Pickle file with many objects

Question

I cannot get my file to store multiple instances of contacts. After adding new contact and trying to print them, it comes up with "IndexError: list index out of range" error. What shall I do to make it work?

import pickle

class People():
    def __init__(self, name, surname, age, mobile_no, home_no):
        self.name = name
        self.surname = surname
        self.age = age
        self.mobile_no = mobile_no
        self.home_no = home_no

    def DisplayContacts(self):
        print("First Name: \t", self.name)
        print("Surname: \t", self.surname)
        print("Age: \t", self.age)
        print("Mobile Number: \t", self.mobile_no)
        print("Home Number: \t", self.home_no)
        print()


def addContact():
    newname = str(input("First name: \t"))
    newsurname = str(input("Surname: \t"))
    newage = int(input("Age: \t"))
    newmobile_no = int(input("Mobile Number: \t"))
    newhome_no = int(input("Home Number: \t"))
    newContact = People(newname, newsurname, newage, newmobile_no, newhome_no) 
    return newContact

cont = 1

contacts = []


while cont == 1:
    user = input("Do you want to add contact? (Y/N)")
    if user == "Y" or user == "y":
        print ("works")
        contacts.append(addContact())
        file = open("CList.pickle", "ab")
        pickle.dump(contacts, file, pickle.HIGHEST_PROTOCOL)
        file.close()
    else:
        print ("111")
        cont = 0


useropen = input("open file? (Y/N)")
if useropen == "Y" or useropen == "y":


    with open ("CList.pickle", "rb") as pickled_file:
        contacts = pickle.load(pickled_file)
        print(contacts[0].surname)
        print(contacts[1].surname)


else:
    print("Null")

Looks like each time you make a new contact, you open the file and append the pickled object to the existing data there. Are you sure that's allowed? I don't think pickle.dumps(a) + pickle.dumps(b) is the same as pickle.dumps([a,b]). — Kevin
– Kevin, Commented Jun 19, 2014 at 19:41

tdelaney · Accepted Answer · 2014-06-19 20:00:55Z

1

Simply appending a picked object to a file is not the same thing as pickling a list. EAch time you append, you've created another pickled record. Read the file multiple times to get your list:

with open ("CList.pickle", "rb") as pickled_file:
    contacts = []
    try:
        while True:
            contacts.append(pickle.load(pickled_file))
    except EOFError:
            pass

Now, instead of appending the list of contacts (which would give you a list of lists with many duplicates), just pickle the new contact:

with open("CList.pickle", "ab") as _file:
    while True:
        user = input("Do you want to add contact? (Y/N)")
        if user == "Y" or user == "y":
            print ("works")
            pickle.dump(addContact(), _file, pickle.HIGHEST_PROTOCOL)
        else:
            print ("111")
            break

edited Jun 19, 2014 at 20:00

answered Jun 19, 2014 at 19:44

tdelaney

77.9k6 gold badges91 silver badges129 bronze badges

Sign up to request clarification or add additional context in comments.

5 Comments

tdelaney Over a year ago

@JoeDoe - have you changed the object you are pickling? You could have old cruft in the file - deleting and starting over may help. I also realized I didn't note that you should change how you pickle also... just pickle the contact object, not the list of contact objects. I am making an update to the anwser.

Joe Doe Over a year ago

I did all as you suggested but when I try to load the file it's just a empty prompt

tdelaney Over a year ago

I ran the patched up code and found that dump/load aren't quite symatric. You end up with an extra new line, so the loop that reads the records gets Person, None, Person, None, ... Solving this is mildly annoying, you could for instance prepend each write with the length of the thing you are writing, or more easily just skip the Nones.

Joe Doe Over a year ago

and what would be the way to skip those?

tdelaney Over a year ago

One option is to load into a variable and then check if its None before adding to the list. Another is to build up your own records using ctypes. Pickle into a string and then use ctypes to write a record of length,pickled_string. The reader would read length-ed chunks.

Collectives™ on Stack Overflow

Pickle file with many objects

1 Answer 1

5 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

5 Comments

Your Answer

Sign up or log in

Post as a guest

Related