Dynamic import and object resolution

April 9, 2025

---

Use the importlib package to dynamically import a Python module by its name. Similarly, use the getattr function to get a class or function by name. This is useful in plugin architectures or when loading things based on configuration files.

Package

To demonstrate, suppose we have a package named pkg as shown below. This package provides a Fruit class and a say_hello function (also shown below).

pkg/
├── __init__.py
├── fruit.py
└── say_hello.py

# pkg/__init__.py

from .fruit import Fruit
from .say_hello import say_hello

__all__ = ["Fruit", "say_hello"]

# pkg/fruit.py

class Fruit:

    def __init__(self, name, qty, price):
        self.name = name
        self.qty = qty
        self.price = price

    def total(self):
        tot = self.qty * self.price
        return tot

# pkg/say_hello.py

def say_hello(s: str):
    greeting = f"hello {s}"
    print(greeting)

Example

The example given here imports the pkg package using the name of the package which is provided as a string to the importlib.import_module. The imported package can then be used like any other package. Next, the getattr function uses the name of a function or class in the package to get the corresponding object.

import importlib

# Import the package and access the say_hello function
pkg = importlib.import_module("pkg")
pkg.say_hello("bart")

# Import the package and get the say_hello function
pkg = importlib.import_module("pkg")
say_hello = getattr(pkg, "say_hello")
say_hello("bart")

# Import the package and get the Fruit class
pkg = importlib.import_module("pkg")
Fruit = getattr(pkg, "Fruit")
fruit = Fruit("apple", 10, 3.58)
tot = fruit.total()
print("Total is", tot)

The example code and package are available in the pythonic repo on GitHub at pythonic/projects/dynamic-import.

Further reading

See the importlib and getattr documentation for more details.