How to check existence of a folder with python and then remove it?

Question

I want to remove dataset folder from dataset3 folder. But the following code is not removing dataset. First I want to check if dataset already exist in dataset then remove dataset.
Can some one please point out my mistake in following code?

for files in os.listdir("dataset3"):
    if os.path.exists("dataset"):
        os.system("rm -rf "+"dataset")

Reynaldoreynard

@HFBrowning - that's not too effective against directories. – Podagra

You don't use the files filename and even if you did, you need to add the original path ('dataset3') to it. – Podagra

Then how I can do that – Reynaldoreynard

By dataset folder, its literally named "dataset"? – Podagra

If you know the name of the directory, there is no need for any checks. os.system("rm -rf dataset3/dataset") does the job. – Podagra

you are looping over all files under "dataset3" and for each file in "dataset3" you are deleting "dataset" - "dataset" has to be on the same directory-level as "dataset3"!? Does this make sense? – Gorget

In my actual code when i run my programe it generates a dataset folder. I shif it in dataset3 folder.But when next time I run the same programe using same parameter I'm unable to move dataset into dataset3 because it already exists there. That's why I want to check its existance that if it already exists then remove it so that I can shift a new dataset into dataset3 – Reynaldoreynard

Answer 1

Python's os.rmdir() only works on empty the directories, however shutil.rmtree() doesn't care (even if there are subdirectories) which makes it very similar to the Linux rm -rf command.

import os
import shutil

dirpath = os.path.join('dataset3', 'dataset')
if os.path.exists(dirpath) and os.path.isdir(dirpath):
    shutil.rmtree(dirpath)

Modern approach

In Python 3.4+ you can do same thing using the pathlib module to make the code more object-oriented and readable:

from pathlib import Path
import shutil

dirpath = Path('dataset3') / 'dataset'
if dirpath.exists() and dirpath.is_dir():
    shutil.rmtree(dirpath)

Ostiary

Is Path a relative path? I prefer absolute paths as is common sense. – Pohl

@Timo: In this case its value happens to be relative. A Path is a subclass of the PurePath class that's defined in the pathlib module and are not absolute nor relative per se — they can be either. These classes make accessing components of path easy to access and allows operations on them to be done in an object-oriented fashion. Also note that sometime using a relative path makes perfect sense — such as when you don't want to hardcode an absolute path into your code or won't know what the root folder is until runtime. – Ostiary

in both ancient and modern code snippets, isn't exists() and is_dir() the same as just is_dir() ? – Connolly

Answer 2

Better to set ignore_errors:

import shutil

shutil.rmtree('/folder_name', ignore_errors=True)

This is much more readable, and concise.

Note that it will ignore all errors, not just dir missing errors.

Crossover

Yeah, so better not to do this, really :) - Otherwise you won't know if it failed to clear the folder. – Developer

Answer 3

os.remove() is to remove a file.

os.rmdir() is to remove an empty directory.

shutil.rmtree() is to delete a directory and all its contents.

import os

folder = "dataset3/"

# Method 1
for files in os.listdir(folder):
    if files == "dataset":
        os.remove(folder + "dataset")

# Method 2
if os.path.exists(folder + "dataset"):
    os.remove(folder + "dataset")

Bigod

Answer 4

try this:

for files in os.listdir("dataset3"):
  if files=="dataset":
    fn=os.path.join("dataset3", files)
    os.system("rm -rf "+fn)
    break

You do not need the os.path.exists() because os.listdir() already told you, that it exists.

And if your foldernames are static, you can do it with:

if os.path.exists("dataset3/dataset"):
  os.system("rm -rf dataset3/dataset")

or as:

try:
  os.system("rm -rf dataset3/dataset")
except:
  pass

Gorget

Answer 5

This will do it:

for files in os.listdir('dataset3'):
     if files == 'dataset':
         os.rmdir(os.path.join(os.getcwd() + 'dataset3', files))

Rios

For os.path.join(), you don't need/want to include the leading slash in the '/dataset3'. – Ostiary

@Ostiary fixed! – Rios