Liu_WIN10\AppData\Local\Google\Chrome\Application\chrome.exe",
|
|
r"C:\Users\Liu-OPEN\AppData\Local\Google\Chrome\Application\chrome.exe",
|
|
r"C:\Program Files\Google\Chrome\Application\chrome.exe",
|
|
r"C:\Program Files (x86)\Google\Chrome\Application\chrome.exe"
|
|
]
|
|
for path in possible_chrome_paths:
|
|
if check_chromium_installed(path):
|
|
browser_path = path
|
|
break
|
|
else:
|
|
import glob
|
|
user_home = os.path.expanduser("~")
|
|
glob_pattern = os.path.join(user_home, ".cache", "ms-playwright", "chromium-*", "chrome-linux", "chrome")
|
|
chrome_paths = glob.glob(glob_pattern)
|
|
for path in sorted(chrome_paths, reverse=True):
|
|
if check_chromium_installed(path):
|
|
browser_path = path
|
|
break
|
|
|
|
launch_args = [
|
|
"--no-sandbox",
|
|
"--disable-setuid-sandbox",
|
|
"--disable-dev-shm-usage",
|
|
"--disable-gpu"
|
|
]
|
|
|
|
if browser_path:
|
|
logger.debug(f"Launch chromium with system chrome: {browser_path}")
|
|
browser = await p.chromium.launch(executable_path=browser_path, args=launch_args)
|
|
else:
|
|
logger.debug("Launch chromium with bundled browser")
|
|
browser = await p.chromium.launch(args=launch_args)
|
|
|
|
try:
|
|
context = await browser.new_context(
|
|
viewport={"width": 780, "height": 960},
|
|
device_scale_factor=1.2
|
|
)
|
|
page = await context.new_page()
|
|
|
|
logger.debug("Set page content")
|
|
await page.set_content(html_content, wait_until='domcontentloaded', timeout=15000)
|
|
|
|
logger.debug("Wait for fonts ready")
|
|
await page.evaluate("document.fonts.ready")
|
|
await asyncio.sleep(0.2)
|
|
|
|
logger.debug(f"Take screenshot: output={output_image}")
|
|
await page.screenshot(
|
|
path=output_image,
|
|
full_page=True,
|
|
timeout=15000,
|
|
animations="disabled"
|
|
)
|
|
if not os.path.exists(output_image):
|
|
raise RuntimeError(f"截图失败，输出文件不存在: {output_image}")
|
|
|
|
finally:
|
|
logger.debug("Closing browser")
|
|
await safe_close_browser(browser)
|
|
|
|
|
|
async def convert_md_str_to_image(md_content: str, output_image: str, max_retries: int = 2) -> str:
|
|
if not md_content:
|
|
raise ValueError("Markdown content cannot be empty")
|
|
|
|
project_root = os.getcwd()
|
|
temp_dir = Path(project_root) / "temp" / "md2image"
|
|
temp_dir.mkdir(parents=True, exist_ok=True)
|
|
output_image_path = temp_dir / output_image
|
|
|
|
last_error = None
|
|
|
|
for attempt in range(max_retries):
|
|
try:
|
|
logger.debug(f"尝试第 {attempt + 1}/{max_retries} 次生成图片")
|
|
|
|
if output_image_path.exists():
|
|
os.remove(str(output_image_path))
|
|
|
|
full_html = await md_str_to_html_content(md_content)
|
|
await html_to_image(full_html, str(output_image_path))
|
|
|
|
image_size = os.path.getsize(str(output_image_path))
|
|
if image_size < 1024:
|
|
raise RuntimeError(f"图片生成异常，大小仅为: {image_size} bytes")
|
|
|
|
logger.info(f"图片成功生成：{output_image_path}")
|
|
return str(output_image_path.resolve())
|
|
|
|
except Exception as e:
|
|
last_error = e
|
|
logger.warning(f"第 {attempt + 1} 次尝试失败: {e}")
|
|
if attempt < max_retries - 1:
|
|
await asyncio.sleep(1.5)
|
|
|
|
raise RuntimeError(f"图片生成失败，已重试 {max_retries} 次。最后错误: {last_error}")
|