[python] How do I remove/delete a folder that is not empty?

I am getting an 'access is denied' error when I attempt to delete a folder that is not empty. I used the following command in my attempt: os.remove("/folder_name").

What is the most effective way of removing/deleting a folder/directory that is not empty?

For Windows, if directory is not empty, and you have read-only files or you get errors like

• Access is denied
• The process cannot access the file because it is being used by another process

Try this, os.system('rmdir /S /Q "{}"'.format(directory))

It's equivalent for rm -rf in Linux/Mac.

import shutil
shutil.rmtree(dest, ignore_errors=True)

Base on kkubasik's answer, check if folder exists before remove, more robust

import shutil
def remove_folder(path):
    # check if folder exists
    if os.path.exists(path):
         # remove if exists
         shutil.rmtree(path)
    else:
         # throw your exception to handle this special scenario
         raise XXError("your exception") 
remove_folder("/folder_name")

From the python docs on os.walk():

# Delete everything reachable from the directory named in 'top',
# assuming there are no symbolic links.
# CAUTION:  This is dangerous!  For example, if top == '/', it
# could delete all your disk files.
import os
for root, dirs, files in os.walk(top, topdown=False):
    for name in files:
        os.remove(os.path.join(root, name))
    for name in dirs:
        os.rmdir(os.path.join(root, name))

If you don't want to use the shutil module you can just use the os module.

from os import listdir, rmdir, remove
for i in listdir(directoryToRemove):
    os.remove(os.path.join(directoryToRemove, i))
rmdir(directoryToRemove) # Now the directory is empty of files

import shutil
shutil.rmtree(dest, ignore_errors=True)

With os.walk I would propose the solution which consists of 3 one-liner Python calls:

python -c "import sys; import os; [os.chmod(os.path.join(rs,d), 0o777) for rs,ds,fs in os.walk(_path_) for d in ds]"
python -c "import sys; import os; [os.chmod(os.path.join(rs,f), 0o777) for rs,ds,fs in os.walk(_path_) for f in fs]"
python -c "import os; import shutil; shutil.rmtree(_path_, ignore_errors=False)"

The first script chmod's all sub-directories, the second script chmod's all files. Then the third script removes everything with no impediments.

I have tested this from the "Shell Script" in a Jenkins job (I did not want to store a new Python script into SCM, that's why searched for a one-line solution) and it worked for Linux and Windows.

Just some python 3.5 options to complete the answers above. (I would have loved to find them here).

import os
import shutil
from send2trash import send2trash # (shutil delete permanently)

Delete folder if empty

root = r"C:\Users\Me\Desktop\test"   
for dir, subdirs, files in os.walk(root):   
    if subdirs == [] and files == []:
           send2trash(dir)
           print(dir, ": folder removed")

Delete also folder if it contains this file

    elif subdirs == [] and len(files) == 1: # if contains no sub folder and only 1 file 
        if files[0]== "desktop.ini" or:  
            send2trash(dir)
            print(dir, ": folder removed")
        else:
            print(dir)

delete folder if it contains only .srt or .txt file(s)

    elif subdirs == []: #if dir doesn’t contains subdirectory
        ext = (".srt", ".txt")
        contains_other_ext=0
        for file in files:
            if not file.endswith(ext):  
                contains_other_ext=True
        if contains_other_ext== 0:
                send2trash(dir)
                print(dir, ": dir deleted")

