feat: add progress callback mechanism for real-time updates in data extraction

- Add ProgressInfo and ProgressCalculator classes for structured progress tracking
- Implement RealtimeOutput for immediate log display without buffering
- Add progress callback support to DiscreteMaterialPlanExtractor
- Use queue-based thread-safe communication for GUI progress updates
- Fix log output issue where all content appeared at once after completion

Co-Authored-By: Claude Sonnet 4.5 <noreply@anthropic.com>

gui/data_extraction_tab.py

@@ -9,13 +9,15 @@
 import os
 import sys
 import threading
+import queue
 import tkinter as tk
 from tkinter import ttk, filedialog, messagebox
 from pathlib import Path
-from io import StringIO
 from contextlib import redirect_stdout
 from gui.widgets import FileSelector, LogText
 from gui.config_manager import ConfigManager
+from gui.progress import ProgressInfo, ProgressCalculator
+from gui.utils import RealtimeOutput
 
 
 class DataExtractionTab(ttk.Frame):
@@ -34,6 +36,11 @@ class DataExtractionTab(ttk.Frame):
         self.extracting = False
         self.extractor = None
         self.extraction_thread = None
+        self.progress_calculator = ProgressCalculator()
+        self.progress_queue = queue.Queue()  # 进度更新队列
+
+        # 启动进度更新轮询
+        self._poll_progress_queue()
 
         self.create_widgets()
 
@@ -163,6 +170,7 @@ class DataExtractionTab(ttk.Frame):
         self.start_button.config(state=tk.DISABLED)
         self.stop_button.config(state=tk.NORMAL)
         self.progress_bar['value'] = 0
+        self.status_label.config(text="正在登录...")
         self.log_text.clear()
         self.log_text.info("开始数据提取...")
 
@@ -195,25 +203,24 @@ class DataExtractionTab(ttk.Frame):
                 verbose=self.verbose_var.get()
             )
 
-            # 捕获 stdout 输出
-            captured_output = StringIO()
+            # 创建实时输出流，每次写入立即更新 GUI
+            realtime_output = RealtimeOutput(lambda line: self._update_log(line, "INFO"))
 
-            # 重定向 stdout 并执行提取
-            with redirect_stdout(captured_output):
+            # 创建进度回调函数
+            def progress_callback(progress_info: ProgressInfo):
+                # 计算总体进度百分比
+                overall_percent = self.progress_calculator.calculate_overall_percent(progress_info)
+                self._update_progress(overall_percent, progress_info.message)
+
+            # 重定向 stdout 并执行提取（带进度回调）
+            with redirect_stdout(realtime_output):
                 result = self.extractor.extract(
                     production_id_file=input_file,
-                    output_file=output_file
+                    output_file=output_file,
+                    progress_callback=progress_callback
                 )
 
-            # 获取捕获的输出并显示到日志
-            output_text = captured_output.getvalue()
-            if output_text:
-                for line in output_text.split('\n'):
-                    if line.strip():
-                        self._update_log(line, "INFO")
-
             if result and self.extracting:
-                self._update_progress(100, "提取完成")
                 self._update_log(f"数据已保存到：{output_file}", "SUCCESS")
             elif not self.extracting:
                 self._update_log("提取已取消", "WARNING")
@@ -233,14 +240,28 @@ class DataExtractionTab(ttk.Frame):
         self.stop_button.config(state=tk.DISABLED)
         self.extractor = None
 
-    def _update_progress(self, value: int, message: str):
-        """线程安全的进度更新"""
-        def update():
-            if self.extracting:
-                self.progress_bar['value'] = value
-                self.status_label.config(text=message)
+    def _poll_progress_queue(self):
+        """轮询进度队列，处理进度更新"""
+        try:
+            while True:
+                # 非阻塞地获取队列中的消息
+                try:
+                    progress_data = self.progress_queue.get_nowait()
+                    value, message = progress_data
+                    self.progress_bar['value'] = value
+                    self.status_label.config(text=message)
+                except queue.Empty:
+                    break
+        finally:
+            # 继续轮询（每 50ms 检查一次）
+            self.after(50, self._poll_progress_queue)
 
-        self.after(0, update)
+    def _update_progress(self, value: int, message: str):
+        """线程安全的进度更新（通过队列）"""
+        try:
+            self.progress_queue.put_nowait((value, message))
+        except:
+            pass  # 队列满时忽略
 
     def _update_log(self, message: str, level: str = "INFO"):
         """线程安全的日志更新"""