点击展开示例：博客格式化提示词
帮我格式化Markdown文本
1. 英文与中文之间加空格
2. 标点符号前的空格
3. 专有名词修正
4. 大体保持原有格式
5. URL、代码块、配置文件保持原状
6. 连续英文词汇保持紧凑（如SYSTEMD-JOURNAL）
7. 路径和变量名保持紧凑（如/opt/monitor）<details>
<summary>点击展开提示</summary>
示例：这里可以写 **Markdown 内容**
</details><div style="padding: 15px; background-color: #e7f3f0; color: #1e4620; border-left: 5px solid #4caf50; border-radius: 4px; margin-bottom: 10px;">
✅ <b>成功：</b> 这是一个绿色的成功提示框。
</div>最近遇到个很迷惑的问题，同一个静态前端网页在其他手机、电脑显示正常，但在我这台 Win10 机器上，彩虹流动的 CSS 动画完全不动，颜色渲染也有点怪，换了 3 个浏览器，结果都一样。
在 AI 的帮助下，浏览器下按F12，并打开控制台，执行matchMedia('(prefers-reduced-motion: reduce)').matches，结果返回true。这意味着系统开启了“减少动态效果”，浏览器检测到后，会主动把动画禁用。
设置 → 轻松使用 → 显示 → 打开在 Windows 中显示动画
控制面板 → 轻松使用 → 轻松使用中心 → 使计算机更易于查看 → 取消勾选关闭所有不必要的动画
添加 CNAME，将 子域名 解析到 www.shopify.com，关闭小橙云，完成优选域名的添加。
其他 Cloudflare 优选域名请自行搜索，如 www.visa.com。
打开一个项目，进入 设置 - 域和路由 - 添加 - 路由，填上形如 子域名.example.com/*
从测速网站上可见，已经成功加速了 Workers 项目的访问速度。
大概是几十天前，我在查询 whois 时发现， SixHz.com 的前任主人已经不再续费了，但域名还没开放购买。于是把相关信息发给 AI，问了下大致的释放时间。从那之后，就会偶尔搜一下，看看能不能买到。
这个世界上，有些事情可遇不可求，而这一次我很幸运，今天终于蹲到了。
域名是在 Cloudflare 购买的，价格是 10.46 美元 / 年。
顺手查了下历史，这个域名之前被用作 Go 组件托管地址 sixhz.com/szamlazzhu，似乎和匈牙利的发票业务有关，还挺有意思的。
想到一个问题：如果有一天我不再更新博客，甚至不再使用原本的服务器，这些博客内容会怎样？经过实测，如果网站不能访问，搜索引擎会在短时间内检测到并“下架”。而我想让它们继续被看见，让它们独立于服务器依赖而存在。
于是就有了这个小项目：一个用 Python + wget 做的极简博客镜像工具，可以把你的 Typecho（或其他兼容的博客）抓成一套完整的、可直接部署的静态网站。
典型用法：把 Typecho 网站静态页面部分爬虫，其他网站自行测试。
./site/，可以直接推送到 Cloudflare Pages / GitHub Pages。img src的图片类资源也会保存并重写为本地相对链接，确保离线可用。--spider 探活；镜像在 site_tmp/ 完成后替换 site/，探活或下载失败都会保留旧的 site/ 不被清空。换句话说：你的博客在任何地方都能继续「活着」。
uv 管理项目。tools/mingw64/bin/。uv run python mirror.pyuv run python mirror.py --no-cleanuv run python mirror.py --spideruv run python mirror.py --url https://example.com/ --output-dir my_site"""
The script prefers the bundled wget on Windows (tools/mingw64/bin/wget.exe),
and uses the system wget on Linux/macOS, falling back to the bundled version
when system wget is unavailable. Output defaults to ./site and can be cleaned
before each run.
