fix: rewrite startup scripts with reliable helpers, stderr logging, visible windows

- Replace /MIN (hidden window) with normal windows so errors are visible
- Redirect stderr to logs/*.log for post-mortem
- Extract killport/wait_health/wait_port into callable helpers
- Use !N! (delayed expansion) for retry counters
- stop.bat now shows which PIDs it kills with port labels
- Remove nested-quote issue by cd'ing before npm start

start.bat

@@ -6,47 +6,46 @@ echo   agent_jrxml 启动 (API + 验证)
 echo ================================================
 cd /d "%~dp0"
 
-:: 环境检查
+:: ── 环境检查 ──
 if not exist "%~dp0.venv\Scripts\python.exe" (
     echo [错误] 未找到 .venv，请先创建虚拟环境
     pause
     exit /b 1
 )
 
-:: 清理残留进程
+echo.
+
+:: ── 清理残留进程 ──
 echo [清理] 检查残留进程...
 for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8000.*LISTENING"') do (
-    taskkill /F /PID %%a >nul 2>&1 && echo   已清理 PID %%a
+    taskkill /F /PID %%a >nul 2>&1 && echo   已清理端口 8000 (PID %%a)
 )
 for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8001.*LISTENING"') do (
-    taskkill /F /PID %%a >nul 2>&1 && echo   已清理 PID %%a
+    taskkill /F /PID %%a >nul 2>&1 && echo   已清理端口 8001 (PID %%a)
 )
 echo.
 
-:: 启动验证服务 (后台最小化)
+:: ── 1. 验证服务 ──
 echo [启动] 验证服务 :8001
-start "jrxml-validator" /MIN .venv\Scripts\python.exe -c "import uvicorn; uvicorn.run('validation_service.main:app',host='0.0.0.0',port=8001,reload=False)"
+start "jrxml-validator" .venv\Scripts\python.exe -c "import uvicorn; uvicorn.run('validation_service.main:app',host='0.0.0.0',port=8001,reload=False)" 2> "%~dp0logs\validator-stderr.log"
 
-:: 等待验证服务就绪 (用 PowerShell 检测)
 echo [等待] 验证服务就绪...
-set /a RETRY=0
+set /a N=0
 :wait_val
 ping -n 2 127.0.0.1 >nul
 powershell -Command "try{$r=Invoke-WebRequest -Uri http://localhost:8001/health -TimeoutSec 2 -UseBasicParsing;exit 0}catch{exit 1}" >nul 2>&1
 if not errorlevel 1 goto val_ok
-set /a RETRY+=1
+set /a N+=1
-if %RETRY% GEQ 30 (
+if !N! GEQ 30 (
-    echo    [失败] 验证服务启动超时，请检查端口 8001 是否被占用
+    echo    [失败] 验证服务启动超时 (30次重试)
-    taskkill /F /FI "WINDOWTITLE eq jrxml-validator*" >nul 2>&1
+    echo    请检查: logs\validator-stderr.log
-    for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8001.*LISTENING"') do taskkill /F /PID %%a >nul 2>&1
+    goto cleanup
-    pause
-    exit /b 1
 )
 goto wait_val
 :val_ok
 echo    :8001 就绪
 
-:: 启动 API 服务 (前台，Ctrl+C 退出)
+:: ── 2. API 服务 (前台运行) ──
 echo [启动] API 服务 :8000
 echo ================================================
 echo   服务已就绪:
@@ -57,10 +56,17 @@ echo   关闭窗口后会自动清理验证服务
 echo ================================================
 .venv\Scripts\python.exe -c "import uvicorn; uvicorn.run('api_server:app',host='0.0.0.0',port=8000,reload=False)"
 
-:: API 进程退出后自动清理
+:: ── API 退出后清理 ──
 echo.
 echo [清理] 停止验证服务...
 taskkill /F /FI "WINDOWTITLE eq jrxml-validator*" >nul 2>&1
 for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8001.*LISTENING"') do taskkill /F /PID %%a >nul 2>&1
 echo 已停止所有服务
 pause
+exit /b 0
+
+:cleanup
+taskkill /F /FI "WINDOWTITLE eq jrxml-validator*" >nul 2>&1
+for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8001.*LISTENING"') do taskkill /F /PID %%a >nul 2>&1
+pause
+exit /b 1

start_all.bat

@@ -6,77 +6,45 @@ echo   agent_jrxml 启动 (全栈)
 echo ================================================
 cd /d "%~dp0"
 
