Archive files in fast growing directories, group a days file to a zip etc. Run via cron or taskscheduler

archive_files_to_zip.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-
from __future__ import with_statement
from contextlib import closing
from zipfile import ZipFile, ZIP_DEFLATED
import os
import datetime
import shutil

def zipdir(basedir, archivename):
	assert os.path.isdir(basedir)
	with closing(ZipFile(archivename, "w", ZIP_DEFLATED)) as z:
		for root, dirs, files in os.walk(basedir):
			#NOTE: ignore empty directories
			for fn in files:
				absfn = os.path.join(root, fn)
				zfn = absfn[len(basedir)+len(os.sep):] #XXX: relative path
				z.write(absfn, zfn)

def archiveToZipByMonth(dir_to_search, file_type):
	source = os.listdir(dir_to_search)
	for files in source:
		#Get only our filetypes
		if files.endswith("."+file_type):
			curpath = os.path.join(dir_to_search, files)
			file_modified = datetime.datetime.fromtimestamp(os.path.getmtime(curpath))
			filedate = str(file_modified.year)+"-"+str(file_modified.month)
			destinationdir = dir_to_search+"\\"+filedate+"\\"
			destinationfile = dir_to_search+"\\"+filedate+"\\"+files
			#Do not move if it is this day
			if not filedate == str(datetime.date.today().year)+"-"+str(datetime.date.today().month):
				#Create dir if not exists
				if not os.path.exists(destinationdir):
					os.mkdir(destinationdir)
				if os.path.exists(destinationfile):
					os.remove(destinationfile)
				shutil.move(dir_to_search+"\\"+files,destinationdir)

	source = os.listdir(dir_to_search)
	for files in source:
		if os.path.isdir(dir_to_search+"\\"+files):
			if not files == str(datetime.date.today().year)+"-"+str(datetime.date.today().month):
				destinationzip = dir_to_search+"\\"+files+".zip"
				if not os.path.exists(destinationzip):
					zipdir(dir_to_search+"\\"+files, destinationzip)
					shutil.rmtree(dir_to_search+"\\"+files)

def archiveToZipByDay(dir_to_search, file_type):
	source = os.listdir(dir_to_search)
	for files in source:
		#Get only our filetypes
		if files.endswith("."+file_type):
			curpath = os.path.join(dir_to_search, files)
			file_modified = datetime.datetime.fromtimestamp(os.path.getmtime(curpath))
			filedate = str(file_modified.year)+"-"+str(file_modified.month)+"-"+str(file_modified.day)
			destinationdir = dir_to_search+"\\"+filedate+"\\"
			destinationfile = dir_to_search+"\\"+filedate+"\\"+files
			#Do not move if it is this day
			if not filedate == str(datetime.date.today().year)+"-"+str(datetime.date.today().month)+"-"+str(datetime.date.today().day):
				#Create dir if not exists
				if not os.path.exists(destinationdir):
					os.mkdir(destinationdir)
				if os.path.exists(destinationfile):
					os.remove(destinationfile)
				shutil.move(dir_to_search+"\\"+files,destinationdir)

	source = os.listdir(dir_to_search)
	for files in source:
		if os.path.isdir(dir_to_search+"\\"+files):
			if not files == str(datetime.date.today().year)+"-"+str(datetime.date.today().month)+"-"+str(datetime.date.today().day):
				destinationzip = dir_to_search+"\\"+files+".zip"
				if not os.path.exists(destinationzip):
					zipdir(dir_to_search+"\\"+files, destinationzip)
					shutil.rmtree(dir_to_search+"\\"+files)

def removeZipsOlderThan(dir_to_search, num_days):
	source = os.listdir(dir_to_search)
	for files in source:
		#Get only our filetypes
		if files.endswith(".zip"):
			curpath = os.path.join(dir_to_search, files)
			file_modified = datetime.datetime.fromtimestamp(os.path.getmtime(curpath))
			if datetime.datetime.now() - file_modified > datetime.timedelta(days=num_days):
				os.remove(curpath)
				print("removed file")
	pass

archiveToZipByDay("C:\\rotate", "rtf")
removeZipsOlderThan("C:\\rotate", 90)