# 安裝核心依賴（tkinter一般Python自帶，無需額外安裝）
pip install pandas openpyxl

python table_tool.py

# 關(guān)鍵代碼片段：讀取Excel并提取圖片
if file_ext == ".xlsx":
    # 用openpyxl讀取工作簿（保留圖片）
    wb = load_workbook(file_path, data_only=True)
    sheets = wb.sheetnames
    
    for sheet in sheets:
        ws = wb[sheet]
        # 1. 讀取表格數(shù)據(jù)到DataFrame（處理表頭）
        df = pd.DataFrame(ws.values)
        if len(df) > 0:
            df.columns = df.iloc[0]  # 第一行設(shè)為列名
            df = df.drop(0).reset_index(drop=True)
        data_dict[sheet] = df
        
        # 2. 提取圖片及位置（核心：保留圖片+行列映射）
        sheet_images = []
        for img in ws._images:
            # 轉(zhuǎn)換為1-based行列號(hào)（Excel原生格式）
            row = img.anchor._from.row + 1  
            col = img.anchor._from.col + 1
            sheet_images.append((img, row, col))
        images_dict[sheet] = sheet_images

# 數(shù)據(jù)預(yù)處理：統(tǒng)一轉(zhuǎn)字符串、填充空值（避免類型不一致導(dǎo)致比對(duì)錯(cuò)誤）
df_a[col_a] = df_a[col_a].astype(str).fillna("")
df_b[col_b] = df_b[col_b].astype(str).fillna("")

# 集合運(yùn)算：快速找差異/共有值（效率遠(yuǎn)高于循環(huán)）
a_vals = set(df_a[col_a].unique())
b_vals = set(df_b[col_b].unique())
a_not_b_vals = a_vals - b_vals  # A有B無
b_not_a_vals = b_vals - a_vals  # B有A無
common_vals = a_vals & b_vals   # 共有值

# 篩選對(duì)應(yīng)行
a_not_b_df = df_a[df_a[col_a].isin(a_not_b_vals)].reset_index(drop=True)

# A格式共有數(shù)據(jù)圖片映射示例
a_common_original_rows = df_a[df_a[col_a].isin(common_vals)].index.tolist()
a_common_images = []
for img, orig_row, col in img_a_list:
    # 原Excel行 → 原DataFrame行（Excel行1=表頭，行2=df第0行）
    orig_df_row = orig_row - 2  
    if orig_df_row in a_common_original_rows:
        # 新Excel行 = 新DataFrame行 + 2（表頭占1行）
        new_row = a_common_original_rows.index(orig_df_row) + 2
        a_common_images.append((img, new_row, col))

cleaned_rows = []
current_row = None
orig_to_new_row = {}  # 原行→新行映射字典
new_row_idx = 0

for idx, row in df.iterrows():
    # 關(guān)鍵列非空 → 新行開始
    if str(row[key_col]).strip() != "":
        if current_row is not None:
            cleaned_rows.append(current_row)
            new_row_idx += 1
        current_row = row.to_dict()
        orig_to_new_row[idx] = new_row_idx
    else:
        # 關(guān)鍵列空 → 補(bǔ)充到當(dāng)前行（僅填充空值）
        if current_row is not None:
            for col in df.columns:
                if str(row[col]).strip() != "" and str(current_row[col]).strip() == "":
                    current_row[col] = row[col]
            orig_to_new_row[idx] = new_row_idx

# 保存最后一行
if current_row is not None:
    cleaned_rows.append(current_row)
cleaned_table_result = pd.DataFrame(cleaned_rows)

# 整理后圖片映射
cleaned_table_images = []
for img, orig_row, col in orig_images:
    orig_df_row = orig_row - 2  # Excel行→原DataFrame行
    if orig_df_row in orig_to_new_row:
        # 新Excel行 = 新DataFrame行 + 2
        new_excel_row = orig_to_new_row[orig_df_row] + 2
        cleaned_table_images.append((img, new_excel_row, col))

if file_ext == ".xlsx":
    # 創(chuàng)建新工作簿
    wb = Workbook()
    ws = wb.active
    ws.title = "數(shù)據(jù)"
    
    # 1. 寫入DataFrame數(shù)據(jù)（保留表頭）
    for r in dataframe_to_rows(df, index=False, header=True):
        ws.append(r)
    
    # 2. 插入圖片到對(duì)應(yīng)位置
    for img, row, col in images:
        new_img = copy.deepcopy(img)  # 復(fù)制圖片避免引用沖突
        new_img.anchor = f"{chr(64+col)}{row}"  # 設(shè)置錨點(diǎn)（如A2、B5）
        ws.add_image(new_img)
    
    wb.save(file_path)
    wb.close()

import tkinter as tk
from tkinter import ttk, filedialog, messagebox
import pandas as pd
import os
import threading
import time
import copy
from openpyxl import load_workbook, Workbook
from openpyxl.drawing.image import Image as OpenpyxlImage
from openpyxl.utils.dataframe import dataframe_to_rows