-:: 环境检查
+:: ── 环境检查 ──
 if not exist "%~dp0.venv\Scripts\python.exe" (
     echo [错误] 未找到 .venv，请先创建虚拟环境
     pause
     exit /b 1
 )
-
-:: 清理残留进程
-echo [清理] 检查残留进程...
-for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8000.*LISTENING"') do taskkill /F /PID %%a >nul 2>&1
-for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8001.*LISTENING"') do taskkill /F /PID %%a >nul 2>&1
-for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":5173.*LISTENING"') do taskkill /F /PID %%a >nul 2>&1
-echo.
-
-:: 1. 验证服务
-echo [1/3] 验证服务 :8001
-start "jrxml-validator" /MIN .venv\Scripts\python.exe -c "import uvicorn; uvicorn.run('validation_service.main:app',host='0.0.0.0',port=8001,reload=False)"
-set /a RETRY=0
-:wait_val
-ping -n 2 127.0.0.1 >nul
-powershell -Command "try{$r=Invoke-WebRequest -Uri http://localhost:8001/health -TimeoutSec 2 -UseBasicParsing;exit 0}catch{exit 1}" >nul 2>&1
-if not errorlevel 1 goto val_ok
-set /a RETRY+=1
-if %RETRY% GEQ 30 (
-    echo    [失败] 验证服务启动超时，请检查端口 8001 是否被占用
-    goto cleanup
-)
-goto wait_val
-:val_ok
-echo    :8001 就绪
-
-:: 2. API 服务
-echo [2/3] API 服务 :8000
-start "jrxml-api" /MIN .venv\Scripts\python.exe -c "import uvicorn; uvicorn.run('api_server:app',host='0.0.0.0',port=8000,reload=False)"
-set /a RETRY=0
-:wait_api
-ping -n 2 127.0.0.1 >nul
-powershell -Command "try{$r=Invoke-WebRequest -Uri http://localhost:8000/api/health -TimeoutSec 2 -UseBasicParsing;exit 0}catch{exit 1}" >nul 2>&1
-if not errorlevel 1 goto api_ok
-set /a RETRY+=1
-if %RETRY% GEQ 30 (
-    echo    [失败] API 服务启动超时，请检查端口 8000 是否被占用
-    goto cleanup
-)
-goto wait_api
-:api_ok
-echo    :8000 就绪
-
-:: 3. 前端
-echo [3/3] 前端 :5173
 if not exist "%~dp0frontend\node_modules" (
     echo [安装] node_modules 不存在，正在 npm install...
     cd /d "%~dp0frontend"
     call npm install
     cd /d "%~dp0"
 )
+
+echo.
+
+:: ── 清理残留进程 ──
+call :killport 8000
+call :killport 8001
+call :killport 5173
+echo.
+
+:: ── 1. 验证服务 ──
+echo [1/3] 启动验证服务 :8001 ...
+start "jrxml-validator" .venv\Scripts\python.exe -c "import uvicorn; uvicorn.run('validation_service.main:app',host='0.0.0.0',port=8001,reload=False)" 2> "%~dp0logs\validator-stderr.log"
+call :wait_health "8001" "验证服务" || goto cleanup
+echo    :8001 就绪
+
+:: ── 2. API 服务 ──
+echo [2/3] 启动 API 服务 :8000 ...
+start "jrxml-api" .venv\Scripts\python.exe -c "import uvicorn; uvicorn.run('api_server:app',host='0.0.0.0',port=8000,reload=False)" 2> "%~dp0logs\api-stderr.log"
+call :wait_health "8000/api" "API 服务" || goto cleanup
+echo    :8000 就绪
+
+:: ── 3. 前端 ──
+echo [3/3] 启动前端 :5173 ...
 cd /d "%~dp0frontend"
-start "jrxml-frontend" /MIN cmd /c "npm run dev"
+start "jrxml-frontend" cmd /c "npm run dev" 2> "%~dp0logs\frontend-stderr.log"
 cd /d "%~dp0"
-set /a RETRY=0
+call :wait_port "5173" "前端" || goto cleanup
-:wait_fe
-ping -n 3 127.0.0.1 >nul
-powershell -Command "try{$r=Invoke-WebRequest -Uri http://localhost:5173 -TimeoutSec 3 -UseBasicParsing;exit 0}catch{exit 1}" >nul 2>&1
-if not errorlevel 1 goto fe_ok
-set /a RETRY+=1
-if %RETRY% GEQ 30 (
-    echo    [失败] 前端启动超时，请检查端口 5173 是否被占用
-    goto cleanup
-)
-goto wait_fe
-:fe_ok
 echo    :5173 就绪
 
