# -*- coding: utf-8 -*-
"""生成备份说明 HTML（每个时间戳子目录一份）。"""
from __future__ import annotations
import html
from collections import defaultdict
from pathlib import Path
def _cat_for(path: str) -> str:
p = path.replace("\\", "/").lower()
if p.startswith("platform_snapshot/dify/") or "/dify_" in p or p.startswith("dify") or "dify_" in Path(p).name:
if "langfuse" in p:
return "langfuse"
return "dify"
if p.startswith("platform_snapshot/langfuse/") or "langfuse" in p:
return "langfuse"
if p.startswith("platform_snapshot/xinference/") or "xinference" in p or "bge-m3" in p or "rerank" in p:
return "xinference"
if p.startswith("platform_snapshot/vllm/") or "vllm" in p or "qwen" in p:
return "vllm"
if p.startswith("platform_snapshot/monitoring/") or "作业8" in path or "prometheus" in p or "grafana" in p:
return "monitoring"
if p.startswith("platform_snapshot/"):
return "platform_snapshot"
if p.startswith("docs/"):
return "docs"
if p.startswith("scripts_cursor/") or p.startswith("scripts_bash_cursor/") or p.startswith("patches_cursor/"):
return "programs"
if p.startswith("data/"):
return "data"
if p.endswith(".env") or p.endswith(".yaml") or p.endswith(".yml") or "config" in Path(p).name.lower():
return "configs"
if p.startswith("tools/") or p.startswith("keep/") or p.startswith(".cursor/"):
return "tools"
return "other"
CATEGORY_META = {
"dify": ("Dify 相关", "控制台、Agent 快照、Docker 编排、诊断与修复脚本"),
"langfuse": ("Langfuse 相关", "观测平台 .env、compose、连接 Dify 的脚本与密钥文件"),
"xinference": ("Xinference 相关", "嵌入/重排模型接口探测与相关脚本"),
"vllm": ("vLLM / GPU 推理", "Qwen 推理服务探测、启动脚本与 GPU 快照"),
"monitoring": ("SLA 监控", "Prometheus / Grafana / Exporter 配置"),
"platform_snapshot": ("207 平台快照", "从服务器拉取的运行时配置与状态"),
"docs": ("文档", "Markdown / HTML 专题与手册"),
"programs": ("程序与补丁", "scripts_cursor / scripts_bash_cursor / patches_cursor"),
"configs": ("配置与密钥", "本地 .env、token、密钥文件（请妥善保管）"),
"data": ("运行数据", "作业产出与本地 data/"),
"tools": ("工具与其它", "tools / keep / Cursor rules"),
"other": ("其它", "未归类文件"),
}
def categorize_files(files: list[dict]) -> dict[str, list[dict]]:
buckets: dict[str, list[dict]] = defaultdict(list)
for ent in files:
buckets[_cat_for(ent["path"])].append(ent)
for k in buckets:
buckets[k].sort(key=lambda x: x["path"])
return dict(buckets)
def _fmt_size(n: int) -> str:
if n < 1024:
return f"{n} B"
if n < 1024 * 1024:
return f"{n / 1024:.1f} KB"
return f"{n / (1024 * 1024):.2f} MB"
def render_backup_guide_html(meta: dict, categories: dict[str, list[dict]]) -> str:
bundle_id = html.escape(meta.get("bundle_id", ""))
created = html.escape(meta.get("created_at", ""))
remote = html.escape(meta.get("remote_folder", "AIinfra_Cursor"))
file_count = meta.get("file_count", 0)
total_bytes = meta.get("total_bytes", 0)
zip_bytes = meta.get("zip_bytes", 0)
zip_sha = html.escape(meta.get("zip_sha256", "") or "")
snap = meta.get("remote_snapshot") or {}
cat_sections = []
order = [
"dify",
"langfuse",
"xinference",
"vllm",
"monitoring",
"platform_snapshot",
"docs",
"programs",
"configs",
"data",
"tools",
"other",
]
for key in order:
items = categories.get(key) or []
if not items:
continue
title, desc = CATEGORY_META.get(key, (key, ""))
rows = []
for ent in items[:200]:
rows.append(
"
%s
%s
"
% (html.escape(ent["path"]), _fmt_size(int(ent.get("size", 0))))
)
more = ""
if len(items) > 200:
more = f"
… 另有 {len(items) - 200} 个文件，详见 manifest.json
"
cat_sections.append(
f"""
{html.escape(title)} {len(items)}
{html.escape(desc)}
路径
大小
{''.join(rows)}
{more}
"""
)
snap_html = ""
if snap:
ok_n = len(snap.get("ok") or [])
miss_n = len(snap.get("missing") or [])
err_n = len(snap.get("errors") or [])
snap_html = f"""