feat: add service config templates and extraction script

Former-commit-id: 1de24b7eb79676d1aba9d799a58c5a753290cf52

scripts/verify_frontend_runtime.py

@@ -0,0 +1,252 @@
+#!/usr/bin/env python3
+# -*- coding: utf-8 -*-
+"""
+前端运行时自动验收（Playwright）：
+1) 打开页面并等待首屏稳定
+2) 采集 console/pageerror/requestfailed
+3) 统计重复资源请求（按 URL 去 query 后聚合）
+4) 输出 markdown 报告 + 截图
+"""
+
+from __future__ import annotations
+
+import argparse
+import asyncio
+import datetime as dt
+import json
+import re
+import sys
+from collections import Counter
+from pathlib import Path
+from typing import Dict, List, Tuple
+from urllib.parse import urlsplit, urlunsplit
+
+from playwright.async_api import Error as PlaywrightError
+from playwright.async_api import async_playwright
+
+
+def _strip_query(url: str) -> str:
+    p = urlsplit(url)
+    return urlunsplit((p.scheme, p.netloc, p.path, "", ""))
+
+
+def _now_tag() -> str:
+    return dt.datetime.now().strftime("%Y%m%d-%H%M%S")
+
+
+def _classify_resource(url: str) -> str:
+    path = urlsplit(url).path.lower()
+    if path.endswith(".js"):
+        return "js"
+    if path.endswith(".css"):
+        return "css"
+    if re.search(r"\.(png|jpg|jpeg|gif|svg|ico|webp)$", path):
+        return "image"
+    if "/api/" in path or "/elevator/" in path:
+        return "api"
+    return "other"
+
+
+async def run_check(url: str, wait_ms: int, screenshot_path: Path) -> Dict[str, object]:
+    console_errors: List[str] = []
+    console_warnings: List[str] = []
+    page_errors: List[str] = []
+    failed_requests: List[str] = []
+    requests: List[Tuple[str, str]] = []
+
+    async with async_playwright() as p:
+        browser = await p.chromium.launch(headless=True)
+        context = await browser.new_context(ignore_https_errors=True)
+        page = await context.new_page()
+        page.set_default_timeout(5_000)
+
+        page.on(
+            "console",
+            lambda msg: (
+                console_errors.append(msg.text)
+                if msg.type == "error"
+                else console_warnings.append(msg.text)
+                if msg.type in {"warning", "warn"}
+                else None
+            ),
+        )
+        page.on("pageerror", lambda err: page_errors.append(str(err)))
+        def _on_request_failed(req) -> None:
+            failure = req.failure
+            if isinstance(failure, dict):
+                err_text = failure.get("errorText", "unknown")
+            elif isinstance(failure, str):
+                err_text = failure
+            elif failure is None:
+                err_text = "unknown"
+            else:
+                err_text = getattr(failure, "error_text", str(failure))
+            failed_requests.append(f"{req.url} | {err_text}")
+
+        page.on("requestfailed", _on_request_failed)
+        page.on("requestfinished", lambda req: requests.append((req.method, req.url)))
+
+        navigation_mode = "domcontentloaded"
+        try:
+            try:
+                await page.goto(url, wait_until="domcontentloaded", timeout=12_000)
+            except PlaywrightError:
+                # 某些历史页面会因脚本阻塞导致 domcontentloaded 不触发，降级到 commit 保证可诊断
+                navigation_mode = "commit(fallback)"
+                await page.goto(url, wait_until="commit", timeout=20_000)
+            await page.wait_for_timeout(wait_ms)
+
+            try:
+                # 避免 full_page 在超长文档上触发超时，先抓取可视区域。
+                await page.screenshot(path=str(screenshot_path), full_page=False, timeout=15_000)
+            except Exception:
+                screenshot_path.write_bytes(b"")
+
+            try:
+                title = await page.title()
+            except Exception:
+                title = ""
+
+            try:
+                body_non_empty = bool(await page.evaluate("Boolean(document.body && document.body.innerText && document.body.innerText.trim())"))
+            except Exception:
+                body_non_empty = False
+
+            try:
+                app_count = await page.locator("#app *").count()
+            except Exception:
+                app_count = -1
+        except PlaywrightError as e:
+            await browser.close()
+            return {
+                "fatal_error": str(e),
+                "console_errors": console_errors,
+                "console_warnings": console_warnings,
+                "page_errors": page_errors,
+                "failed_requests": failed_requests,