"""
from __future__ import annotations
import argparse
import hashlib
import os
import platform
import re
import shutil
import subprocess
import sys
from pathlib import Path
from typing import Iterable, List
from urllib.error import URLError
from urllib.parse import urlsplit
from urllib.request import urlopen
DEFAULT_URL = "https://blog.example.com/"
DEFAULT_OUTPUT_DIR = "site"
REJECT_REGEX = r"/(admin|login|register|action|feed)/"
def project_root() -> Path:
"""Return the directory containing this script."""
return Path(__file__).resolve().parent
def resolve_output_dir(root: Path, output_dir: str) -> Path:
"""Resolve and validate the output directory inside the project root."""
target = (root / output_dir).resolve()
try:
target.relative_to(root)
except ValueError as exc:
raise ValueError("Output directory must stay under the project root") from exc
return target
def temp_output_dir(output_dir: Path) -> Path:
"""Derive a temporary output directory alongside the final output."""
return output_dir.with_name(f"{output_dir.name}_tmp")
def bundled_wget_path(root: Path) -> Path:
"""Path to the repository-bundled wget executable."""
return root / "tools" / "mingw64" / "bin" / "wget.exe"
def find_wget(root: Path) -> Path:
"""Pick the appropriate wget executable."""
system_name = platform.system().lower()
bundled = bundled_wget_path(root)
system_wget = shutil.which("wget")
if system_name == "windows":
if bundled.exists():
return bundled
if system_wget:
return Path(system_wget)
raise FileNotFoundError(
"wget not found. Expected bundled wget at tools/mingw64/bin/wget.exe "
"or a system wget in PATH."
)
if system_wget:
return Path(system_wget)
if bundled.exists():
return bundled
raise FileNotFoundError(
"wget not found. Install wget or place it at tools/mingw64/bin/wget.exe."
)
def cleanup_directory(path: Path) -> None:
"""Remove a directory tree if it exists."""
if path.exists():
shutil.rmtree(path)
def prepare_temp_directory(temp_dir: Path, seed_from: Path | None) -> None:
"""Create a fresh temporary directory, optionally seeded from an existing tree."""
cleanup_directory(temp_dir)
temp_dir.parent.mkdir(parents=True, exist_ok=True)
if seed_from and seed_from.exists():
shutil.copytree(seed_from, temp_dir)
else:
temp_dir.mkdir(parents=True, exist_ok=True)
def replace_directory(src: Path, dst: Path) -> None:
"""Replace dst with src, keeping the previous dst until replacement succeeds."""
backup = dst.with_name(f"{dst.name}_backup")
if backup.exists():
shutil.rmtree(backup)
dst.parent.mkdir(parents=True, exist_ok=True)
dst_existed = dst.exists()
if dst_existed:
dst.rename(backup)
try:
src.rename(dst)
except OSError:
if dst_existed and backup.exists():
backup.rename(dst)
raise
if backup.exists():
shutil.rmtree(backup)
def build_wget_command(
wget_path: Path, output_dir: Path, url: str, spider: bool
) -> List[str]:
"""Construct the wget command for the mirror job."""
command: List[str] = [
str(wget_path),
"--mirror",
"--convert-links",
"--adjust-extension",
"--page-requisites",
"--no-parent",
"--restrict-file-names=windows",
f"--reject-regex={REJECT_REGEX}",
"-P",
str(output_dir),
"-nH",
]
if spider:
command.append("--spider")
command.append(url)
return command
def stream_process_output(command: Iterable[str]) -> int:
"""Run a process and stream stdout/stderr to the console."""
with subprocess.Popen(
list(command),
stdout=subprocess.PIPE,
stderr=subprocess.STDOUT,
text=True,
encoding="utf-8",
errors="replace",
) as proc:
if proc.stdout:
for line in proc.stdout:
print(line, end="")
return_code = proc.wait()
return return_code
def rewrite_links_to_local(output_dir: Path, base_url: str) -> None:
"""Post-process downloaded files to point base-domain assets to local copies."""