Delete folder if its size is less than 400kb :

def get_tree_size(path):
    """Return total size of files in given path and subdirs."""
    total = 0
    for entry in os.scandir(path):
        if entry.is_dir(follow_symlinks=False):
            total += get_tree_size(entry.path)
        else:
            total += entry.stat(follow_symlinks=False).st_size
    return total


for dir, subdirs, files in os.walk(root):   
    If get_tree_size(dir) < 400000:  # ˜ 400kb
        send2trash(dir)
    print(dir, "dir deleted")

def deleteDir(dirPath):
    deleteFiles = []
    deleteDirs = []
    for root, dirs, files in os.walk(dirPath):
        for f in files:
            deleteFiles.append(os.path.join(root, f))
        for d in dirs:
            deleteDirs.append(os.path.join(root, d))
    for f in deleteFiles:
        os.remove(f)
    for d in deleteDirs:
        os.rmdir(d)
    os.rmdir(dirPath)

from python 3.4 you may use :

import pathlib

def delete_folder(pth) :
    for sub in pth.iterdir() :
        if sub.is_dir() :
            delete_folder(sub)
        else :
            sub.unlink()
    pth.rmdir() # if you just want to delete the dir content but not the dir itself, remove this line

where pth is a pathlib.Path instance. Nice, but may not be the fastest.

I'd like to add a "pure pathlib" approach:

from pathlib import Path
from typing import Union

def del_dir(target: Union[Path, str], only_if_empty: bool = False):
    """
    Delete a given directory and its subdirectories.

    :param target: The directory to delete
    :param only_if_empty: Raise RuntimeError if any file is found in the tree
    """
    target = Path(target).expanduser()
    assert target.is_dir()
    for p in sorted(target.glob('**/*'), reverse=True):
        if not p.exists():
            continue
        p.chmod(0o666)
        if p.is_dir():
            p.rmdir()
        else:
            if only_if_empty:
                raise RuntimeError(f'{p.parent} is not empty!')
            p.unlink()
    target.rmdir()

This relies on the fact that Path is orderable, and longer paths will always sort after shorter paths, just like str. Therefore, directories will come before files. If we reverse the sort, files will then come before their respective containers, so we can simply unlink/rmdir them one by one with one pass.

Benefits:

• It's NOT relying on external binaries: everything uses Python's batteries-included modules (Python >= 3.6)
  • Which means that it does not need to repeatedly start a new subprocess to do unlinking
• It's quite fast & simple; you don't have to implement your own recursion
• It's cross-platform (at least, that's what pathlib promises in Python 3.6; no operation above stated to not run on Windows)
• If needed, one can do a very granular logging, e.g., log each deletion as it happens.

I have found a very easy way to Delete any folder(Even NOT Empty) or file on WINDOWS OS.

os.system('powershell.exe  rmdir -r D:\workspace\Branches\*%s* -Force' %CANDIDATE_BRANCH)

from python 3.4 you may use :

import pathlib

def delete_folder(pth) :
    for sub in pth.iterdir() :
        if sub.is_dir() :
            delete_folder(sub)
        else :
            sub.unlink()
    pth.rmdir() # if you just want to delete the dir content but not the dir itself, remove this line

where pth is a pathlib.Path instance. Nice, but may not be the fastest.

def deleteDir(dirPath):
    deleteFiles = []
    deleteDirs = []
    for root, dirs, files in os.walk(dirPath):
        for f in files:
            deleteFiles.append(os.path.join(root, f))
        for d in dirs:
            deleteDirs.append(os.path.join(root, d))
    for f in deleteFiles:
        os.remove(f)
    for d in deleteDirs:
        os.rmdir(d)
    os.rmdir(dirPath)

Recursion-based, pure pathlib solution:

from pathlib import Path

def remove_path(path: Path):
    if path.is_file() or path.is_symlink():
        path.unlink()
        return
    for p in path.iterdir():
        remove_path(p)
    path.rmdir()

Supports Windows and symbolic links

Just some python 3.5 options to complete the answers above. (I would have loved to find them here).

import os
import shutil
from send2trash import send2trash # (shutil delete permanently)

Delete folder if empty

root = r"C:\Users\Me\Desktop\test"   
for dir, subdirs, files in os.walk(root):   
    if subdirs == [] and files == []:
           send2trash(dir)
           print(dir, ": folder removed")