# 全局變量存儲(chǔ)文件數(shù)據(jù)（新增圖片存儲(chǔ)字段）
file_a_data = {"sheets": [], "data": {}, "path": "", "images": {}}  # images: {sheet: [(img, row, col), ...]}
file_b_data = {"sheets": [], "data": {}, "path": "", "images": {}}
# 新增：存儲(chǔ)共有數(shù)據(jù)及對(duì)應(yīng)圖片
compare_result = {
    "a_not_b": None, "b_not_a": None, "a_common": None, "b_common": None,
    "a_common_images": [], "b_common_images": []
}
# 整理表格相關(guān)全局變量（新增圖片存儲(chǔ)）
clean_file_data = {"sheets": [], "data": {}, "path": "", "images": {}}
cleaned_table_result = None
cleaned_table_images = []  # 整理后的圖片：[(img, new_row, col), ...]

class TableCompareApp:
    def __init__(self, root):
        self.root = root
        self.root.title("表格比對(duì)&整理工具")
        self.root.geometry("950x850")
        
        # 進(jìn)度條變量
        self.progress_var = tk.DoubleVar()
        
        # 構(gòu)建UI
        self._create_widgets()
    
    def _create_widgets(self):
        # 1. 文件上傳區(qū)域（原有比對(duì)功能）
        upload_frame = ttk.LabelFrame(self.root, text="文件上傳（比對(duì)功能）")
        upload_frame.pack(padx=10, pady=10, fill="x")
        
        # 文件A上傳
        ttk.Label(upload_frame, text="文件A：").grid(row=0, column=0, padx=5, pady=5, sticky="w")
        self.file_a_label = ttk.Label(upload_frame, text="未選擇文件", foreground="gray")
        self.file_a_label.grid(row=0, column=1, padx=5, pady=5, sticky="w")
        ttk.Button(upload_frame, text="選擇文件", command=lambda: self.load_file("A")).grid(row=0, column=2, padx=5, pady=5)
        
        # 文件B上傳
        ttk.Label(upload_frame, text="文件B：").grid(row=1, column=0, padx=5, pady=5, sticky="w")
        self.file_b_label = ttk.Label(upload_frame, text="未選擇文件", foreground="gray")
        self.file_b_label.grid(row=1, column=1, padx=5, pady=5, sticky="w")
        ttk.Button(upload_frame, text="選擇文件", command=lambda: self.load_file("B")).grid(row=1, column=2, padx=5, pady=5)
        
        # 2. 文件基本信息區(qū)域（原有）
        info_frame = ttk.LabelFrame(self.root, text="文件基本信息")
        info_frame.pack(padx=10, pady=5, fill="x")
        
        self.file_a_info = ttk.Label(info_frame, text="文件A信息：無")
        self.file_a_info.pack(padx=5, pady=2, anchor="w")
        
        self.file_b_info = ttk.Label(info_frame, text="文件B信息：無")
        self.file_b_info.pack(padx=10, pady=2, anchor="w")
        
        # 3. 比對(duì)配置區(qū)域（原有）
        config_frame = ttk.LabelFrame(self.root, text="比對(duì)配置（支持不同列比對(duì)）")
        config_frame.pack(padx=10, pady=5, fill="x")
        
        # 子表格選擇
        ttk.Label(config_frame, text="文件A子表格：").grid(row=0, column=0, padx=5, pady=5, sticky="w")
        self.sheet_a_combobox = ttk.Combobox(config_frame, state="disabled")
        self.sheet_a_combobox.grid(row=0, column=1, padx=5, pady=5, sticky="w")
        
        ttk.Label(config_frame, text="文件B子表格：").grid(row=1, column=0, padx=5, pady=5, sticky="w")
        self.sheet_b_combobox = ttk.Combobox(config_frame, state="disabled")
        self.sheet_b_combobox.grid(row=1, column=1, padx=5, pady=5, sticky="w")
        
        # 比對(duì)列選擇
        ttk.Label(config_frame, text="A表格比對(duì)列：").grid(row=2, column=0, padx=5, pady=5, sticky="w")
        self.compare_col_a_combobox = ttk.Combobox(config_frame, state="disabled")
        self.compare_col_a_combobox.grid(row=2, column=1, padx=5, pady=5, sticky="w")
        
        ttk.Label(config_frame, text="B表格比對(duì)列：").grid(row=3, column=0, padx=5, pady=5, sticky="w")
        self.compare_col_b_combobox = ttk.Combobox(config_frame, state="disabled")
        self.compare_col_b_combobox.grid(row=3, column=1, padx=5, pady=5, sticky="w")
        