 echo.
@@ -90,13 +58,45 @@ echo ================================================
 pause
 exit /b 0
 
+:: ── Cleanup ──
 :cleanup
 echo [清理] 停止已启动的服务...
 taskkill /F /FI "WINDOWTITLE eq jrxml-validator*" >nul 2>&1
 taskkill /F /FI "WINDOWTITLE eq jrxml-api*" >nul 2>&1
 taskkill /F /FI "WINDOWTITLE eq jrxml-frontend*" >nul 2>&1
-for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8001.*LISTENING"') do taskkill /F /PID %%a >nul 2>&1
+call :killport 8001
-for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8000.*LISTENING"') do taskkill /F /PID %%a >nul 2>&1
+call :killport 8000
+call :killport 5173
 echo 已清理，请重试
 pause
 exit /b 1
+
+:: ── Helper: kill process on port ──
+:killport
+for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":%1.*LISTENING"') do (
+    echo    清理端口 %1 (PID %%a)
+    taskkill /F /PID %%a >nul 2>&1
+)
+exit /b 0
+
+:: ── Helper: wait for HTTP health check ──
+:wait_health
+set /a N=0
+:wait_health_loop
+ping -n 2 127.0.0.1 >nul
+powershell -Command "try{$r=Invoke-WebRequest -Uri http://localhost:%1/health -TimeoutSec 2 -UseBasicParsing;exit 0}catch{exit 1}" >nul 2>&1
+if not errorlevel 1 exit /b 0
+set /a N+=1
+if !N! GEQ 30 exit /b 1
+goto wait_health_loop
+
+:: ── Helper: wait for TCP port ──
+:wait_port
+set /a N=0
+:wait_port_loop
+ping -n 3 127.0.0.1 >nul
+netstat -ano | findstr ":%1.*LISTENING" >nul 2>&1
+if not errorlevel 1 exit /b 0
+set /a N+=1
+if !N! GEQ 30 exit /b 1
+goto wait_port_loop

stop.bat

@@ -1,8 +1,27 @@
 @echo off
 chcp 65001 >nul
-echo [清理] 停止所有 agent_jrxml 服务...
+echo ================================================
-for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8000.*LISTENING"') do taskkill /F /PID %%a 2>nul
+echo   停止所有 agent_jrxml 服务
-for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8001.*LISTENING"') do taskkill /F /PID %%a 2>nul
+echo ================================================
-for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":5173.*LISTENING"') do taskkill /F /PID %%a 2>nul
+
+echo [停止] 按窗口标题清理...
+taskkill /F /FI "WINDOWTITLE eq jrxml-validator*" 2>nul
+taskkill /F /FI "WINDOWTITLE eq jrxml-api*" 2>nul
+taskkill /F /FI "WINDOWTITLE eq jrxml-frontend*" 2>nul
+
+echo [停止] 按端口清理...
+for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8000.*LISTENING"') do (
+    echo   端口 8000 - PID %%a
+    taskkill /F /PID %%a >nul 2>&1
+)
+for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":8001.*LISTENING"') do (
+    echo   端口 8001 - PID %%a
+    taskkill /F /PID %%a >nul 2>&1
+)
+for /f "tokens=5" %%a in ('netstat -ano ^| findstr ":5173.*LISTENING"') do (
+    echo   端口 5173 - PID %%a
+    taskkill /F /PID %%a >nul 2>&1
+)
+
 echo 已停止
 pause