Delete also folder if it contains this file

    elif subdirs == [] and len(files) == 1: # if contains no sub folder and only 1 file 
        if files[0]== "desktop.ini" or:  
            send2trash(dir)
            print(dir, ": folder removed")
        else:
            print(dir)

delete folder if it contains only .srt or .txt file(s)

    elif subdirs == []: #if dir doesn’t contains subdirectory
        ext = (".srt", ".txt")
        contains_other_ext=0
        for file in files:
            if not file.endswith(ext):  
                contains_other_ext=True
        if contains_other_ext== 0:
                send2trash(dir)
                print(dir, ": dir deleted")

Delete folder if its size is less than 400kb :

def get_tree_size(path):
    """Return total size of files in given path and subdirs."""
    total = 0
    for entry in os.scandir(path):
        if entry.is_dir(follow_symlinks=False):
            total += get_tree_size(entry.path)
        else:
            total += entry.stat(follow_symlinks=False).st_size
    return total


for dir, subdirs, files in os.walk(root):   
    If get_tree_size(dir) < 400000:  # ˜ 400kb
        send2trash(dir)
    print(dir, "dir deleted")

if you are sure, that you want to delete the entire dir tree, and are no more interested in contents of dir, then crawling for entire dir tree is stupidness... just call native OS command from python to do that. It will be faster, efficient and less memory consuming.

RMDIR c:\blah /s /q 

or *nix

rm -rf /home/whatever 

In python, the code will look like..

import sys
import os

mswindows = (sys.platform == "win32")

def getstatusoutput(cmd):
    """Return (status, output) of executing cmd in a shell."""
    if not mswindows:
        return commands.getstatusoutput(cmd)
    pipe = os.popen(cmd + ' 2>&1', 'r')
    text = pipe.read()
    sts = pipe.close()
    if sts is None: sts = 0
    if text[-1:] == '\n': text = text[:-1]
    return sts, text


def deleteDir(path):
    """deletes the path entirely"""
    if mswindows: 
        cmd = "RMDIR "+ path +" /s /q"
    else:
        cmd = "rm -rf "+path
    result = getstatusoutput(cmd)
    if(result[0]!=0):
        raise RuntimeError(result[1])

For Windows, if directory is not empty, and you have read-only files or you get errors like

• Access is denied
• The process cannot access the file because it is being used by another process

Try this, os.system('rmdir /S /Q "{}"'.format(directory))

It's equivalent for rm -rf in Linux/Mac.

Ten years later and using Python 3.7 and Linux there are still different ways to do this:

import subprocess
from pathlib import Path

#using pathlib.Path
path = Path('/path/to/your/dir')
subprocess.run(["rm", "-rf", str(path)])

#using strings
path = "/path/to/your/dir"
subprocess.run(["rm", "-rf", path])

Essentially it's using Python's subprocess module to run the bash script $ rm -rf '/path/to/your/dir as if you were using the terminal to accomplish the same task. It's not fully Python, but it gets it done.

The reason I included the pathlib.Path example is because in my experience it's very useful when dealing with many paths that change. The extra steps of importing the pathlib.Path module and converting the end results to strings is often a lower cost to me for development time. It would be convenient if Path.rmdir() came with an arg option to explicitly handle non-empty dirs.

import os
import stat
import shutil

def errorRemoveReadonly(func, path, exc):
    excvalue = exc[1]
    if func in (os.rmdir, os.remove) and excvalue.errno == errno.EACCES:
        # change the file to be readable,writable,executable: 0777
        os.chmod(path, stat.S_IRWXU | stat.S_IRWXG | stat.S_IRWXO)  
        # retry
        func(path)
    else:
        # raiseenter code here

shutil.rmtree(path, ignore_errors=False, onerror=errorRemoveReadonly) 

