Namespace in Python

Posted on Edited on In Views: Word count in article: 215 Reading time ≈ 1 mins.
Namespaces are actually dictionaries in python which store the objects that compiler can access to and the mapping relationship between names and objects.

There are totally 4 types of namespace in python, that is LEGB. They are all independent from each other. However, there is priority among them. When compiler searches for a variable, it will search in the sequence of LEGB:

  • Locals: The namespace inside a function, which including local variables and formal parameters: 
    1
    2
    3
    4
    def func1(x, y):
        a = 1
        print(locals())
    func1(1, 1)
    1
    2
    Output:
    {'x': 1, 'y': 1, 'a': 1}
  • Enclosing function: The namespace of outer function in nested function: 
    1
    2
    3
    4
    5
    6
    7
    8
    9
    def func1(x, y):
        a = 1

        def func2(z, k):
            print(locals())

        func2(1, 1)

    func1(1, 1)
    1
    2
    output:
    {'z': 1, 'k': 1}
    1
    2
    3
    4
    5
    6
    7
    8
    9
    10
    11
    def func1(x, y):
        a = 1
        
        def func2(z, k):
            z += x
            k += y
            print(locals())
        
        func2(1, 1)

    func1(1, 1)
    1
    2
    output:
    {'z': 2, 'k': 2, 'x': 1, 'y': 1}
    This shows that the inner function can access to the outer function (only read) and add variables in enclosing function to its local namespace.
  • Globals: The namespace of current module, that is .py file, including variables, functions, classes and so on.
  • __builtins__: The namespace of built-in variables and built-in functions.

Variables in higher-priority namespaces will override variables with the same name in lower-priority namespaces.

Reference: