使用 os.walk 后如何使用 Pandas 读取已知文件名的特定文件？

Question

I am using os.walk() to get all the files in a certain directory; however, there are a ton of files in that directory that I don't need. I know the specific names of the files that I want to read as they update everyday and only the date of the filename changes.

import pandas as pd
import os
from os import listdir, walk
from os.path import isfile, join

td_str = pd.to_datetime('today').strftime('%Y%m%d') # Returns '20200923'
path = 'C:\\Users\\myuser\\subdirectory\\' + td_str
for root, directories, files in os.walk(path, topdown=False):
    for name in files:
        print(os.path.join(root, name))


### The Output
# C:\Users\myuser\subdirectory\20200923\20200923_file_a.csv
# C:\Users\myuser\subdirectory\20200923\20200923_file_b.csv 
# C:\Users\myuser\subdirectory\20200923\20200923_file_c.csv
# C:\Users\myuser\subdirectory\20200923\20200923_file_d.csv

I know I want to read file_b and file_c and put them into respective dataframes.

df_file_b = pd.read_csv('C:\Users\myuser\subdirectory\20200923\20200923_file_b.csv')
df_file_c = pd.read_csv('C:\Users\myuser\subdirectory\20200923\20200923_file_c.csv')

How can I read those specific files only with pandas? I currently assume I might have to do something like have the desired filenames in a list and do an if loop to check if os.walk() finds it, but is there a more efficient way to do this?

Thanks much.

Answer 1

You can make use of the glob module, supporting patterns: 10.7. glob — Unix style pathname pattern expansion .

A simple solution would be to create a filter with file_ and the character range b to c :

import glob
print(glob.glob('*file_[b-c]*'))

This will print

['20200923_file_b.csv', '20200923_file_c.csv']