If ignore_errors is set, errors are ignored; otherwise, if onerror is set, it is called to handle the error with arguments (func, path, exc_info) where func is os.listdir, os.remove, or os.rmdir; path is the argument to that function that caused it to fail; and exc_info is a tuple returned by sys.exc_info(). If ignore_errors is false and onerror is None, an exception is raised.enter code here

From docs.python.org:

This example shows how to remove a directory tree on Windows where some of the files have their read-only bit set. It uses the onerror callback to clear the readonly bit and reattempt the remove. Any subsequent failure will propagate.

import os, stat
import shutil

def remove_readonly(func, path, _):
    "Clear the readonly bit and reattempt the removal"
    os.chmod(path, stat.S_IWRITE)
    func(path)

shutil.rmtree(directory, onerror=remove_readonly)

To delete a folder even if it might not exist (avoiding the race condition in Charles Chow's answer) but still have errors when other things go wrong (e.g. permission problems, disk read error, the file isn't a directory)

For Python 3.x:

import shutil

def ignore_absent_file(func, path, exc_inf):
    except_instance = exc_inf[1]
    if isinstance(except_instance, FileNotFoundError):
        return
    raise except_instance

shutil.rmtree(dir_to_delete, onerror=ignore_absent_file)

The Python 2.7 code is almost the same:

import shutil
import errno

def ignore_absent_file(func, path, exc_inf):
    except_instance = exc_inf[1]
    if isinstance(except_instance, OSError) and \
        except_instance.errno == errno.ENOENT:
        return
    raise except_instance

shutil.rmtree(dir_to_delete, onerror=ignore_absent_file)

In my case the only way to delete was by using all possibilities because my code was supposed to run either by cmd.exe or powershell.exe. If it is your case, just create a function with this code and you will be fine:

        #!/usr/bin/env python3

        import shutil
        from os import path, system
        import sys

        # Try to delete the folder ---------------------------------------------
        if (path.isdir(folder)):
            shutil.rmtree(folder, ignore_errors=True)

        if (path.isdir(folder)):
            try:
                system("rd -r {0}".format(folder))
            except Exception as e:
                print("WARN: Failed to delete => {0}".format(e),file=sys.stderr)

        if (path.isdir(self.backup_folder_wrk)):
            try:
                system("rd /s /q {0}".format(folder))
            except Exception as e:
                print("WARN: Failed to delete => {0}".format(e),file=sys.stderr)

        if (path.isdir(folder)):
            print("WARN: Failed to delete {0}".format(folder),file=sys.stderr)
        # -------------------------------------------------------------------------------------

if you are sure, that you want to delete the entire dir tree, and are no more interested in contents of dir, then crawling for entire dir tree is stupidness... just call native OS command from python to do that. It will be faster, efficient and less memory consuming.

RMDIR c:\blah /s /q 

or *nix

rm -rf /home/whatever 

In python, the code will look like..

import sys
import os

mswindows = (sys.platform == "win32")

def getstatusoutput(cmd):
    """Return (status, output) of executing cmd in a shell."""
    if not mswindows:
        return commands.getstatusoutput(cmd)
    pipe = os.popen(cmd + ' 2>&1', 'r')
    text = pipe.read()
    sts = pipe.close()
    if sts is None: sts = 0
    if text[-1:] == '\n': text = text[:-1]
    return sts, text


def deleteDir(path):
    """deletes the path entirely"""
    if mswindows: 
        cmd = "RMDIR "+ path +" /s /q"
    else:
        cmd = "rm -rf "+path
    result = getstatusoutput(cmd)
    if(result[0]!=0):
        raise RuntimeError(result[1])

I'd like to add a "pure pathlib" approach:

from pathlib import Path
from typing import Union

def del_dir(target: Union[Path, str], only_if_empty: bool = False):
    """
    Delete a given directory and its subdirectories.