        # 4. 操作按鈕區(qū)域（新增：導(dǎo)出共有數(shù)據(jù)按鈕）
        compare_btn_frame = ttk.Frame(self.root)
        compare_btn_frame.pack(padx=10, pady=10)
        
        self.compare_btn = ttk.Button(compare_btn_frame, text="開始比對(duì)", command=self.start_compare, state="disabled")
        self.compare_btn.pack(side="left", padx=5)
        
        self.export_a_btn = ttk.Button(compare_btn_frame, text="導(dǎo)出A有B無的數(shù)據(jù)", command=lambda: self.export_result("a_not_b"), state="disabled")
        self.export_a_btn.pack(side="left", padx=5)
        
        self.export_b_btn = ttk.Button(compare_btn_frame, text="導(dǎo)出B有A無的數(shù)據(jù)", command=lambda: self.export_result("b_not_a"), state="disabled")
        self.export_b_btn.pack(side="left", padx=5)
        
        # 新增：導(dǎo)出共有數(shù)據(jù)按鈕
        self.export_common_btn = ttk.Button(compare_btn_frame, text="導(dǎo)出共有數(shù)據(jù)", command=self.export_common_result, state="disabled")
        self.export_common_btn.pack(side="left", padx=5)
        
        # ========== 表格整理功能區(qū)域 ==========
        clean_frame = ttk.LabelFrame(self.root, text="表格整理功能（解決一行數(shù)據(jù)占多行問題）")
        clean_frame.pack(padx=10, pady=15, fill="x")
        
        # 選擇要整理的文件
        ttk.Label(clean_frame, text="待整理文件：").grid(row=0, column=0, padx=5, pady=5, sticky="w")
        self.clean_file_label = ttk.Label(clean_frame, text="未選擇文件", foreground="gray")
        self.clean_file_label.grid(row=0, column=1, padx=5, pady=5, sticky="w")
        ttk.Button(clean_frame, text="選擇文件", command=self.load_clean_file).grid(row=0, column=2, padx=5, pady=5)
        
        # 選擇待整理的子表格
        ttk.Label(clean_frame, text="待整理子表格：").grid(row=1, column=0, padx=5, pady=5, sticky="w")
        self.clean_sheet_combobox = ttk.Combobox(clean_frame, state="disabled")
        self.clean_sheet_combobox.grid(row=1, column=1, padx=5, pady=5, sticky="w")
        
        # 選擇關(guān)鍵列（判斷新行的依據(jù)：關(guān)鍵列非空=新行）
        ttk.Label(clean_frame, text="新行判斷列：").grid(row=2, column=0, padx=5, pady=5, sticky="w")
        self.clean_key_col_combobox = ttk.Combobox(clean_frame, state="disabled")
        self.clean_key_col_combobox.grid(row=2, column=1, padx=5, pady=5, sticky="w")
        
        # 整理/導(dǎo)出按鈕
        clean_btn_frame = ttk.Frame(clean_frame)
        clean_btn_frame.grid(row=3, column=0, columnspan=3, pady=10)
        
        self.clean_btn = ttk.Button(clean_btn_frame, text="開始整理表格", command=self.start_clean_table, state="disabled")
        self.clean_btn.pack(side="left", padx=5)
        
        self.export_clean_btn = ttk.Button(clean_btn_frame, text="導(dǎo)出整理后表格", command=self.export_cleaned_table, state="disabled")
        self.export_clean_btn.pack(side="left", padx=5)
        
        # ========== 進(jìn)度條區(qū)域（共用） ==========
        progress_frame = ttk.Frame(self.root)
        progress_frame.pack(padx=10, pady=5, fill="x")
        
        self.progress_bar = ttk.Progressbar(progress_frame, variable=self.progress_var, maximum=100)
        self.progress_bar.pack(fill="x", padx=5, pady=5)
        
        self.progress_label = ttk.Label(progress_frame, text="進(jìn)度：0%")
        self.progress_label.pack(anchor="center")
    
    # ========== 核心修復(fù)：支持讀取圖片 ==========
    def load_file(self, file_type):
        """加載比對(duì)用的A/B文件（支持讀取圖片）"""
        file_path = filedialog.askopenfilename(
            title=f"選擇{file_type}文件",
            filetypes=[("Excel文件", "*.xlsx"), ("CSV文件", "*.csv"), ("所有文件", "*.*")]
        )
        if not file_path:
            return
        
