文件信息整理demo

import os
import re
from subprocess import getoutput
import numpy as np
import pandas as pd
from tqdm import tqdm
import time


def filesize_float_function(filepath):
    '''
    获得文件大小
    Args:
        filepath:

    Returns:

    '''
    return os.path.getsize(filepath) / float(1024 ** 2)

def filesize_to_MB_function(filepath):
    '''
    获得文件大小的MB单位
    Args:
        filepath:

    Returns:

    '''
    return "%.10f MB" % (os.path.getsize(filepath) / float(1024 ** 2))

def filesize_to_GB_function(filepath):
    '''
    获得文件大小的GB单位
    Args:
        filepath:

    Returns:

    '''
    return "%.10f GB" % (os.path.getsize(filepath) / float(1024 ** 3))

def filesize_to_TB_function(filepath):
    '''
    获得文件大小的TB单位
    Args:
        filepath:
    Returns:
    '''
    return "%.10f TB" % (os.path.getsize(filepath) / float(1024 ** 4))

def get_filename_function(filepath):
    '''
    从路径中获得文件名
    Args:
        filepath:

    Returns:
    '''
    return os.path.splitext(filepath)[-1]

def get_filetype_function(filepath
                               , types=['.py', 'txt', 'log', 'csv'
                                        ,'docx','doc'
                                        , 'dp' , "pdf",'ipynb',"jpeg"
                                        ,'raw',"jpg" ,'bmp','jpg'
                                        ,'svg','png','tif','gif'
                                        ,'pcx','tga','exif','fpx'
                                        ,'psd','cdr','pcd','dxf'
                                        ,'ufo','eps','ai','WMF'
                                        ,'webp']):
    '''
    从文件名中获得疑似的文件类型
    Args:
        filename: 文件的绝对路径地址
        types: 选择的文件类型

    Returns:

    '''
    file_type =  os.path.splitext(filepath)[-1].split('.')[-1]
    if file_type in types:
        return file_type
    else:
        return False
    
def get_time_stamp_to_time_function(timestamp):
    """
    把时间戳转化为时间
    """
    time_struct = time.localtime(timestamp)
    return time.strftime('%Y-%m-%d %H:%M:%S',time_struct)
    
def get_access_time_function(filepath):
    """
    获取访问时间
    """
    return os.path.getatime(filepath)

def get_create_time_function(filepath):
    """
    获取创建时间
    """
    return os.path.getctime(filepath)

def get_modify_time_function(filepath):
    """
    获取修改时间
    """
    return os.path.getmtime(filepath)

def floa_time_function(strptime):
    """
    时间戳转浮点数
    """
    return time.mktime(time.strptime(strptime,'%Y-%m-%d %H:%M:%S'))
    


def filter_file_paths(dir_path=os.path.expanduser("~/")):
    counter = 0
    result = {"文件地址":[],"文件类型":[],"创建时间":[]
              ,"访问时间":[],"修改时间":[],"文件大小浮点数":[]
              ,"文件大小mb":[],"创建时间戳":[],"访问时间戳":[]
              ,"修改时间戳":[]}
    for dir_, folder, files in tqdm(os.walk(dir_path)):
        for file in files:
            counter += 1
            temp_path = os.path.join(dir_, file)
            check_file = os.path.isfile(temp_path)
            if check_file:
                result["文件地址"].append(temp_path)
                result["文件类型"].append(get_filetype_function(temp_path))
                result["创建时间"].append(get_create_time_function(temp_path))
                result["访问时间"].append(get_access_time_function(temp_path))
                result["修改时间"].append(get_modify_time_function(temp_path))
                result["创建时间戳"].append(get_time_stamp_to_time_function(get_create_time_function(temp_path)))
                result["访问时间戳"].append(get_time_stamp_to_time_function(get_access_time_function(temp_path)))
                result["修改时间戳"].append(get_time_stamp_to_time_function(get_modify_time_function(temp_path)))
                result["文件大小浮点数"].append(filesize_float_function(temp_path))
                result["文件大小mb"].append(filesize_to_MB_function(temp_path))
                
    return pd.DataFrame(result)
            
filter_file_paths(os.path.expanduser(os.getcwd()))