How to build Python-C++ libraries

on YouTube ↗49m

" summary

Explains how to produce redistributable Python wheels for C++ extensions by using scikit-build + CMake to control compiler flags and dependencies, CI (Travis/Circle/AppVeyor) to build across multiple OSes and Python versions, and auditwheel/multi-build to bundle transitive native libs and create manylinux-compatible wheels. Shows a minimal C++ math extension, automates rebuilds, runs ~11 CI jobs per change, and performs a one‑click deploy to PyPI triggered by a git tag — yielding reproducible, tested wheels that install on target platforms.

AI-written summary · about

" chapters (11)

1. 0:00Introduction to pip packaging
2. 2:15Understanding package structure
3. 7:22The history of packaging tools
4. 9:45Creating C extensions
5. 13:05Challenges of native builds
6. 21:51Introduction to scikit-build
7. 27:44Automating build workflows
8. 31:58Testing with CI services
9. 36:21Distributing wheels
10. 39:53Cross-platform automation
11. 44:26Conclusion and Q&A

" related talks

• 40mBuilding a binary extensionHenry Fredrick Schreiner III · PyCon US 2022
• 24mA tale of Python C extensions and cross-platform wheelsVinayak Mehta · PyConline AU 2021
• 24mThe wheelhouse of horrorsLaszlo Kiss Kollar · PyCon US 2023
• 37mScikit build: A buildJean Christophe Fillion Robin, Matt McCormick, Max Smolens, Omar Padron · PyData Carolinas 2016
• 36mIntroduction to C++ python extensions and embedding Python in C++ AppsDiego Rodriguez-Losada · CppCon 2016
• 29mGenerating Python bindings to C++ librariesDavid Vo · PyCon AU 2023
• 25mHow to bring your C++ project to Python landSharon Woods · Qt World Summit 2019
• 01h 01mExtend and Embed Python with C++Rainer Grimm · Meeting C++ 2022