Finding and deleting files using python script [duplicate] - linux

Question

This question already has answers here:
Get a filtered list of files in a directory
(14 answers)
Closed 6 years ago.
I am writing a Python script to find and remove all .py files having corresponding .pyc files.
How to extract this file list and remove them?
For example : consider there some file in /foo/bar:
file.py
file.pyc
file3.py
file2.py
file2.pyc...etc
I want to delete file.py,file2.py and not file3.py as it do not have corresponding .pyc file.
and I want to do in all folders under '/'.
Is there one-liner bash code for the same?
P.S : I am using CentOS 6.8, having python2.7

Answer 1

Here's my solution:
import os
ab=[]
for roots,dirs,files in os.walk("/home/foo/bar/"):
for file in files:
if file.endswith(".py"):
ab.append(os.path.join(roots,file))
bc=[]
for i in range(len(ab)):
bc.append(ab[i]+"c")
xy=[]
for roots,dirs,files in os.walk("/home/foo/bar/"):
for file in files:
if file.endswith(".pyc"):
xy.append(os.path.join(roots,file))
ex=[x[:-1] for x in bc if x in xy]
for i in ex:
os.remove(i)
P.S: Newbie in python scriptiing.

Answer 2

Bash solution:
#!/bin/bash
find /foo/bar -name "*.py" -exec ls {} \; > file1.txt
find /foo/bar/ -name "*.pyc" -exec ls {} \; > file2.txt
p=`wc -l file1.txt| cut -d' ' -f1`
for ((c=1;c<=$p;c++))
do
grep `sed -n ${c}p file1.txt | sed s/$/c/g` file2.txt > /dev/null
if [ $? -eq 0 ]
then
list=`sed -n ${c}p file1.txt`
echo " exist : $list"
rm -rf `sed -n ${c}p file1.txt`
fi
done

Answer 3

this is a very operating-system-near solution
maybe make a shell script from the following commands and invoke it from python using subprocess.call (How to call a shell script from python code?, Calling an external command in Python)
find . -name "*.pyc" > /tmp/pyc.txt
find . -name "*.py" > /tmp/py.txt
from the entries of these files remove path and file ending using sed or basename:
for f in $(cat /tmp/pyc.txt) ; do
sed 's/.*\///' remove path
sed 's/\.[^.]*$//' remove file ending
done
for f in $(cat /tmp/py.txt) ; do
sed 's/.*\///' remove path
sed 's/\.[^.]*$//' remove file ending
done
(https://unix.stackexchange.com/questions/44735/how-to-get-only-filename-using-sed)
awk 'FNR==NR{a[$1];next}($1 in a){print}' /tmp/pyc.txt /tmp/py.txt > /tmp/rm.txt (https://unix.stackexchange.com/questions/125155/compare-two-files-for-matching-lines-and-store-positive-results)
for f in $(cat /tmp/rm.txt) ; do
rm $f
done (Unix: How to delete files listed in a file)

Answer 4

The following code will work for a single layer directory. (Note: I wasn't sure how you wanted to handle multiple layers of folders --- e.g. if you have A.py in one folder and A.pyc in another, does it count as having both present, or do they have to be in the same layer of the same folder? If the latter case, it should be fairly simple to just loop through the folders and just call this code within each loop.)
import os
# Produces a sorted list of all files in a directory
dirList = os.listdir(folder_path) # Use os.listdir() if want current directory
dirList.sort()
# Takes advantage of fact that both py and pyc files will share same name and
# that pyc files will appear immediately after their py counterparts in dirList
lastPyName = ""
for file in dirList:
if file[-3:] == ".py":
lastPyName = file[:-3]
elif file[-4:] == ".pyc":
if lastPyName == file[:-4]:
os.remove(lastPyName + ".py")
os.remove(lastPyName + ".pyc") # In case you want to delete this too