Comparison of Packaging Tools in 2023 - Peacock

歡迎來到 PyCon TW 2023 共筆

共筆入口：https://hackmd.io/@pycontw/2023
手機版請點選上方 按鈕展開議程列表。
Welcome to PyCon TW 2023 Collaborative Writing
Collaborative Writing Workplace：https://hackmd.io/@pycontw/2023
Using mobile please tap to unfold the agenda.

Collaborative writing start from below
從這裡開始共筆

Slide: https://slides.p3ac0ck.net/pycontw2023/index.html

• PEPs: 517, 621, and 660
• pyproject.toml format was defined

pyproject.toml https://pip.pypa.io/en/stable/reference/build-system/pyproject-toml/

pyproject.toml format

• build-system
• project
  • dependencies

Pipenv

• Pros
  • supports lock-file
  • easy to add/ upgrade pacakges
  • wraps vitualenv
• Cons
  • slow dependency resolver
  • limited support for pep 621 style pyproject.toml

Poetry

• Pros
  • Dependency management with lock-file
  • Wrapper virtual env
  • Include Task-runner poetry run
  • Helper for the building wheels
• Cons
  • Not support PEP621 style pyproject.toml

PDM

• Pros
  • Fast dependency resolver (you can choose another resolver)
  • Support PEP 621
  • Include Task runner (pdm run)
• Cons
  • Not include wheel building helper

pip-tools

• Pros
  • Simple, only 2 commands
  • Support PEP 621
    • Combine-able with pip or hatch
  • Maintained by the jazzband community
• Cons
  • does not included task runner or wheel building tools

Hatch

• Pros
  • Configurable to a backend for builing project
  • Maintained acitvely by PyPA nearly official
  • Support PEP 621
• Cons

pip

• Pros
  • Build-in in Python, very simple
  • Support PEP 621
• Cons
  • cannot wrap virtualenv
  • does not support file-managed dependencies
  • does not include task-runnder and wheel builder

Summary

Case 1: Library developing

• Hatch (with pip-tools): Very configurable and flexible

Case 2: Application Developing

• PDM: All-in-one tool and easy to use
• resaons:
  • fast dependency resolver
  • supports PEP 621 style pyproject.toml

Case 3: Automation Script Developing

• pip Simple to use, not built-in in Python)
• reasons:
  • Lightweight, non-dep
  • Simple to use

Appendix

Rye

• pros
  • single-binary, easy to install and use
  • works with pyproject.toml spec PEP 621 style
  • wraps virtualenv, including package builder and publisher
  • inlcude python version manager
• cons
  • experimental, not stable

Q & A

1. is requirement.txt dead?

• no, but it has some problems

Below is the part that speaker updated the talk/tutorial after speech
講者於演講後有更新或勘誤投影片的部份