    :param target: The directory to delete
    :param only_if_empty: Raise RuntimeError if any file is found in the tree
    """
    target = Path(target).expanduser()
    assert target.is_dir()
    for p in sorted(target.glob('**/*'), reverse=True):
        if not p.exists():
            continue
        p.chmod(0o666)
        if p.is_dir():
            p.rmdir()
        else:
            if only_if_empty:
                raise RuntimeError(f'{p.parent} is not empty!')
            p.unlink()
    target.rmdir()

This relies on the fact that Path is orderable, and longer paths will always sort after shorter paths, just like str. Therefore, directories will come before files. If we reverse the sort, files will then come before their respective containers, so we can simply unlink/rmdir them one by one with one pass.

Benefits:

• It's NOT relying on external binaries: everything uses Python's batteries-included modules (Python >= 3.6)
  • Which means that it does not need to repeatedly start a new subprocess to do unlinking
• It's quite fast & simple; you don't have to implement your own recursion
• It's cross-platform (at least, that's what pathlib promises in Python 3.6; no operation above stated to not run on Windows)
• If needed, one can do a very granular logging, e.g., log each deletion as it happens.

In my case the only way to delete was by using all possibilities because my code was supposed to run either by cmd.exe or powershell.exe. If it is your case, just create a function with this code and you will be fine:

        #!/usr/bin/env python3

        import shutil
        from os import path, system
        import sys

        # Try to delete the folder ---------------------------------------------
        if (path.isdir(folder)):
            shutil.rmtree(folder, ignore_errors=True)

        if (path.isdir(folder)):
            try:
                system("rd -r {0}".format(folder))
            except Exception as e:
                print("WARN: Failed to delete => {0}".format(e),file=sys.stderr)

        if (path.isdir(self.backup_folder_wrk)):
            try:
                system("rd /s /q {0}".format(folder))
            except Exception as e:
                print("WARN: Failed to delete => {0}".format(e),file=sys.stderr)

        if (path.isdir(folder)):
            print("WARN: Failed to delete {0}".format(folder),file=sys.stderr)
        # -------------------------------------------------------------------------------------

If you don't want to use the shutil module you can just use the os module.

from os import listdir, rmdir, remove
for i in listdir(directoryToRemove):
    os.remove(os.path.join(directoryToRemove, i))
rmdir(directoryToRemove) # Now the directory is empty of files

From docs.python.org:

This example shows how to remove a directory tree on Windows where some of the files have their read-only bit set. It uses the onerror callback to clear the readonly bit and reattempt the remove. Any subsequent failure will propagate.

import os, stat
import shutil

def remove_readonly(func, path, _):
    "Clear the readonly bit and reattempt the removal"
    os.chmod(path, stat.S_IWRITE)
    func(path)

shutil.rmtree(directory, onerror=remove_readonly)

Base on kkubasik's answer, check if folder exists before remove, more robust

import shutil
def remove_folder(path):
    # check if folder exists
    if os.path.exists(path):
         # remove if exists
         shutil.rmtree(path)
    else:
         # throw your exception to handle this special scenario
         raise XXError("your exception") 
remove_folder("/folder_name")

To delete a folder even if it might not exist (avoiding the race condition in Charles Chow's answer) but still have errors when other things go wrong (e.g. permission problems, disk read error, the file isn't a directory)

For Python 3.x:

import shutil

def ignore_absent_file(func, path, exc_inf):
    except_instance = exc_inf[1]
    if isinstance(except_instance, FileNotFoundError):
        return
    raise except_instance

shutil.rmtree(dir_to_delete, onerror=ignore_absent_file)

The Python 2.7 code is almost the same:

import shutil
import errno

def ignore_absent_file(func, path, exc_inf):
    except_instance = exc_inf[1]
    if isinstance(except_instance, OSError) and \
        except_instance.errno == errno.ENOENT:
        return
    raise except_instance

shutil.rmtree(dir_to_delete, onerror=ignore_absent_file)

You can use os.system command for simplicity:

import os
os.system("rm -rf dirname")

As obvious, it actually invokes system terminal to accomplish this task.