        # 初始化數(shù)據(jù)
        if file_type == "A":
            file_a_data.update({"sheets": [], "data": {}, "path": file_path, "images": {}})
            self.file_a_label.config(text=os.path.basename(file_path))
            target_data = file_a_data
        else:
            file_b_data.update({"sheets": [], "data": {}, "path": file_path, "images": {}})
            self.file_b_label.config(text=os.path.basename(file_path))
            target_data = file_b_data
        
        try:
            file_ext = os.path.splitext(file_path)[1].lower()
            sheets = []
            data_dict = {}
            images_dict = {}  # 存儲(chǔ)每個(gè)sheet的圖片 (img, row, col)
            
            if file_ext == ".xlsx":
                # 用openpyxl讀取工作簿（保留圖片）
                wb = load_workbook(file_path, data_only=True)
                sheets = wb.sheetnames
                
                for sheet in sheets:
                    ws = wb[sheet]
                    # 讀取數(shù)據(jù)到DataFrame
                    df = pd.DataFrame(ws.values)
                    # 設(shè)置列名（第一行）
                    if len(df) > 0:
                        df.columns = df.iloc[0]
                        df = df.drop(0).reset_index(drop=True)
                    else:
                        df.columns = []
                    data_dict[sheet] = df
                    
                    # 提取圖片及位置（openpyxl行/列從0開始，轉(zhuǎn)換為1-based）
                    sheet_images = []
                    for img in ws._images:
                        row = img.anchor._from.row + 1  # 轉(zhuǎn)換為1-based行號(hào)
                        col = img.anchor._from.col + 1  # 轉(zhuǎn)換為1-based列號(hào)
                        sheet_images.append((img, row, col))
                    images_dict[sheet] = sheet_images
                
                wb.close()
            
            elif file_ext == ".csv":
                sheets = ["默認(rèn)表格"]
                df = pd.read_csv(file_path)
                data_dict["默認(rèn)表格"] = df
                images_dict["默認(rèn)表格"] = []  # CSV無圖片
            else:
                messagebox.showerror("錯(cuò)誤", "僅支持xlsx和csv格式文件！")
                return
            
            # 更新數(shù)據(jù)和圖片
            target_data["sheets"] = sheets
            target_data["data"] = data_dict
            target_data["images"] = images_dict
            
            # 更新信息顯示（包含圖片數(shù)）
            if file_type == "A":
                first_sheet = sheets[0]
                first_df = data_dict[first_sheet]
                img_count = len(images_dict[first_sheet])
                info_text = f"文件A信息：格式={file_ext[1:].upper()} | 子表格數(shù)={len(sheets)} | "
                info_text += f"首個(gè)表格行數(shù)={len(first_df)} | 列數(shù)={len(first_df.columns)} | 圖片數(shù)={img_count}"
                self.file_a_info.config(text=info_text)
                self.sheet_a_combobox.config(state="normal", values=sheets)
                self.sheet_a_combobox.current(0)
                self.update_compare_columns()
            else:
                first_sheet = sheets[0]
                first_df = data_dict[first_sheet]
                img_count = len(images_dict[first_sheet])
                info_text = f"文件B信息：格式={file_ext[1:].upper()} | 子表格數(shù)={len(sheets)} | "
                info_text += f"首個(gè)表格行數(shù)={len(first_df)} | 列數(shù)={len(first_df.columns)} | 圖片數(shù)={img_count}"
                self.file_b_info.config(text=info_text)
                self.sheet_b_combobox.config(state="normal", values=sheets)
                self.sheet_b_combobox.current(0)
                self.update_compare_columns()
            
            self.check_compare_btn_state()
        
        except Exception as e:
            messagebox.showerror("讀取失敗", f"文件{file_type}讀取錯(cuò)誤：{str(e)}")
    
    def load_clean_file(self):
        """加載待整理的表格文件（支持讀取圖片）"""
        file_path = filedialog.askopenfilename(
            title="選擇待整理的表格文件",
            filetypes=[("Excel文件", "*.xlsx"), ("CSV文件", "*.csv"), ("所有文件", "*.*")]
        )
        if not file_path:
            return
        