+                "requests": requests,
+            }
+
+        await browser.close()
+
+    normalized_urls = [_strip_query(u) for _, u in requests]
+    url_counter = Counter(normalized_urls)
+    duplicate_urls = [(u, c) for u, c in url_counter.items() if c > 1]
+    duplicate_urls.sort(key=lambda x: x[1], reverse=True)
+
+    resource_counter = Counter(_classify_resource(u) for u in normalized_urls)
+    return {
+        "fatal_error": "",
+        "navigation_mode": navigation_mode,
+        "title": title,
+        "body_non_empty": body_non_empty,
+        "app_count": app_count,
+        "console_errors": console_errors,
+        "console_warnings": console_warnings,
+        "page_errors": page_errors,
+        "failed_requests": failed_requests,
+        "requests_total": len(requests),
+        "requests_unique": len(url_counter),
+        "resource_counter": dict(resource_counter),
+        "duplicate_urls": duplicate_urls[:30],
+    }
+
+
+def _render_report(url: str, result: Dict[str, object], screenshot_path: Path) -> str:
+    lines: List[str] = []
+    lines.append("# 前端运行时自动验收报告")
+    lines.append("")
+    lines.append(f"- URL: `{url}`")
+    lines.append(f"- 时间: `{dt.datetime.now().isoformat(timespec='seconds')}`")
+    lines.append(f"- 截图: `{screenshot_path}`")
+    lines.append("")
+
+    fatal = str(result.get("fatal_error", "")).strip()
+    if fatal:
+        lines.append("## 结论")
+        lines.append("- **失败**：页面打开阶段异常")
+        lines.append(f"- 异常: `{fatal}`")
+        return "\n".join(lines) + "\n"
+
+    lines.append("## 页面状态")
+    lines.append(f"- 导航模式: `{result.get('navigation_mode', '')}`")
+    lines.append(f"- title: `{result.get('title', '')}`")
+    lines.append(f"- body 是否非空: `{result.get('body_non_empty', False)}`")
+    lines.append(f"- `#app` 子节点数: `{result.get('app_count', 0)}`")
+    lines.append("")
+
+    lines.append("## 请求统计")
+    lines.append(f"- 总请求数: `{result.get('requests_total', 0)}`")
+    lines.append(f"- 去重后 URL 数: `{result.get('requests_unique', 0)}`")
+    lines.append(f"- 类型分布: `{json.dumps(result.get('resource_counter', {}), ensure_ascii=False)}`")
+    lines.append("")
+
+    dup = result.get("duplicate_urls", [])
+    lines.append("## 重复请求（Top）")
+    if dup:
+        for u, c in dup:
+            lines.append(f"- `{c}x` {u}")
+    else:
+        lines.append("- 无重复 URL 请求")
+    lines.append("")
+
+    def _section(title: str, rows: List[str]) -> None:
+        lines.append(f"## {title}")
+        if rows:
+            for row in rows[:50]:
+                lines.append(f"- {row}")
+        else:
+            lines.append("- 无")
+        lines.append("")
+
+    _section("Console Error", result.get("console_errors", []))
+    _section("Page Error", result.get("page_errors", []))
+    _section("Request Failed", result.get("failed_requests", []))
+    _section("Console Warning", result.get("console_warnings", []))
+
+    return "\n".join(lines)
+
+
+async def _amain() -> int:
+    parser = argparse.ArgumentParser(description="自动检测前端运行时错误/重复加载")
+    parser.add_argument("--url", default="http://127.0.0.1:8090/login", help="待检测 URL")
+    parser.add_argument(
+        "--wait-ms", type=int, default=5000, help="页面打开后额外等待毫秒数（默认 5000）"
+    )
+    parser.add_argument(
+        "--out-dir",
+        default="artifacts/frontend-check",
+        help="报告与截图输出目录（默认 artifacts/frontend-check）",
+    )
+    args = parser.parse_args()
+
+    out_dir = Path(args.out_dir).resolve()
+    out_dir.mkdir(parents=True, exist_ok=True)
+    tag = _now_tag()
+    screenshot = out_dir / f"frontend-{tag}.png"
+    report = out_dir / f"frontend-{tag}.md"
+
+    result = await run_check(args.url, args.wait_ms, screenshot)
+    report.write_text(_render_report(args.url, result, screenshot), encoding="utf-8")
+
+    print(f"[frontend-check] report={report}")
+    print(f"[frontend-check] screenshot={screenshot}")
+    if result.get("fatal_error"):
+        return 2
+    if result.get("console_errors") or result.get("page_errors") or result.get("failed_requests"):
+        return 1
+    return 0
+
+
+def main() -> int:
+    try:
+        return asyncio.run(_amain())
+    except KeyboardInterrupt:
+        return 130
+    except Exception as e:
+        print(f"[frontend-check] unexpected error: {e}", file=sys.stderr)
+        return 2
+
+
+if __name__ == "__main__":
+    raise SystemExit(main())