python 清理文件名不同的相同文件

#!/usr/bin/python  
#coding:utf8

import sys
import filecmp
import os
import hashlib

allrootpath = [
    "D:\\360down\\now",
    "F:\\pl",
    "C:\\newgif",
    "H:\games\\newdoc\\360down",
]

all_types = [".rar", ".gif", ".zip", ".jpg", ".png"]

allfiles = []

runforall = False

def is_valid_type(filepath):
    #valid_types = all_types
    valid_types = [".gif"]
    re = False
    for types in valid_types:
        if filepath.find(types) > -1:
            re = True
            break
    return re

def CalcMD5(filepath):
    with open(filepath,'rb') as f:
        md5obj = hashlib.md5()
        md5obj.update(f.read())
        hash = md5obj.hexdigest()
        return hash

def traverseDirByOSWalk(path):
    global allfiles
    path = os.path.expanduser(path)
    for (dirname, subdir, subfile) in os.walk(path):
        for f in subfile:
            fullfile = os.path.join(dirname, f)
            if runforall or is_valid_type(fullfile):
                allfiles.append(fullfile)

def dirlist(path, allfile):  
    filelist =  os.listdir(path)  
  
    for filename in filelist:  
        filepath = os.path.join(path, filename)  
        if os.path.isdir(filepath):  
            dirlist(filepath, allfile)  
        else:  
            allfile.append(filepath)  
    return allfile

for rootpath in allrootpath:
    traverseDirByOSWalk(rootpath)

print len(allfiles)

fileMd5Dir = {}

counter = 0

for file in allfiles:
    code = CalcMD5(file)
    if fileMd5Dir.has_key(code) == False:
        fileMd5Dir[code] = []
        fileMd5Dir[code].append(file)
    else:
        isExist = False
        for existFile in fileMd5Dir[code]:
            if filecmp.cmp(existFile, file) == True:
                isExist = True
                break
        if isExist == True:
            os.remove(file)
            print "Delete " + file + " !!!"
            counter = counter + 1
            print counter
        else:
            fileMd5Dir[code].append(file)