        # 清空原有整理文件數(shù)據(jù)
        clean_file_data.update({"sheets": [], "data": {}, "path": file_path, "images": {}})
        self.clean_file_label.config(text=os.path.basename(file_path))
        
        try:
            file_ext = os.path.splitext(file_path)[1].lower()
            sheets = []
            data_dict = {}
            images_dict = {}
            
            if file_ext == ".xlsx":
                # 用openpyxl讀取工作簿（保留圖片）
                wb = load_workbook(file_path, data_only=True)
                sheets = wb.sheetnames
                
                for sheet in sheets:
                    ws = wb[sheet]
                    # 讀取數(shù)據(jù)到DataFrame
                    df = pd.DataFrame(ws.values)
                    # 設(shè)置列名（第一行）
                    if len(df) > 0:
                        df.columns = df.iloc[0]
                        df = df.drop(0).reset_index(drop=True)
                    else:
                        df.columns = []
                    data_dict[sheet] = df
                    
                    # 提取圖片及位置
                    sheet_images = []
                    for img in ws._images:
                        row = img.anchor._from.row + 1
                        col = img.anchor._from.col + 1
                        sheet_images.append((img, row, col))
                    images_dict[sheet] = sheet_images
                
                wb.close()
            
            elif file_ext == ".csv":
                sheets = ["默認(rèn)表格"]
                df = pd.read_csv(file_path)
                data_dict["默認(rèn)表格"] = df
                images_dict["默認(rèn)表格"] = []
            else:
                messagebox.showerror("錯(cuò)誤", "僅支持xlsx和csv格式文件！")
                return
            
            # 存儲(chǔ)數(shù)據(jù)和圖片
            clean_file_data["sheets"] = sheets
            clean_file_data["data"] = data_dict
            clean_file_data["images"] = images_dict
            
            # 更新子表格下拉框
            self.clean_sheet_combobox.config(state="normal", values=sheets)
            self.clean_sheet_combobox.current(0)
            
            # 更新關(guān)鍵列下拉框（新行判斷列）
            self.update_clean_key_columns()
            
            # 啟用整理按鈕
            self.clean_btn.config(state="normal")
            
            # 顯示文件信息（包含圖片數(shù)）
            first_sheet = sheets[0]
            first_df = data_dict[first_sheet]
            img_count = len(images_dict[first_sheet])
            info_text = f"待整理文件信息：格式={file_ext[1:].upper()} | 子表格數(shù)={len(sheets)} | "
            info_text += f"首個(gè)表格行數(shù)={len(first_df)} | 列數(shù)={len(first_df.columns)} | 圖片數(shù)={img_count}"
            messagebox.showinfo("文件加載成功", info_text)
        
        except Exception as e:
            messagebox.showerror("讀取失敗", f"待整理文件讀取錯(cuò)誤：{str(e)}")
    
    # ========== 通用導(dǎo)出函數(shù)（核心修復(fù)：支持導(dǎo)出圖片） ==========
    def export_with_images(self, df, images, title):
        """通用導(dǎo)出函數(shù)（支持圖片保留）"""
        # 選擇導(dǎo)出路徑
        file_path = filedialog.asksaveasfilename(
            title=title,
            filetypes=[("Excel文件", "*.xlsx"), ("CSV文件", "*.csv")],
            defaultextension=".xlsx"
        )
        if not file_path:
            return
        
        file_ext = os.path.splitext(file_path)[1].lower()
        
        try:
            if file_ext == ".xlsx":
                # 創(chuàng)建新工作簿
                wb = Workbook()
                ws = wb.active
                ws.title = "數(shù)據(jù)"
                
                # 寫入DataFrame數(shù)據(jù)（保留表頭）
                for r in dataframe_to_rows(df, index=False, header=True):
                    ws.append(r)
                
                # 插入圖片到對(duì)應(yīng)位置
                for img, row, col in images:
                    # 復(fù)制圖片對(duì)象避免引用沖突
                    new_img = copy.deepcopy(img)
                    # 設(shè)置圖片錨點(diǎn)（列字母+行號(hào)）
                    new_img.anchor = f"{chr(64+col)}{row}"
                    ws.add_image(new_img)
                
                # 保存工作簿
                wb.save(file_path)
                wb.close()
                messagebox.showinfo("導(dǎo)出成功", f"數(shù)據(jù)已導(dǎo)出至：{file_path}\n包含圖片：{len(images)}張")
            
            elif file_ext == ".csv":
                # CSV不支持圖片，提示用戶
                if images:
                    messagebox.showwarning("格式提示", "CSV格式不支持存儲(chǔ)圖片，圖片將丟失！")
                df.to_csv(file_path, index=False, encoding="utf-8-sig")
                messagebox.showinfo("導(dǎo)出成功", f"數(shù)據(jù)已導(dǎo)出至：{file_path}\n注意：CSV格式不包含圖片")
        
        except Exception as e:
            messagebox.showerror("導(dǎo)出失敗", f"導(dǎo)出錯(cuò)誤：{str(e)}")
    
