checksum

checksum.py

#!/usr/bin/env python

import os
import sys
import hashlib
import pprint


def getSHA1(path):
    sha1 = hashlib.sha1()
    with open(path, "rb") as f:
        data = f.read(1024)
        while data:
            sha1.update(data)
            data = f.read(1024)

    f.close()
    return sha1.hexdigest()


def findDuplicates(path):
    fileHash = {}
    for current, dirs, files in os.walk(path):
        for f in files:
            if os.path.splitext(f)[1].lower() in ['.mp3', '.wav', '.wma', '.ogg']:
                print "Processing: %s" % os.path.join(current, f)
                sha = getSHA1(os.path.join(current, f))
                if fileHash.has_key(sha):
                    fileHash[sha]["files"].append(os.path.join(current, f))
                else:
                    fileHash[sha] = { "files": [os.path.join(current, f)] }

    # pp = pprint.PrettyPrinter(indent=2)                        
    # pp.pprint(fileHash)
    for sha1 in fileHash:
        if len(fileHash[sha1]["files"]) > 1:
            print "-"*80
            print "[%s - %s - %s]" % (sha1, os.path.split(fileHash[sha1]["files"][0])[1], len(fileHash[sha1]["files"]))
            for f in fileHash[sha1]["files"]:
                print "\t%s" % f

            print ""
            for f in fileHash[sha1]["files"][1:]:
                os.remove(f)
                print "\tRemoved: %s" % f

def main():
    if len(sys.argv) > 1:
        for path in sys.argv[1:]:
            findDuplicates(path)
    else:
        print "Error: Too few arguments!"
        print "Usage: $python %s path/to/directory" % sys.argv[0]


if __name__ == "__main__":
    main()