Apps Artificial Intelligence CSS DevOps Go JavaScript Laravel Linux MongoDB MySQL PHP Python Rust Svelte Vue

Understanding `__init__.py` in Python

1 min read .
Updated on
Understanding `__init__.py` in Python

The __init__.py file is commonly used to define a regular Python package and control what happens when that package is imported. It can be empty, expose a package-level API, define metadata, or run lightweight initialization code.

1. A Basic Package

Consider this structure:

my_package/
├── __init__.py
├── module1.py
└── module2.py

With __init__.py present, my_package is a regular package and its modules can be imported normally:

from my_package import module1
from my_package.module2 import some_function

Since Python 3.3, namespace packages can exist without __init__.py, so the file is no longer required for every possible package layout. It is still the standard choice for ordinary application and library packages.

2. Re-export a Public API

Suppose module1.py and module2.py contain useful functions. You can expose them at the package level:

# my_package/__init__.py
from .module1 import some_function
from .module2 import another_function

Callers can then write:

from my_package import some_function, another_function

This can provide a simpler and more stable public API even if the internal module structure changes later.

3. What __all__ Does

You may define:

from .module1 import some_function
from .module2 import another_function

__all__ = ["some_function", "another_function"]

__all__ primarily controls which names are imported by:

from my_package import *

It can also communicate the intended public API to readers and tooling, but it does not make other attributes truly private.

4. Package Metadata

Small constants can live in __init__.py:

__version__ = "1.0.0"

For installed package versions, modern projects may instead read metadata through importlib.metadata so the version has a single source of truth.

5. Avoid Heavy Import-Time Work

Code in __init__.py runs when the package is imported. Avoid slow network requests, large database connections, expensive filesystem scans, or other surprising side effects there.

Prefer lightweight imports and definitions:

from .client import Client

__all__ = ["Client"]

6. Relative Imports

Inside a package, a leading dot refers to the current package:

from .module1 import some_function

Two dots refer to the parent package:

from ..shared import helper

Use relative imports deliberately; absolute imports are often easier to understand across larger codebases.

Conclusion

__init__.py remains an important part of conventional Python packages. Keep it lightweight, use it to expose a clear package-level API when helpful, and remember that namespace packages are a separate mechanism that can work without the file.

Python

Related Posts

Catching Multiple Exceptions in Python
Updated on
Catching Multiple Exceptions in Python
Python lets one except clause handle several exception types when they should receive the same response. This keeps error handling concise without hiding unrelated failures. Catch Several Exception Types Place the exception classes in a tuple: Copy try: value = int(user_input) result = 100 / value except (ValueError, ZeroDivisionError) as exc: print(f"Invalid input: {exc}") ValueError handles non-numeric input, while ZeroDivisionError handles zero.
Python
Check Whether a Key Exists in a Python Dictionary
Updated on
Check Whether a Key Exists in a Python Dictionary
Python dictionaries provide fast key-based lookup, and checking whether a key exists is a common operation. Use the in Operator The clearest and most idiomatic solution is: Copy my_dict = {"name": "Alice", "age": 30, "city": "New York"} if "name" in my_dict: print("The key 'name' exists.") else: print("The key 'name' does not exist.") Membership tests on a dictionary check keys by default.
Python Dictionary
How to Check for a Specific Key in a Python Dictionary
Updated on
How to Check for a Specific Key in a Python Dictionary
Python dictionaries store key-value pairs and provide efficient key lookup. Before reading an optional key, you may want to check whether it exists. 1. Use in The recommended approach is the membership operator: Copy my_dict = {"name": "Alice", "age": 30, "city": "New York"} if "name" in my_dict: print("The key exists.") else: print("The key does not exist.") This checks keys directly and clearly expresses the intent.
Python
How to Flatten a List of Lists in Python
Updated on
How to Flatten a List of Lists in Python
Python lists can contain other lists, which is useful for representing grouped or nested data. When you need a single sequence instead, you can flatten the nested structure in several ways. 1. Flatten One Level with a List Comprehension For a list where every top-level item is another list: Copy nested_list = [[1, 2, 3], [4, 5], [6, 7, 8]] flat_list = [item for sublist in nested_list for item in sublist] print(flat_list) Output:
Python
How to List Files in a Directory with Python
Updated on
How to List Files in a Directory with Python
Listing files in a directory is a common Python task for automation, data processing, uploads, and filesystem utilities. The standard library provides several good approaches. 1. os.listdir() os.listdir() returns names for both files and subdirectories, so filter the results when you need only regular files: Copy import os def list_files(directory): try: entries = os.listdir(directory) return [ name for name in entries if os.path.isfile(os.path.join(directory, name)) ] except FileNotFoundError: return [] 2. os.scandir() os.scandir() returns DirEntry objects that can expose file-type information efficiently:
Python
How to Use a Global Variable Inside a Python Function
Updated on
How to Use a Global Variable Inside a Python Function
Python variables follow scope rules that determine where a name can be read or assigned. A module-level variable can be read inside a function, but assigning to that same name requires special handling. Reading a Global Variable A function can read a module-level variable without the global keyword: Copy app_name = "Nalar" def print_app_name(): print(app_name) print_app_name() Because the function does not assign to app_name, Python resolves the name from the enclosing module scope.
Python
chevron-up