parsed = urlsplit(base_url)
host = parsed.netloc
if not host:
return
prefixes = {f"{scheme}://{host}" for scheme in ("http", "https")}
prefixes.add(f"//{host}")
pattern = re.compile(
r"(?P<prefix>" + "|".join(re.escape(p) for p in prefixes) + r")(?P<path>/[^\s\"'>)]+)"
)
for file_path in output_dir.rglob("*"):
if file_path.suffix.lower() not in {".html", ".htm", ".css", ".js"}:
continue
try:
original = file_path.read_text(encoding="utf-8", errors="ignore")
except OSError:
continue
changed = False
def _replace(match: re.Match[str]) -> str:
nonlocal changed
url_path = match.group("path")
local_target = (output_dir / url_path.lstrip("/")).resolve()
if local_target.exists():
relative = Path(
os.path.relpath(local_target, start=file_path.parent.resolve())
)
changed = True
return str(relative).replace("\\", "/")
return match.group(0)
rewritten = pattern.sub(_replace, original)
if changed:
try:
file_path.write_text(rewritten, encoding="utf-8")
except OSError:
pass
def _hash_filename(url: str, default_ext: str = ".bin") -> str:
parsed = urlsplit(url)
ext = Path(parsed.path).suffix or default_ext
digest = hashlib.sha1(url.encode("utf-8", "ignore")).hexdigest()
return f"{digest}{ext}"
def download_external_images(output_dir: Path, base_url: str) -> None:
"""Download external img/src assets and rewrite HTML to local relative paths."""
base_host = urlsplit(base_url).netloc
external_dir = output_dir / "external_assets"
external_dir.mkdir(parents=True, exist_ok=True)
img_pattern = re.compile(
r'(<img[^>]+src=["\'])(?P<src>https?:\/\/[^"\']+)(["\'])',
flags=re.IGNORECASE,
)
replacements: dict[str, Path] = {}
html_files = [
p for p in output_dir.rglob("*") if p.suffix.lower() in {".html", ".htm"}
]
for file_path in html_files:
try:
content = file_path.read_text(encoding="utf-8", errors="ignore")
except OSError:
continue
changed = False
def _handle_match(match: re.Match[str]) -> str:
nonlocal changed
src_url = match.group("src")
host = urlsplit(src_url).netloc
if not host or host == base_host:
return match.group(0)
if src_url not in replacements:
filename = _hash_filename(src_url, default_ext=".img")
dest_path = external_dir / filename
if not dest_path.exists():
try:
with urlopen(src_url, timeout=20) as resp, open(
dest_path, "wb"
) as out_f:
shutil.copyfileobj(resp, out_f)
except (URLError, OSError):
return match.group(0)
replacements[src_url] = dest_path
dest_path = replacements[src_url]
relative = Path(
os.path.relpath(dest_path.resolve(), start=file_path.parent.resolve())
)
changed = True
new_src = str(relative).replace("\\", "/")
return f"{match.group(1)}{new_src}{match.group(3)}"
rewritten = img_pattern.sub(_handle_match, content)
if changed:
try:
file_path.write_text(rewritten, encoding="utf-8")
except OSError:
pass
def parse_args(argv: list[str] | None = None) -> argparse.Namespace:
parser = argparse.ArgumentParser(
description="Mirror https://blog.sixhz.com/ into a local static site directory."
)
parser.add_argument(
"--url",
default=DEFAULT_URL,
help="Root URL to mirror (default: %(default)s)",
)
parser.add_argument(
"--output-dir",
default=DEFAULT_OUTPUT_DIR,
help="Directory (relative to project root) to store the mirrored site (default: %(default)s)",
)
clean_group = parser.add_mutually_exclusive_group()
clean_group.add_argument(
"--clean",
dest="clean",
action="store_true",
help="Start from a fresh temporary directory instead of seeding from the existing output (default).",
)
clean_group.add_argument(
"--no-clean",
dest="clean",
action="store_false",
help="Seed the temporary download directory from the existing output before mirroring.",
)
parser.set_defaults(clean=True)
parser.add_argument(
"--spider",
action="store_true",
help="Only run wget spider mode to test links without downloading files.",
)
return parser.parse_args(argv)
def main(argv: list[str] | None = None) -> int:
args = parse_args(argv)
root = project_root()
try:
output_dir = resolve_output_dir(root, args.output_dir)
temp_dir = temp_output_dir(output_dir)
output_dir.parent.mkdir(parents=True, exist_ok=True)
temp_dir.parent.mkdir(parents=True, exist_ok=True)
wget_path = find_wget(root)
except (FileNotFoundError, ValueError) as exc:
print(exc, file=sys.stderr)
return 1
print(f"Using wget at: {wget_path}")
print(f"Output directory: {output_dir}")
print(f"Temporary directory: {temp_dir}")
