Upgrading to Python 3.12: Essential Syntax and Performance Changes

Prerequisites

To get the most out of this guide, you should have a solid grasp of Python 3.11 fundamentals, including list comprehensions, basic type hinting, and object-oriented programming. Familiarity with the Global Interpreter Lock will help you understand the experimental threading improvements.

Key Libraries & Tools

• pathlib: Updated with a new walk() method for easier directory navigation.
• typing: Updated to support new syntax for generics and the @override decorator.
• asyncio: Now the primary package for concurrent operations as older modules are retired.

Code Walkthrough

Simplified Generics

Previously, defining generic classes required manual TypeVar declarations. PEP 695 introduces a cleaner, inline syntax.

# The Old Way
from typing import TypeVar, Generic
T = TypeVar("T", bound=str)
class Container(Generic[T]):
    ...

# The Python 3.12 Way
class Container[T: str]:
    def get_data() -> T:
        ...

This new syntax eliminates the need for separate variable assignments and makes generic definitions look more like other modern languages.

The Override Decorator

To prevent accidental errors when subclassing, use the @override decorator to ensure you are actually modifying a base class method.

from typing import override

class Base:
    def get_color(self):
        return "blue"

class Child(Base):
    @override
    def get_color(self):  # Type checkers will error if this is misspelled
        return "red"

Syntax Notes

f-strings are now fully formalized using the new PEG parser. You can now nest double quotes within f-strings without switching to single quotes, allowing for complex expressions like f"{strings["key"]}" directly within the template.

Practical Examples

Use the new pathlib.Path.walk() to recursively process files without relying on the older os.walk interface. This keeps your code consistent within the pathlib ecosystem. Additionally, utilize Immortal Objects for global constants; because their reference counts never change, they avoid cache invalidation in multi-threaded environments.

Tips & Gotchas

Watch out for removed modules! Python 3.12 officially deletes distutils, asyncore, and asynchat. If your project still relies on these, migrate to setuptools or asyncio immediately. Also, note that while per-interpreter Global Interpreter Lock support exists in the C-API, you won't see the full Python-level interface until 3.13.