Question

I have been reading a lot about Python OOP here and in the Python tutorial. But some questions bug me on class attributes.

Example:

class Par(object):
    def __init__(self, chip, fig):
        self.fruit = chip
        self.fig = fig
        self.pear =10

1. Shouldn’t self.fruit = chip be self.chip = chip?
2. How come fruit does not appear in the attribute list in the __init__() brackets but it is used?
3. What is the difference between self.fruit = chip and self.chip = chip?
4. What is the difference between declaring an attribute in the __init__() and declaring it outside __init__() e.g. self.pear =10

Answer 1

self will be an instance of Par. self.__dict__ is a dict which holds attributes of self. When you say

self.fruit = chip

an entry in the dict is created: self.__dict__ will include {'fruit':chip}.

When you say

def __init__(self, chip, fig):

you are declaring that Par must be passed two values, chip and fig. (The self instance will be passed to __init__ for you.)

So inside the __init__, the local variables (not attributes!) chip and fig are known.

self.fruit = chip is creating an attribute fruit with value chip.

---

For example,

class Par(object):
    def __init__(self, chip, fig):
        self.fruit = chip
        self.fig = fig
        self.pear = 10

# 99 is being passed in as the value of `chip`
# 'bar' is being passed in as the value of `fig`    
par = Par(99,'bar') 
print(par.__dict__)

yields

{'fruit': 99, 'pear': 10, 'fig': 'bar'}

---

Note that to access the attributes you would normally use, for example, par.fruit rather than par.__dict__['fruit']. I show par.__dict__ above only to give you a picture of what is going on behind the scenes.