mypy

mypy
www.python.org · GPL · source
Name	mypy
Developer	Dmitry Vasilevich?; Python Software Foundation?
Released	2012
Programming language	Python
Operating system	Cross-platform
License	MIT

mypy mypy is a static type checker for the Python programming language that performs optional compile-time type checking to help developers find type-related errors before runtime. It complements projects using Python implementations and tooling from organizations such as the Python Software Foundation, interacts with package ecosystems like PyPI, and is often used alongside editors and platforms such as Visual Studio Code, JetBrains products, and GitHub continuous integration. The project has influenced typing discussions involving standards bodies and conferences including Python Enhancement Proposal processes and presentations at events like PyCon.

Overview

mypy provides static analysis that analyses annotated Python code written against libraries and frameworks such as Django, Flask, NumPy, Pandas, and TensorFlow. It integrates with build and automation tools like Travis CI, CircleCI, GitLab CI/CD, and Jenkins to guard code quality in repositories hosted on platforms such as GitHub, GitLab, and Bitbucket. Contributors and maintainers often coordinate via communication channels including GitHub Issues, Gerrit-style workflows, and mailing lists associated with the Python Software Foundation community.

Features and Design

The tool supports gradual typing that interoperates with core Python language machinery as specified in PEP 484 and related proposals such as PEP 526 and PEP 561. It implements features like type annotations for functions, classes, and variables tailored to patterns used in libraries including asyncio, SQLAlchemy, and Requests. Design choices reflect influences from statically typed languages and systems such as Haskell, TypeScript, OCaml, and Java, while maintaining compatibility with runtime semantics from implementations like CPython, PyPy, and MicroPython. The architecture exposes command-line interfaces and APIs that facilitate integration with tools such as flake8, black (software), isort, and language server implementations like Language Server Protocol-aware servers used by Visual Studio Code and Sublime Text.

Usage and Integration

Developers adopt the checker in projects ranging from web applications built with Django and Flask to data science pipelines using NumPy, Pandas, and SciPy. It is applied in continuous integration pipelines on services like Travis CI and GitHub Actions and in corporate environments using platforms such as Azure DevOps and Google Cloud Build. IDE and editor integrations target environments provided by JetBrains (for example PyCharm), Visual Studio Code, and Emacs with packages maintained by contributors from communities tied to organizations such as Mozilla and Red Hat. Commercial users often combine it with static analysis suites from vendors like SonarSource and security platforms referencing guidelines from OWASP.

Type System and Type Checking

The checker implements a form of structural and nominal typing that uses typing constructs standardized in PEP 484 and enhanced by proposals like PEP 544 and PEP 563. It supports generics, protocols, type variables, union types, and callable signatures to model APIs from packages such as Django REST framework, FastAPI, and aiohttp. Type stubs and stub packages follow conventions used in typeshed and community-maintained repositories that mirror types for libraries including stdlib modules and popular third-party packages on PyPI. Error reporting and gradual enforcement strategies are designed to fit workflows used by teams at companies such as Google, Dropbox, Microsoft, and Facebook.

Development History

Origins trace to academic and industry influences around static typing research and pragmatic typing adoption driven by engineers associated with projects at organizations like Dropbox and contributors active within the Python Software Foundation ecosystem. The roadmap has been shaped through discussions in bug trackers, proposal threads on mailing lists, and talks at conferences like PyCon, EuroPython, and ICSE. Over time the project incorporated community-contributed features, compatibility layers for Python language changes introduced by versions like Python 3.5, Python 3.6, and later minor releases, and interoperability with packaging conventions promoted by PyPA.

Adoption and Ecosystem

Adoption spans open-source projects, academic codebases, and enterprise code maintained by companies such as Dropbox, Microsoft, Google, Spotify, and Instagram. The ecosystem includes type stub repositories, linters, formatters, language servers, and third-party extensions developed by contributors associated with foundations and firms like Python Software Foundation, NumFOCUS, and various university research groups. Training resources and community support appear in tutorials, books, and workshops presented at venues including PyCon, Strata Data Conference, and university courses taught at institutions such as MIT, Stanford University, and UC Berkeley.

Category:Python (programming language) tools