    # ========== 比對(duì)功能（修復(fù)：處理圖片） ==========
    def update_compare_columns(self):
        """更新比對(duì)列下拉框"""
        if file_a_data["sheets"] and self.sheet_a_combobox.get():
            sheet_a = self.sheet_a_combobox.get()
            cols_a = file_a_data["data"][sheet_a].columns.tolist()
            self.compare_col_a_combobox.config(state="normal", values=cols_a)
            self.compare_col_a_combobox.current(0)
        
        if file_b_data["sheets"] and self.sheet_b_combobox.get():
            sheet_b = self.sheet_b_combobox.get()
            cols_b = file_b_data["data"][sheet_b].columns.tolist()
            self.compare_col_b_combobox.config(state="normal", values=cols_b)
            self.compare_col_b_combobox.current(0)
    
    def check_compare_btn_state(self):
        """檢查比對(duì)按鈕狀態(tài)"""
        if (file_a_data["path"] and file_b_data["path"] and 
            self.compare_col_a_combobox.get() and self.compare_col_b_combobox.get()):
            self.compare_btn.config(state="normal")
        else:
            self.compare_btn.config(state="disabled")
    
    def update_progress(self, value):
        """更新進(jìn)度條"""
        self.progress_var.set(value)
        self.progress_label.config(text=f"進(jìn)度：{int(value)}%")
        self.root.update_idletasks()
    
    def compare_task(self):
        """比對(duì)任務(wù)（修復(fù)：處理共有數(shù)據(jù)圖片）"""
        try:
            # 1. 獲取配置信息
            sheet_a = self.sheet_a_combobox.get()
            sheet_b = self.sheet_b_combobox.get()
            col_a = self.compare_col_a_combobox.get()
            col_b = self.compare_col_b_combobox.get()
            
            # 2. 獲取數(shù)據(jù)和圖片
            df_a = file_a_data["data"][sheet_a].copy()
            df_b = file_b_data["data"][sheet_b].copy()
            img_a_list = file_a_data["images"].get(sheet_a, [])
            img_b_list = file_b_data["images"].get(sheet_b, [])
            
            # 3. 進(jìn)度條初始化
            self.update_progress(0)
            time.sleep(0.1)
            
            # 4. 數(shù)據(jù)預(yù)處理
            self.update_progress(20)
            df_a[col_a] = df_a[col_a].astype(str).fillna("")
            df_b[col_b] = df_b[col_b].astype(str).fillna("")
            
            # 5. 核心比對(duì)邏輯
            self.update_progress(50)
            a_vals = set(df_a[col_a].unique())
            b_vals = set(df_b[col_b].unique())
            
            # A有B無的行
            a_not_b_vals = a_vals - b_vals
            a_not_b_df = df_a[df_a[col_a].isin(a_not_b_vals)].reset_index(drop=True)
            
            # B有A無的行
            b_not_a_vals = b_vals - a_vals
            b_not_a_df = df_b[df_b[col_b].isin(b_not_a_vals)].reset_index(drop=True)
            
            # 共有數(shù)據(jù)
            common_vals = a_vals & b_vals
            a_common_df = df_a[df_a[col_a].isin(common_vals)].reset_index(drop=True)
            b_common_df = df_b[df_b[col_b].isin(common_vals)].reset_index(drop=True)
            
            # 處理共有數(shù)據(jù)的圖片（映射行號(hào)）
            # A格式共有數(shù)據(jù)圖片
            a_common_original_rows = df_a[df_a[col_a].isin(common_vals)].index.tolist()
            a_common_images = []
            for img, orig_row, col in img_a_list:
                orig_df_row = orig_row - 2  # Excel行1=表頭，行2=df第0行
                if orig_df_row in a_common_original_rows:
                    new_row = a_common_original_rows.index(orig_df_row) + 2  # 新表行2開始是數(shù)據(jù)
                    a_common_images.append((img, new_row, col))
            
            # B格式共有數(shù)據(jù)圖片
            b_common_original_rows = df_b[df_b[col_b].isin(common_vals)].index.tolist()
            b_common_images = []
            for img, orig_row, col in img_b_list:
                orig_df_row = orig_row - 2
                if orig_df_row in b_common_original_rows:
                    new_row = b_common_original_rows.index(orig_df_row) + 2
                    b_common_images.append((img, new_row, col))
            
            # 6. 存儲(chǔ)結(jié)果
            compare_result["a_not_b"] = a_not_b_df
            compare_result["b_not_a"] = b_not_a_df
            compare_result["a_common"] = a_common_df
            compare_result["b_common"] = b_common_df
            compare_result["a_common_images"] = a_common_images
            compare_result["b_common_images"] = b_common_images
            