Ten years later and using Python 3.7 and Linux there are still different ways to do this:

import subprocess
from pathlib import Path

#using pathlib.Path
path = Path('/path/to/your/dir')
subprocess.run(["rm", "-rf", str(path)])

#using strings
path = "/path/to/your/dir"
subprocess.run(["rm", "-rf", path])

Essentially it's using Python's subprocess module to run the bash script $ rm -rf '/path/to/your/dir as if you were using the terminal to accomplish the same task. It's not fully Python, but it gets it done.

The reason I included the pathlib.Path example is because in my experience it's very useful when dealing with many paths that change. The extra steps of importing the pathlib.Path module and converting the end results to strings is often a lower cost to me for development time. It would be convenient if Path.rmdir() came with an arg option to explicitly handle non-empty dirs.

From the python docs on os.walk():

# Delete everything reachable from the directory named in 'top',
# assuming there are no symbolic links.
# CAUTION:  This is dangerous!  For example, if top == '/', it
# could delete all your disk files.
import os
for root, dirs, files in os.walk(top, topdown=False):
    for name in files:
        os.remove(os.path.join(root, name))
    for name in dirs:
        os.rmdir(os.path.join(root, name))

Recursion-based, pure pathlib solution:

from pathlib import Path

def remove_path(path: Path):
    if path.is_file() or path.is_symlink():
        path.unlink()
        return
    for p in path.iterdir():
        remove_path(p)
    path.rmdir()

Supports Windows and symbolic links

import os
import stat
import shutil

def errorRemoveReadonly(func, path, exc):
    excvalue = exc[1]
    if func in (os.rmdir, os.remove) and excvalue.errno == errno.EACCES:
        # change the file to be readable,writable,executable: 0777
        os.chmod(path, stat.S_IRWXU | stat.S_IRWXG | stat.S_IRWXO)  
        # retry
        func(path)
    else:
        # raiseenter code here

shutil.rmtree(path, ignore_errors=False, onerror=errorRemoveReadonly) 

If ignore_errors is set, errors are ignored; otherwise, if onerror is set, it is called to handle the error with arguments (func, path, exc_info) where func is os.listdir, os.remove, or os.rmdir; path is the argument to that function that caused it to fail; and exc_info is a tuple returned by sys.exc_info(). If ignore_errors is false and onerror is None, an exception is raised.enter code here

With os.walk I would propose the solution which consists of 3 one-liner Python calls:

python -c "import sys; import os; [os.chmod(os.path.join(rs,d), 0o777) for rs,ds,fs in os.walk(_path_) for d in ds]"
python -c "import sys; import os; [os.chmod(os.path.join(rs,f), 0o777) for rs,ds,fs in os.walk(_path_) for f in fs]"
python -c "import os; import shutil; shutil.rmtree(_path_, ignore_errors=False)"

The first script chmod's all sub-directories, the second script chmod's all files. Then the third script removes everything with no impediments.

I have tested this from the "Shell Script" in a Jenkins job (I did not want to store a new Python script into SCM, that's why searched for a one-line solution) and it worked for Linux and Windows.

From the python docs on os.walk():

# Delete everything reachable from the directory named in 'top',
# assuming there are no symbolic links.
# CAUTION:  This is dangerous!  For example, if top == '/', it
# could delete all your disk files.
import os
for root, dirs, files in os.walk(top, topdown=False):
    for name in files:
        os.remove(os.path.join(root, name))
    for name in dirs:
        os.rmdir(os.path.join(root, name))

I have found a very easy way to Delete any folder(Even NOT Empty) or file on WINDOWS OS.

os.system('powershell.exe  rmdir -r D:\workspace\Branches\*%s* -Force' %CANDIDATE_BRANCH)

You can use os.system command for simplicity:

import os
os.system("rm -rf dirname")

As obvious, it actually invokes system terminal to accomplish this task.