数据集搜集器（百科）008

对数据集搜集器（百科）007进行一下改进：
错误处理：增加更多的错误处理，比如网络请求超时、解析错误等。
用户界面：增加一些提示信息，让用户更清楚当前的操作状态。
多线程处理：确保多线程处理更加安全，避免多个线程同时操作同一资源。
日志记录：增加更多的日志记录，方便调试和追踪问题。
配置文件：引入配置文件，方便修改一些常量（如记录文件夹名称）。
完善后的代码
python

import tkinter as tk
from tkinter import filedialog, messagebox
import requests
from bs4 import BeautifulSoup
import json
import os
import threading
from tkinter import ttk
import logging# 配置日志
logging.basicConfig(filename='app.log', level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')# 读取配置文件
CONFIG_FILE = 'config.json'
DEFAULT_CONFIG = {"record_folder": "记录","log_file": "app.log"
}def load_config():if os.path.exists(CONFIG_FILE):with open(CONFIG_FILE, 'r', encoding='utf-8') as file:return json.load(file)return DEFAULT_CONFIGconfig = load_config()class BaikeSearchApp:def __init__(self, root):self.root = rootself.root.title("百度百科查询工具")# 创建输入框self.input_label = tk.Label(root, text="输入问题:")self.input_label.pack(pady=5)self.input_entry = tk.Entry(root, width=80)self.input_entry.pack(pady=5)# 创建文本框self.text = tk.Text(root, wrap='word', height=20, width=80)self.text.pack(pady=10)# 创建按钮self.load_button = tk.Button(root, text="加载文件", command=self.load_file)self.load_button.pack(side=tk.LEFT, padx=10)self.query_button = tk.Button(root, text="获取回答", command=self.get_answer)self.query_button.pack(side=tk.LEFT, padx=10)self.save_button = tk.Button(root, text="保存记录", command=self.save_record)self.save_button.pack(side=tk.LEFT, padx=10)self.history_button = tk.Button(root, text="查看历史记录", command=self.show_history)self.history_button.pack(side=tk.LEFT, padx=10)self.help_button = tk.Button(root, text="帮助", command=self.show_help)self.help_button.pack(side=tk.LEFT, padx=10)# 创建状态栏self.status_var = tk.StringVar()self.status_bar = tk.Label(root, textvariable=self.status_var, bd=1, relief=tk.SUNKEN, anchor=tk.W)self.status_bar.pack(side=tk.BOTTOM, fill=tk.X)# 创建进度条self.progress = ttk.Progressbar(root, orient="horizontal", length=300, mode="determinate")self.progress.pack(pady=10)# 初始化历史记录self.history = []self.root.protocol("WM_DELETE_WINDOW", self.on_closing)def on_closing(self):if hasattr(self, 'thread') and self.thread.is_alive():messagebox.showinfo("提示", "请等待所有任务完成后再关闭窗口。")else:self.root.destroy()def load_file(self):file_path = filedialog.askopenfilename(filetypes=[("Text files", "*.txt")])if file_path:with open(file_path, 'r', encoding='utf-8') as file:lines = file.readlines()total_lines = len(lines)self.progress["maximum"] = total_linesfor i, line in enumerate(lines):self.text.insert(tk.END, f"问题: {line.strip()}\n")self.get_answer(line.strip())self.progress["value"] = i + 1self.root.update_idletasks()self.status_var.set(f"已加载文件: {file_path}")def get_answer(self, query=None):if not query:query = self.input_entry.get().strip()if not query:query = self.text.get("insert linestart", "insert lineend").strip()if not query:messagebox.showwarning("警告", "请先输入或选择一个问题")returnself.status_var.set(f"正在查询: {query}")logging.info(f"开始查询: {query}")self.thread = threading.Thread(target=self._get_answer, args=(query,))self.thread.start()def _get_answer(self, query):url = f"https://baike.baidu.com/item/{query}"try:response = requests.get(url, timeout=10)response.raise_for_status()soup = BeautifulSoup(response.content, 'html.parser')# 从<meta>标签中提取描述description_tag = soup.find('meta', attrs={'name': 'description'})if description_tag and 'content' in description_tag.attrs:content = description_tag['content']else:content = "未找到相关词条"answer = {"question": query,"human_answers": [content],"chatgpt_answers": [content]}formatted_answer = f"问题: {query}\n答案: {content}\n\n"self.text.insert(tk.END, formatted_answer)self.history.append(answer)self.status_var.set(f"查询完成: {query}")logging.info(f"查询完成: {query}")except requests.RequestException as e:self.text.insert(tk.END, f"请求失败: {e}\n")self.status_var.set("请求失败")logging.error(f"请求失败: {e}")def save_record(self):record_folder = config["record_folder"]if not os.path.exists(record_folder):os.makedirs(record_folder)with open(os.path.join(record_folder, "bata.txt"), 'w', encoding='utf-8') as file:for record in self.history:file.write(json.dumps(record, ensure_ascii=False) + "\n")self.status_var.set("记录已保存")def show_history(self):history_window = tk.Toplevel(self.root)history_window.title("历史记录")history_text = tk.Text(history_window, wrap='word', height=20, width=80)history_text.pack(pady=10)for record in self.history:history_text.insert(tk.END, json.dumps(record, ensure_ascii=False) + "\n")clear_button = tk.Button(history_window, text="清空历史记录", command=self.clear_history)clear_button.pack(pady=10)def clear_history(self):self.history = []self.text.delete(1.0, tk.END)self.status_var.set("历史记录已清空")def show_help(self):help_window = tk.Toplevel(self.root)help_window.title("帮助文档")help_text = tk.Text(help_window, wrap='word', height=20, width=80)help_text.pack(pady=10)help_content = """使用说明:1. 在输入框中输入问题，点击“获取回答”按钮查询答案。2. 点击“加载文件”按钮，选择包含问题的文本文件，批量查询答案。3. 查询结果会显示在文本框中，并自动保存到历史记录。4. 点击“保存记录”按钮，将历史记录保存到文件中。5. 点击“查看历史记录”按钮，查看和管理历史记录。6. 点击“帮助”按钮，查看使用说明。"""help_text.insert(tk.END, help_content)if __name__ == "__main__":root = tk.Tk()app = BaikeSearchApp(root)root.mainloop()