createmodel-doc/search_excel.py

import os
import pandas as pd

def find_excel_files(directory, search_str):
    search_str_lower = search_str.lower()
    matched_files = []
    
    for root, _, files in os.walk(directory):
        for file in files:
            if file.lower().endswith(('.xls', '.xlsx')):
                file_path = os.path.join(root, file)
                try:
                    xls = pd.ExcelFile(file_path)
                    found = False
                    for sheet_name in xls.sheet_names:
                        df = pd.read_excel(xls, sheet_name=sheet_name)
                        
                        # 检查列名
                        cols_lower = [str(col).lower() for col in df.columns]
                        if any(search_str_lower in col for col in cols_lower):
                            matched_files.append(file_path)
                            found = True
                            break
                        
                        # 检查数据单元格
                        mask = df.astype(str).apply(
                            lambda col: col.str.contains(search_str, case=False, regex=False, na=False)
                        )
                        if mask.any().any():
                            matched_files.append(file_path)
                            found = True
                            break
                    if found:
                        break  # 跳到下一个文件
                except Exception as e:
                    print(f"错误处理文件 {file_path}: {e}")
    return matched_files

if __name__ == "__main__":
    # 用户输入
    target_dir = input("请输入要搜索的目录路径：")
    search_text = input("请输入要查找的字符串：")
    
    # 查找文件
    results = find_excel_files(target_dir, search_text)
    
    # 输出结果
    print("\n找到以下包含指定字符串的文件：")
    for file in results:
        print(file)