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"""

gui/progress.py

@@ -0,0 +1,97 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+进度信息模块
+
+定义用于进度回调的数据结构。
+"""
+
+from dataclasses import dataclass, field
+from typing import Dict, Any, Optional
+
+
+@dataclass
+class ProgressInfo:
+    """
+    进度信息
+
+    用于在后台任务和 GUI 之间传递进度信息。
+    """
+    stage: str  # 阶段标识: 'login', 'query', 'download', 'logout', 'convert', 'complete'
+    current: int  # 当前进度值
+    total: int  # 总量
+    message: str  # 显示给用户的消息
+    detail: Dict[str, Any] = field(default_factory=dict)  # 额外详细信息
+
+    @property
+    def percent(self) -> int:
+        """计算进度百分比（0-100）"""
+        if self.total <= 0:
+            return 0
+        return int(self.current * 100 / self.total)
+
+    def __repr__(self) -> str:
+        return f"ProgressInfo(stage={self.stage}, {self.current}/{self.total}, {self.message})"
+
+
+class ProgressCalculator:
+    """
+    进度计算器
+
+    将各阶段的进度映射到总体进度百分比。
+    """
+
+    # 各阶段在总进度中的占比
+    STAGE_WEIGHTS = {
+        'login': 5,      # 登录: 0-5%
+        'query': 5,      # 查询: 5-10%
+        'download': 65,  # 下载: 10-75%
+        'logout': 5,     # 注销: 75-80%
+        'convert': 15,   # 转换: 80-95%
+        'complete': 5,   # 完成: 95-100%
+    }
+
+    def __init__(self):
+        """初始化进度计算器"""
+        self._stage_offsets = self._calculate_stage_offsets()
+
+    def _calculate_stage_offsets(self) -> Dict[str, int]:
+        """计算各阶段的起始偏移量（百分比）"""
+        offsets = {}
+        offset = 0
+        for stage, weight in self.STAGE_WEIGHTS.items():
+            offsets[stage] = offset
+            offset += weight
+        return offsets
+
+    def calculate_overall_percent(self, progress: ProgressInfo) -> int:
+        """
+        计算总体进度百分比
+
+        Args:
+            progress: 进度信息
+
+        Returns:
+            总体进度百分比 (0-100)
+        """
+        stage = progress.stage
+
+        if stage == 'complete':
+            return 100
+
+        if stage not in self._stage_offsets:
+            return 0
+
+        # 计算阶段起始百分比
+        stage_offset = self._stage_offsets[stage]
+
+        # 计算阶段内的进度百分比
+        stage_percent = progress.percent
+
+        # 计算该阶段的权重
+        stage_weight = self.STAGE_WEIGHTS[stage]
+
+        # 总进度 = 阶段偏移 + (阶段内进度 * 阶段权重 / 100)
+        overall = stage_offset + int(stage_percent * stage_weight / 100)
+
+        return min(overall, 100)

gui/utils.py

@@ -0,0 +1,39 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+GUI 工具模块
+
+提供 GUI 相关的工具类和函数。
+"""
+
+
+class RealtimeOutput:
+    """实时输出流，每次写入立即回调通知"""
+
+    def __init__(self, callback):
+        """
+        初始化实时输出流
+
+        Args:
+            callback: 写入时的回调函数，接收文本内容
+        """
+        self.callback = callback
+        self.buffer = []
+
+    def write(self, text):
+        """写入文本"""
+        if text:
+            # 将文本按行分割，逐行回调
+            lines = text.split('\n')
+            for line in lines:
+                if line:  # 忽略空行（由 split 产生）
+                    self.callback(line)
+        return len(text)
+
+    def flush(self):
+        """刷新（兼容性方法）"""
+        pass
+
+    def isatty(self):
+        """返回 False，表示不是终端"""
+        return False