Goals and Architecture Overview

This document gives an overview of the goals and architecture of PyPy. If you’re interested in using PyPy or hacking on it, have a look at our getting started section.

Mission statement

We aim to provide a compliant, flexible and fast implementation of the Python Language which uses the RPython toolchain to enable new advanced high-level features without having to encode the low-level details. We call this PyPy.

High Level Goals

Our main motivation for developing the translation framework is to provide a full featured, customizable, fast and very compliant Python implementation, working on and interacting with a large variety of platforms and allowing the quick introduction of new advanced language features.

This Python implementation is written in RPython as a relatively simple interpreter, in some respects easier to understand than CPython, the C reference implementation of Python. We are using its high level and flexibility to quickly experiment with features or implementation techniques in ways that would, in a traditional approach, require pervasive changes to the source code. For example, PyPy’s Python interpreter can optionally provide lazily computed objects - a small extension that would require global changes in CPython. Another example is the garbage collection technique: changing CPython to use a garbage collector not based on reference counting would be a major undertaking, whereas in PyPy it is an issue localized in the translation framework, and fully orthogonal to the interpreter source code.

PyPy Python Interpreter

PyPy’s Python Interpreter is written in RPython and implements the full Python language. This interpreter very closely emulates the behavior of CPython. It contains the following key components:

  • a bytecode compiler responsible for producing Python code objects from the source code of a user application;
  • a bytecode evaluator responsible for interpreting Python code objects;
  • a standard object space, responsible for creating and manipulating the Python objects seen by the application.

The bytecode compiler is the preprocessing phase that produces a compact bytecode format via a chain of flexible passes (tokenizer, lexer, parser, abstract syntax tree builder, bytecode generator). The bytecode evaluator interprets this bytecode. It does most of its work by delegating all actual manipulations of user objects to the object space. The latter can be thought of as the library of built-in types. It defines the implementation of the user objects, like integers and lists, as well as the operations between them, like addition or truth-value-testing.

This division between bytecode evaluator and object space gives a lot of flexibility. One can plug in different object spaces to get different or enriched behaviours of the Python objects.



RPython is the language in which we write interpreters. Not the entire PyPy project is written in RPython, only the parts that are compiled in the translation process. The interesting point is that RPython has no parser, it’s compiled from the live python objects, which makes it possible to do all kinds of metaprogramming during import time. In short, Python is a meta programming language for RPython.

The RPython standard library is to be found in the rlib subdirectory.

Consult Getting Started with RPython for further reading


The translation toolchain - this is the part that takes care of translating RPython to flow graphs and then to C. There is more in the architecture document written about it.

It lives in the rpython directory: flowspace, annotator and rtyper.

PyPy Interpreter

This is in the pypy directory. pypy/interpreter is a standard interpreter for Python written in RPython. The fact that it is RPython is not apparent at first. Built-in modules are written in pypy/module/*. Some modules that CPython implements in C are simply written in pure Python; they are in the top-level lib_pypy directory. The standard library of Python (with a few changes to accomodate PyPy) is in lib-python.

JIT Compiler

Just-in-Time Compiler (JIT): we have a tracing JIT that traces the interpreter written in RPython, rather than the user program that it interprets. As a result it applies to any interpreter, i.e. any language. But getting it to work correctly is not trivial: it requires a small number of precise “hints” and possibly some small refactorings of the interpreter. The JIT itself also has several almost-independent parts: the tracer itself in rpython/jit/metainterp, the optimizer in rpython/jit/metainterp/optimizer that optimizes a list of residual operations, and the backend in rpython/jit/backend/<machine-name> that turns it into machine code. Writing a new backend is a traditional way to get into the project.

Garbage Collectors

Garbage Collectors (GC): as you may notice if you are used to CPython’s C code, there are no Py_INCREF/Py_DECREF equivalents in RPython code. Garbage Collection in RPython is inserted during translation. Moreover, this is not reference counting; it is a real GC written as more RPython code. The best one we have so far is in rpython/memory/gc/incminimark.py.