添加环境变量映射，支持完整性检查任务和工具类任务的执行

etl_billiards/gui/workers/task_worker.py

@@ -0,0 +1,175 @@
+# -*- coding: utf-8 -*-
+"""任务执行工作线程"""
+
+import subprocess
+import sys
+import os
+from pathlib import Path
+from typing import List, Optional, Dict
+
+from PySide6.QtCore import QThread, Signal
+
+from ..utils.app_settings import app_settings
+
+
+class TaskWorker(QThread):
+    """任务执行工作线程"""
+    
+    # 信号
+    output_received = Signal(str)  # 收到输出行
+    task_finished = Signal(int, str)  # 任务完成 (exit_code, summary)
+    error_occurred = Signal(str)  # 发生错误
+    progress_updated = Signal(int, int)  # 进度更新 (current, total)
+    
+    def __init__(self, command: List[str], working_dir: Optional[str] = None, 
+                 extra_env: Optional[Dict[str, str]] = None, parent=None):
+        super().__init__(parent)
+        self.command = command
+        self.extra_env = extra_env or {}
+        
+        # 工作目录优先级: 参数 > 应用设置 > 自动检测
+        if working_dir is not None:
+            self.working_dir = working_dir
+        elif app_settings.etl_project_path:
+            self.working_dir = app_settings.etl_project_path
+        else:
+            # 回退到源码目录
+            self.working_dir = str(Path(__file__).resolve().parents[2])
+        
+        self.process: Optional[subprocess.Popen] = None
+        self._stop_requested = False
+        self._exit_code: Optional[int] = None
+        self._output_lines: List[str] = []
+    
+    def run(self):
+        """执行任务"""
+        try:
+            self._stop_requested = False
+            self._output_lines = []
+            
+            # 设置环境变量
+            env = os.environ.copy()
+            env["PYTHONIOENCODING"] = "utf-8"
+            env["PYTHONUNBUFFERED"] = "1"
+            
+            # 添加项目根目录到 PYTHONPATH
+            project_root = self.working_dir
+            existing_path = env.get("PYTHONPATH", "")
+            if existing_path:
+                env["PYTHONPATH"] = f"{project_root}{os.pathsep}{existing_path}"
+            else:
+                env["PYTHONPATH"] = project_root
+            
+            # 添加额外的环境变量
+            if self.extra_env:
+                for key, value in self.extra_env.items():
+                    env[key] = str(value)
+                    self.output_received.emit(f"[环境变量] {key}={value}")
+            
+            self.output_received.emit(f"[工作目录] {self.working_dir}")
+            self.output_received.emit(f"[执行命令] {' '.join(self.command)}")
+            
+            # 启动进程
+            self.process = subprocess.Popen(
+                self.command,
+                stdout=subprocess.PIPE,
+                stderr=subprocess.STDOUT,
+                text=True,
+                encoding="utf-8",
+                errors="replace",
+                cwd=self.working_dir,
+                env=env,
+                creationflags=subprocess.CREATE_NO_WINDOW if sys.platform == "win32" else 0,
+            )
+            
+            # 读取输出
+            if self.process.stdout:
+                for line in iter(self.process.stdout.readline, ""):
+                    if self._stop_requested:
+                        break
+                    
+                    line = line.rstrip("\n\r")
+                    if line:
+                        self._output_lines.append(line)
+                        self.output_received.emit(line)
+                        
+                        # 解析进度信息（如果有）
+                        self._parse_progress(line)
+            
+            # 等待进程结束
+            if self.process:
+                self.process.wait()
+                self._exit_code = self.process.returncode
+            
+            # 生成摘要
+            summary = self._generate_summary()
+            self.task_finished.emit(self._exit_code or 0, summary)
+            
+        except FileNotFoundError as e:
+            self.error_occurred.emit(f"找不到 Python 解释器: {e}")
+            self.task_finished.emit(-1, f"执行失败: {e}")
+        except Exception as e:
+            self.error_occurred.emit(f"执行出错: {e}")
+            self.task_finished.emit(-1, f"执行失败: {e}")
+        finally:
+            self.process = None
+    
+    def stop(self):
+        """停止任务"""
+        self._stop_requested = True
+        if self.process:
+            try:
+                self.process.terminate()
+                # 给进程一些时间来终止
+                try:
+                    self.process.wait(timeout=5)
+                except subprocess.TimeoutExpired:
+                    self.process.kill()
+            except Exception:
+                pass
+    
+    def _parse_progress(self, line: str):
+        """解析进度信息"""
+        # 尝试从日志中解析进度
+        # 示例: "[INFO] 处理进度: 50/100"
+        import re
+        match = re.search(r'进度[:\s]*(\d+)/(\d+)', line)
+        if match:
+            current = int(match.group(1))
+            total = int(match.group(2))
+            self.progress_updated.emit(current, total)
+    
+    def _generate_summary(self) -> str:
+        """生成执行摘要"""
+        if not self._output_lines:
+            return "无输出"
+        
+        # 查找关键信息
+        summary_parts = []
+        
+        for line in self._output_lines[-20:]:  # 只看最后 20 行
+            line_lower = line.lower()
+            if "success" in line_lower or "完成" in line or "成功" in line:
+                summary_parts.append(line)
+            elif "error" in line_lower or "失败" in line or "错误" in line:
+                summary_parts.append(line)
+            elif "inserted" in line_lower or "updated" in line_lower:
+                summary_parts.append(line)
+            elif "fetched" in line_lower or "抓取" in line:
+                summary_parts.append(line)
+        
+        if summary_parts:
+            return "\n".join(summary_parts[-5:])  # 最多返回 5 行
+        
+        # 如果没有找到关键信息，返回最后一行
+        return self._output_lines[-1] if self._output_lines else "执行完成"
+    
+    @property
+    def exit_code(self) -> Optional[int]:
+        """获取退出码"""
+        return self._exit_code
+    
+    @property
+    def output(self) -> str:
+        """获取完整输出"""
+        return "\n".join(self._output_lines)