compares two directories, and list all difference based on contents and size ...

compareDirs.py

#!/usr/bin/env python
# -*- coding: utf-8 -*-

import os, sys
import filecmp
holderlist=[]

def compare_dirs(dir1, dir2):
    dircomp=filecmp.dircmp(dir1,dir2)
    only_in_source=dircomp.left_only
    diff_in_source=dircomp.diff_files
    dirpath=os.path.abspath(dir1)
    [holderlist.append((os.path.abspath(os.path.join(dir1,x)), os.path.getsize(os.path.abspath(os.path.join(dir1,x))))) for x in only_in_source]
    [holderlist.append((os.path.abspath(os.path.join(dir1,x)), os.path.getsize(os.path.abspath(os.path.join(dir1,x))))) for x in diff_in_source]
    if len(dircomp.common_dirs) > 0:
        for item in dircomp.common_dirs:
            compare_dirs(os.path.abspath(os.path.join(dir1,item)), os.path.abspath(os.path.join(dir2,item)))
        return holderlist

def size_format(size):
    units = ['B', 'KB', 'MB', 'GB', 'TB']
    for u in units:
        if size < 1024:  return '%f %s' %(size, u)
        size /= 1024
    return '%f %s' %(size, units[-1])

def main(args):
    #print args
    if len(args) != 2:
        print "Usage: <script> <source> <destination>\n"
        print "< the script compares two folders >\n"
        return

    dir1, dir2 = args[0], args[1]
    if not os.path.isdir(dir1):
        print "the source folder is not there\n"
        print "< the script compares two folders >\n"
        return
    if not os.path.isdir(dir2):
        print "the destination folder is not there\n"
        print "< the script compares two folders >\n"
        return
    print "<< comparision process start >>\n"

    output_string = ''
    compare_dirs(dir1, dir2)
    import operator
    import datetime
    output_string  += "The source location is:" + dir1 + "\n"
    output_string  += "The destination location is:" + dir2 + "\n\n"

    if not holderlist:
        print "there are no difference between the two folders.\n"
    else:
        print "Success. all differences have been found between the two folders.\n"
        holderlist.sort(key=operator.itemgetter(1))
    sizeFile = 0
    for item in holderlist:
        sizeFile += item[1]
        output_string += item[0] + "," + size_format(item[1]) + "\n"

    output_string += "\nThe total size of the difference is:," + str(size_format(sizeFile)) + "\n"
    diff_file = '/nas/projects/Tactic/bilal/render/temp/logs/%s_%s_%s_diff_list.csv' % (str(datetime.date.today()), dir1.split('/')[-1], dir2.split('/')[-1])
    if os.path.isfile(diff_file):
        os.remove(diff_file)
    #size_file = '/test/folder/%s_diff_list.csv' % (str(datetime.date.today()))
    diffFile = open(diff_file, 'w')
    diffFile.writelines(output_string)
    diffFile.close()

    print "<< process is complete.>>\n"
    print "<< for details, please check the file: %s>>\n" % diff_file

if __name__ == '__main__':
    main(sys.argv[1:])