5

more basic questions i'm struggling with...

Give the basic code below... how does the person object get the address "attached" to it. 

class Person(object):
    def __init__(self, fn, ln):
        self.uid    = Id_Class.new_id("Person")
        self.f_name = fn
        self.l_name = ln

class Address(object):
    def __init__(self, st, sub):
        self.uid    = Id_Class.new_id("Address")
        self.street = st
        self.suburb = sub

s = Person('John', 'Doe')

hm = Address('Queen St.', 'Sydney')
Improve this question

4 Answers 4

Reset to default
9

Try:

class Person(object):
    def __init__(self, fn, ln, address):
        self.uid    = Id_Class.new_id("Person")
        self.f_name = fn
        self.l_name = ln
        self.address = address

class Address(object):
    def __init__(self, st, sub):
        self.uid    = Id_Class.new_id("Address")
        self.street = st
        self.suburb = sub

hm = Address('Queen St.', 'Sydney')

s = Person('John', 'Doe', hm)
Improve this answer
Sign up to request clarification or add additional context in comments.

3 Comments

+1. Passing it into the constructor is almost always the answer.
ahh nope, didn't try that... :) Sorry, just to clarify, this is passing the reference to the address object to the attribute Person.address?
No, it's passing it when you make Person. A Person isn't valid without an Address, so you pass it along with the name.
1

However you want. Perhaps the simplest way is:

s.address = hm

But you don't say what you mean by "attach". If you want something more elaborate (e.g., if you want the address to be created when the Person object is created) then you'll need to explain more.

Improve this answer

1 Comment

i think i'm overcomplicating the idea of oop in my own thinking
1

you cound always attach it by assigning to it (which would in fact update an internal dict dictionnary of your instace)

s.address = hm
print s.address
>>> <object Address ...>

Or better, do it from within your constructor, namely

class Person : 
    def __init__(self,fn,ln,my_address) : 
        [... snip]
        self.address = my_address

which is exactly the same, but you will be sure you have an address attribute and can always have default arguments for no values, such as 

def __init__(self,fn,ln,my_address=None) : ...
Improve this answer

Comments

1

As BrenBam askes, by what do you mean by attach. You should ask yourself how much data are you talking about, how reliable you need it. I see you seperate the address class from the person class. This implies these different objects will be used in differentiating ways.

You should be aware of common database design practices and why you would seperate the address from the person or conversely why you wouldn't.

for example:

addresses = {}
persons = {}
records = {}

address = Address(...)
person1 = Person(...)
person2 = Person(...)

addresses[address.uid] = address
persons[person1.uid] = person1
persons[person2.uid] = person2

records[address.uid] = [person1.uid, person2.uid]

This is a better solution for address with lots of people at it, or people that move around alot, or for apps that don't care who lives or works at an address, just that there is a number of them each deserving of some very important junk mail needing to inspire them for your products. A military troop mailing app would benefit as things and people move all over the place.

Improve this answer

8 Comments

Thanks for the response @DevPlayer, I'm still pretty green but from what i've read - you create objects with specific data ie person, address, car - hense why i split it out. Would def be easier to have it all contained
Oh sorry I meant to ask that "why do you split person and address seperately?" question rhetorically. it was an excersize in "how to chunk data". If in your application the person is never without an address and an address has one person and only one person object and a person object and address object are stored in file with each field for every record then there would be little reason to split the address and person fields into their respective classes. By splitting them you gain storage space and consistant 4matng 4 "1600 Washington Ave" will be the same format for all peoples living there.
My gist of my comment above is you'll want a good reason to seperate data, usually because each data subset is handled and or processed differently. BUT that point leads to some kind of answer to your original question; how to attach the two data types. Do you link them by object, by reference, by merging,. There's many was to attach data and pros and cons to "attaching" data by those means. My example was using a pythonic way of doing "a-poor-mans-relational database method". The key and value of records was the "link" between subdata sets person and address.
Instead of an outside 3rd object, ie "records", we could have put a reference "inside" the person object referring to the address object. Or you could use the Python ducktyping method. where you have a Person class and a subclass of a Person class called Address where the address has all the Person attributes as well as some new ones, kind of a "merge" of fields. That is just a flatter version of putting an attribute in Person that -points- to an address object.
this is what i've been struggling with: how is the best way to create an app with data storage and speed in the front of my mind. I feel i'm getting a head of my self (with regard to programming ability). Do i chunk up data so it is "easier" to store in a relational db or do i embed all the data necessary to store in a nonRel db like couch or mongrel??
|

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.