            # 7. 完成進(jìn)度
            self.update_progress(100)
            
            # 8. 啟用導(dǎo)出按鈕
            self.export_a_btn.config(state="normal")
            self.export_b_btn.config(state="normal")
            self.export_common_btn.config(state="normal")
            
            # 顯示結(jié)果（包含圖片數(shù)）
            messagebox.showinfo("比對(duì)完成", 
                                f"比對(duì)結(jié)果：\nA有B無的數(shù)據(jù)行數(shù)：{len(a_not_b_df)}\n"
                                f"B有A無的數(shù)據(jù)行數(shù)：{len(b_not_a_df)}\n"
                                f"共有數(shù)據(jù)行數(shù)（A格式）：{len(a_common_df)} | 圖片數(shù)：{len(a_common_images)}\n"
                                f"共有數(shù)據(jù)行數(shù)（B格式）：{len(b_common_df)} | 圖片數(shù)：{len(b_common_images)}")
        
        except Exception as e:
            messagebox.showerror("比對(duì)失敗", f"比對(duì)過程出錯(cuò)：{str(e)}")
            self.update_progress(0)
    
    def start_compare(self):
        """啟動(dòng)比對(duì)"""
        self.compare_btn.config(state="disabled")
        self.export_a_btn.config(state="disabled")
        self.export_b_btn.config(state="disabled")
        self.export_common_btn.config(state="disabled")
        
        compare_thread = threading.Thread(target=self.compare_task)
        compare_thread.daemon = True
        compare_thread.start()
    
    def export_result(self, result_type):
        """導(dǎo)出比對(duì)結(jié)果（修復(fù)：支持圖片）"""
        if compare_result[result_type] is None:
            messagebox.showwarning("提示", "暫無可導(dǎo)出的數(shù)據(jù)！")
            return
        
        # 篩選對(duì)應(yīng)圖片
        if result_type == "a_not_b":
            df = compare_result["a_not_b"]
            sheet_a = self.sheet_a_combobox.get()
            img_a_list = file_a_data["images"].get(sheet_a, [])
            col_a = self.compare_col_a_combobox.get()
            a_not_b_vals = set(df[col_a].unique())
            a_not_b_original_rows = file_a_data["data"][sheet_a][file_a_data["data"][sheet_a][col_a].isin(a_not_b_vals)].index.tolist()
            images = []
            for img, orig_row, col in img_a_list:
                orig_df_row = orig_row - 2
                if orig_df_row in a_not_b_original_rows:
                    new_row = a_not_b_original_rows.index(orig_df_row) + 2
                    images.append((img, new_row, col))
            title = "保存A有B無的數(shù)據(jù)"
        elif result_type == "b_not_a":
            df = compare_result["b_not_a"]
            sheet_b = self.sheet_b_combobox.get()
            img_b_list = file_b_data["images"].get(sheet_b, [])
            col_b = self.compare_col_b_combobox.get()
            b_not_a_vals = set(df[col_b].unique())
            b_not_a_original_rows = file_b_data["data"][sheet_b][file_b_data["data"][sheet_b][col_b].isin(b_not_a_vals)].index.tolist()
            images = []
            for img, orig_row, col in img_b_list:
                orig_df_row = orig_row - 2
                if orig_df_row in b_not_a_original_rows:
                    new_row = b_not_a_original_rows.index(orig_df_row) + 2
                    images.append((img, new_row, col))
            title = "保存B有A無的數(shù)據(jù)"
        else:
            return
        
        self.export_with_images(df, images, title)
    
    def export_common_result(self):
        """導(dǎo)出共有數(shù)據(jù)（修復(fù)：支持圖片）"""
        if compare_result["a_common"] is None or compare_result["b_common"] is None:
            messagebox.showwarning("提示", "暫無共有數(shù)據(jù)可導(dǎo)出！")
            return
        
        # 選擇導(dǎo)出格式
        choice = messagebox.askquestion(
            "選擇導(dǎo)出格式",
            "請(qǐng)選擇共有數(shù)據(jù)的導(dǎo)出格式：\n【是】= 按A表格格式導(dǎo)出（保留A表格列結(jié)構(gòu)）\n【否】= 按B表格格式導(dǎo)出（保留B表格列結(jié)構(gòu)）",
            icon="question"
        )
        
        if choice == "yes":
            df = compare_result["a_common"]
            images = compare_result["a_common_images"]
            title = "保存按A表格格式的共有數(shù)據(jù)"
        elif choice == "no":
            df = compare_result["b_common"]
            images = compare_result["b_common_images"]
            title = "保存按B表格格式的共有數(shù)據(jù)"
        else:
            return
        
        self.export_with_images(df, images, title)
    
