python 统计磁盘大小到excel

fengjicheng

scan_disk_size

[Python] 纯文本查看 复制代码

import os
import math
import pandas as pd
import datetime

# -*- coding: utf-8 -*-
# @Time    :  2024/05/22
# [url=home.php?mod=space&uid=686208]@AuThor[/url]  :  fengjicheng
# [url=home.php?mod=space&uid=267492]@file[/url]    :  scan_disk_size.py
# @Software:  根据层级统计文件夹


def get_drives():
    """获取Windows系统中的所有盘符"""
    drives = []
    for letter in 'F':
        if os.path.exists(letter + ':\\'):
            drives.append(letter + ':\\')
    return drives



def dfs_calculate_dir_size_and_export_to_excel(dir_path='.', current_depth=0, max_depth=4, excel_writer=None, sheet_name=None):
    """递归计算指定目录的总大小，并将结果输出到单个Excel文件的不同工作表中。控制搜索深度。
    
    :param dir_path: 要计算大小的目录路径，默认为当前目录
    :param current_depth: 当前递归深度，默认为0
    :param max_depth: 最大递归深度，默认为4
    :param excel_writer: ExcelWriter对象，用于写入数据
    :param sheet_name: Excel的sheet名称
    :return: None
    """
    try:
        #  如果到达最大深度，则计算目录和文件的大小
        for entry in os.scandir(dir_path):
            #  如果为回收站 和系统模块则跳过
            if  entry.name  in ['System Volume Information', '$Recycle.Bin','$RECYCLE.BIN','$360Honeypot']:
                continue
            row_data = {
                    '盘符': dir_path.split(':')[0],
                    '是否为文件夹':'是' if not  entry.is_file() else '否',
                    '文件路径': entry.path,
                    '文件大小': getFileFolderSize(entry.path) if not  entry.is_file() else entry.stat().st_size,
                    '友好的文件大小': '',
                    '搜索层级': current_depth
                    }
            row_data['友好的文件大小'] = convert_size(row_data['文件大小'])          
            df = pd.DataFrame([row_data])
            if excel_writer is not None and sheet_name is not None:
                if sheet_name not in excel_writer.sheets:
                    df.to_excel(excel_writer, sheet_name=sheet_name, index=False, header=True)
                else:
                    df.to_excel(excel_writer, sheet_name=sheet_name, index=False, header=False, startrow=excel_writer.sheets[sheet_name].max_row)
            if entry.is_dir() and  current_depth < max_depth:
                dfs_calculate_dir_size_and_export_to_excel(entry.path, current_depth + 1, max_depth, excel_writer, sheet_name)
    except PermissionError:
        print(f"Permission denied for accessing: {dir_path}. Skipping this directory.")
    except FileNotFoundError:
        print(f"Directory not found: {dir_path}. Continuing to the next one.")
    except Exception as e:
        print(f"An unexpected error occurred: {e}. Skipping this directory.")
                    
def convert_size(size_bytes):
    """将字节转换为更易读的格式（KB, MB, GB）"""
    if size_bytes == 0:
        return "0B"
    size_name = ("B", "KB", "MB", "GB", "TB", "PB", "EB", "ZB", "YB")
    i = int(math.floor(math.log(size_bytes, 1024)))
    p = math.pow(1024, i)
    s = round(size_bytes / p, 2)
    return "%s %s" % (s, size_name[i])

def getFileFolderSize(fileOrFolderPath):
  """获得目录大小"""
  totalSize = 0
 
  if not os.path.exists(fileOrFolderPath):
    return totalSize
  
  if os.path.isfile(fileOrFolderPath):
    totalSize = os.path.getsize(fileOrFolderPath) # 5041481
    return totalSize
 
  if os.path.isdir(fileOrFolderPath):
    with os.scandir(fileOrFolderPath) as dirEntryList:
      for curSubEntry in dirEntryList:
        curSubEntryFullPath = os.path.join(fileOrFolderPath, curSubEntry.name)
        if curSubEntry.is_dir():
          curSubFolderSize = getFileFolderSize(curSubEntryFullPath) # 5800007
          totalSize += curSubFolderSize
        elif curSubEntry.is_file():
          curSubFileSize = os.path.getsize(curSubEntryFullPath) # 1891
          totalSize += curSubFileSize
      return totalSize


def main():
    drives = get_drives()
    # 获取当前时间
    now = datetime.datetime.now()
    # 输出文件名
    output_file = now.strftime("%Y%m%d%H%M%S") + "_search_results.xlsx"
    # 需要提前 pip install  openpyxl
    with pd.ExcelWriter(output_file, engine='openpyxl') as writer:
        # 遍历盘符
        for drive in drives:
            dfs_calculate_dir_size_and_export_to_excel(drive,current_depth=0,max_depth=3, excel_writer=writer, sheet_name='Search Results')
    print("搜索完成，结果已输出到各驱动器对应的Excel文件中。")
    

if __name__ == "__main__":
    main()

扫描文件夹，并把文件夹大小保存到excel中

koogg

加油，再加上其他信息统计收集，最后可以做成一个局域网内电脑配置信息一键收集并上传指定位置的工具了，这样以后统计电脑信息就方便多了

10m2go

感谢，这个功能可以用脚本实现不

lvtaode0657

再加上点儿其他统计信息，就功能更好了。

cxrm

很不错的功能，棒

huaxiaotian

假如有上千万的文件，这个会卡死吧

fengjicheng

huaxiaotian 发表于 2024-10-2 12:22
假如有上千万的文件，这个会卡死吧

会有可能，不过可以改造，

xiaolinge566

加油楼主，期待功能更加完善