How Python Runs Your Code¶

Your Code (e.g., `example.py`):¶

A Python file that contains your source code, like:

Python
print("Hello World")

Interpreter / Part of CPython Implementation:¶

The interpreter is the program that runs your Python code.
CPython is the default and most widely used implementation of Python, written in C. The interpreter, a key part of CPython, executes bytecode, whether it’s from a .pyc file or generated in memory. When you install Python from python.org, you get the CPython implementation.

Lexing & Parsing:¶

Before running the code, Python analyzes it in two steps to understand its structure:
- Lexing: Breaks code into tokens (e.g., print, "Hello World").
- Parsing: Arranges these tokens into a tree-like structure called an AST (Abstract Syntax Tree), which represents the logic and flow of your code.

Bytecode and Compilation¶

The AST is then compiled into bytecode, a low-level, platform-independent set of instructions. This bytecode is generated in memory and executed immediately by the Python Virtual Machine (PVM).

For modules that are imported into the main script, the bytecode is saved to disk in a .pyc file inside the __pycache__ directory. This caching speeds up future imports by avoiding the need to recompile the module every time it is loaded.

🔹 Note: The .pyc file is only written in disk when a module (e.g, .py file) is imported, Running the main script directly (e.g, without any imported .py file) only uses in-memory bytecode.

🔹 Bytecode is not machine code—it’s an intermediate representation that Python understands.

Python Virtual Machine (PVM):¶

The PVM is the runtime engine that reads and executes the bytecode instruction by instruction. It’s a core part of the Python interpreter.
It interacts with the OS, manages memory, handles I/O, and displays output (like "Hello World" on the screen).

Compiler & PVM Are Both Parts of the Python Interpreter¶

The Python interpreter consists of two essential components:
- 🔸 A compiler that translates .py files into bytecode.
- 🔸 The PVM that executes that bytecode.

This Is the Flow of the Whole Process¶

Your code → Lexer & Parser → AST → Compiler → Bytecode → PVM → Output

What About JIT?¶

CPython doesn’t use JIT (Just-In-Time compilation) by default; it interprets bytecode directly. Other implementations like PyPy use JIT for better performance.
However, CPython is experimenting with JIT! In CPython 3.13, a JIT compiler is being developed, but as of April 7, 2025, it’s experimental and disabled by default. Learn more here: What docs say about JIT.

CPython vs. JIT: How They Work¶

CPython compiles Python source code into bytecode once, then interprets it step by step using the Python Virtual Machine (PVM). It’s a slow process no machine code is made directly.

JIT compiles code into machine code during runtime, and reusing the most used parts of the program.

CPython keeps things simple; JIT focuses on speed.

Diagram¶

Diagram