    # ========== 表格整理功能（修復(fù)：處理圖片） ==========
    def update_clean_key_columns(self):
        """更新整理功能的關(guān)鍵列下拉框"""
        if clean_file_data["sheets"] and self.clean_sheet_combobox.get():
            sheet = self.clean_sheet_combobox.get()
            cols = clean_file_data["data"][sheet].columns.tolist()
            self.clean_key_col_combobox.config(state="normal", values=cols)
            self.clean_key_col_combobox.current(0)
    
    def clean_table_task(self):
        """表格整理任務(wù)（修復(fù)：合并數(shù)據(jù)+保留圖片）"""
        global cleaned_table_result, cleaned_table_images
        try:
            # 獲取配置
            sheet = self.clean_sheet_combobox.get()
            key_col = self.clean_key_col_combobox.get()
            df = clean_file_data["data"][sheet].copy()
            orig_images = clean_file_data["images"].get(sheet, [])
            
            # 進(jìn)度條初始化
            self.update_progress(0)
            time.sleep(0.1)
            
            # 預(yù)處理
            self.update_progress(10)
            df = df.fillna("")
            
            # 核心整理邏輯
            self.update_progress(20)
            cleaned_rows = []
            current_row = None
            orig_to_new_row = {}  # 原df行號(hào) → 新df行號(hào)映射
            new_row_idx = 0
            
            total_rows = len(df)
            for idx, row in df.iterrows():
                # 更新進(jìn)度
                progress = 20 + (idx / total_rows) * 70
                self.update_progress(progress)
                
                # 關(guān)鍵列非空 = 新行開始
                if str(row[key_col]).strip() != "":
                    if current_row is not None:
                        cleaned_rows.append(current_row)
                        new_row_idx += 1
                    current_row = row.to_dict()
                    orig_to_new_row[idx] = new_row_idx
                else:
                    # 補(bǔ)充到當(dāng)前行
                    if current_row is not None:
                        for col in df.columns:
                            if str(row[col]).strip() != "" and str(current_row[col]).strip() == "":
                                current_row[col] = row[col]
                        orig_to_new_row[idx] = new_row_idx
            
            # 保存最后一行
            if current_row is not None:
                cleaned_rows.append(current_row)
                new_row_idx += 1
            
            # 轉(zhuǎn)換為DataFrame
            self.update_progress(90)
            cleaned_table_result = pd.DataFrame(cleaned_rows)
            
            # 處理圖片（映射到新行號(hào)）
            cleaned_table_images = []
            for img, orig_row, col in orig_images:
                orig_df_row = orig_row - 2  # Excel行 → df行號(hào)
                if orig_df_row in orig_to_new_row:
                    # 新Excel行號(hào) = 新df行號(hào) + 2（行1是表頭）
                    new_excel_row = orig_to_new_row[orig_df_row] + 2
                    cleaned_table_images.append((img, new_excel_row, col))
            
            # 完成進(jìn)度
            self.update_progress(100)
            
            # 啟用導(dǎo)出按鈕
            self.export_clean_btn.config(state="normal")
            
            messagebox.showinfo("整理完成", 
                                f"表格整理成功！\n原行數(shù)：{len(df)} | 整理后行數(shù)：{len(cleaned_table_result)}\n"
                                f"原圖片數(shù)：{len(orig_images)} | 整理后圖片數(shù)：{len(cleaned_table_images)}")
        
        except Exception as e:
            messagebox.showerror("整理失敗", f"表格整理出錯(cuò)：{str(e)}")
            self.update_progress(0)
    
    def start_clean_table(self):
        """啟動(dòng)表格整理"""
        if not self.clean_key_col_combobox.get():
            messagebox.showwarning("提示", "請(qǐng)選擇新行判斷列！")
            return
        
        self.clean_btn.config(state="disabled")
        self.export_clean_btn.config(state="disabled")
        
        clean_thread = threading.Thread(target=self.clean_table_task)
        clean_thread.daemon = True
        clean_thread.start()
    
    def export_cleaned_table(self):
        """導(dǎo)出整理后的表格（修復(fù)：保留圖片）"""
        global cleaned_table_result, cleaned_table_images
        if cleaned_table_result is None:
            messagebox.showwarning("提示", "暫無整理后的數(shù)據(jù)！")
            return
        
        self.export_with_images(cleaned_table_result, cleaned_table_images, "保存整理后的表格")

if __name__ == "__main__":
    root = tk.Tk()
    # 解決tkinter中文亂碼（Windows）
    try:
        root.option_add("*Font", "SimHei 9")
    except:
        pass
    
    app = TableCompareApp(root)
    root.mainloop()