# Spider-only mode remains available for manual checks.
if args.spider:
cleanup_directory(temp_dir)
spider_command = build_wget_command(wget_path, temp_dir, args.url, spider=True)
print("Running spider command:")
print(" ".join(spider_command))
spider_code = stream_process_output(spider_command)
cleanup_directory(temp_dir)
if spider_code != 0:
print(f"Spider check failed with code {spider_code}", file=sys.stderr)
return spider_code
# Pre-flight: spider the site before attempting a mirror.
cleanup_directory(temp_dir)
spider_command = build_wget_command(wget_path, temp_dir, args.url, spider=True)
print("Running spider check before mirroring:")
print(" ".join(spider_command))
spider_code = stream_process_output(spider_command)
if spider_code != 0:
print(
"Skipping mirroring because spider check failed; keeping existing output.",
file=sys.stderr,
)
cleanup_directory(temp_dir)
return spider_code
try:
seed_source = output_dir if not args.clean else None
prepare_temp_directory(temp_dir, seed_from=seed_source)
except OSError as exc:
print(f"Failed to prepare temporary directory: {exc}", file=sys.stderr)
cleanup_directory(temp_dir)
return 1
mirror_command = build_wget_command(wget_path, temp_dir, args.url, spider=False)
print("Running mirror command:")
print(" ".join(mirror_command))
return_code = stream_process_output(mirror_command)
if return_code != 0:
print(f"wget exited with code {return_code}", file=sys.stderr)
cleanup_directory(temp_dir)
return return_code
# Post-process links to ensure assets point to local copies for offline deploy.
rewrite_links_to_local(temp_dir, args.url)
download_external_images(temp_dir, args.url)
try:
replace_directory(temp_dir, output_dir)
except OSError as exc:
print(f"Failed to replace output directory: {exc}", file=sys.stderr)
cleanup_directory(temp_dir)
return 1
return return_code
if __name__ == "__main__":
sys.exit(main())可以参考run_mirror.sh，借助Github，将本项目部署在服务器上，以实现自动任务。
#!/bin/bash
set -e
# 确保能找到 uv 等命令
export PATH="/root/.local/bin:/usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin"
# 1. 到仓库目录
cd "$(dirname "$0")"
# 2. 确保是最新代码（可选）
git pull --rebase || true
# 3. 运行镜像脚本
uv run python mirror.py
# 4. 把新生成的内容加入 git
git add .
# 5. 提交（如果没有变更就跳过）
if ! git diff --cached --quiet; then
git commit -m "chore: daily mirror $(date -Iseconds)"
git push origin master
fi本项目基于 GNU GPL v3，你可以自由地使用、分发、修改。
作用：让不开放公网 80/443 端口的 NAT 小主机，也可以通过一个 HTTPS 域名被全球访问。
路径：用户浏览器 → Cloudflare 网络 → (Tunnel) → NAT 机
控制台 - 外观 - 编辑当前外观 - 编辑文件 header.php ：
<!-- 添加网站图标 -->
<link rel="apple-touch-icon" sizes="180x180" href="https://academic.sixhz.com/images/favicon/apple-touch-icon.png" />
<link rel="icon" type="image/x-icon" href="https://academic.sixhz.com/images/favicon/favicon.ico" />
<link rel="icon" type="image/png" sizes="512x512" href="https://academic.sixhz.com/images/favicon/android-chrome-512x512.png" />
<link rel="icon" type="image/png" sizes="192x192" href="https://academic.sixhz.com/images/favicon/android-chrome-192x192.png" />
<link rel="icon" type="image/png" sizes="32x32" href="https://academic.sixhz.com/images/favicon/favicon-32x32.png" />
<link rel="icon" type="image/png" sizes="16x16" href="https://academic.sixhz.com/images/favicon/favicon-16x16.png" />
<link rel="manifest" href="https://academic.sixhz.com/images/favicon/site.webmanifest" />本站当前使用IDC服务站为"博客列车"，感谢站长提供的托管服务。
网址为： https://blogtrain.com/
活动地址： https://idcflare.com/t/topic/32449
优惠码为IDCFLARE，购买日期为2025年10月25日。
优惠码使用后免费1年，第二年继续免费续期，每次续期1年。
在这里记录一下，以免忘记了。
