decode.py

# An adapter that takes CSV as input, performs a MIME/utf-8 decoding, then returns the CSV results
import csv,sys,commands,socket,email.header,re,codecs,traceback

def decode_string(encoded_string):
    match = re.search('(=\?.*?\?(Q|B)\?)', encoded_string)
    if not match:
        return encoded_string

    try:
      header_string = email.header.decode_header(encoded_string)
    except:
      return encoded_string

    try:
      codecs.lookup(header_string[0][1])
    except:
      return "Codec not found"

    decoded = header_string[0][0].decode(header_string[0][1],"replace")

    decoded_utf8 = decoded.encode('utf-8')
    return decoded_utf8

def main():
    if len(sys.argv) != 3:
        print "Usage: python decode_string.py [encoded field] [decoded field]"
        sys.exit(0)

    encoded_stringf = sys.argv[1]
    decoded_stringf = sys.argv[2]
    r = csv.reader(sys.stdin)
    w = None
    header = []
    first = True

    for line in r:
        if first:
            header = line
            if encoded_stringf not in header or decoded_stringf not in header:
                print "Encoded string field and decoded string field must exist in CSV data"
                sys.exit(0)
            csv.writer(sys.stdout).writerow(header)
            w = csv.DictWriter(sys.stdout, header)
            first = False
            continue

        # Read the result
        result = {}
        i = 0
        while i < len(header):
            if i < len(line):
                result[header[i]] = line[i]
            else:
                result[header[i]] = ''
            i += 1

        if len(result[encoded_stringf]):
            result[decoded_stringf] = decode_string(result[encoded_stringf])
            if len(result[decoded_stringf]):
                w.writerow(result)

if __name__=="__main__":
    try:
        main()
    except:
        traceback.print_exc(file=open("/tmp/errlog.txt","a"))
        sys.exit(1)