Generates pip requirements.txt file for any project by analysing package imports
Project description
rqmts
rqmts - Generate pip requirements.txt file for any project by analysing package imports
Click here to see the demo.
Click here for documentation.
About the project
rqmts is a fantastic stand-alone tool which generates pip requirements.txt
file for any project by analysing package imports.
It does not requires any dependency (works out-of-the-box), not needs internet to work (is completely offline, upto this moment), nor uses regular expressions in such a violent way as existing projects do. Instead, it uses simple heuristic techniques and parse conditional trees, which is a better method for extracting imported names from statements, functions, etc.
Why this project
Questions
- Why not just use pip’s freeze command to generate a
requirements.txt
file for my project ? - Why to re-invent the wheel when there are modules such as pipreqs, pigar already present ?
- Why not manually ?
Answers
- Why not just pip freeze?
pip freeze
only saves the packages that are installed withpip install
in your environment.pip freeze
saves all packages in the environment including those that you don't use in your current project. (if you don't have virtualenv)
- Why re-invent the wheel ?
- pipreqs fails on many occasions (see - pipreqs/issues)
- I found this repository and thought, "Hmm.. I think I can simply this problem while trying to match pipreqs results"
- pigar queries pypi servers, big no-no. Ideally, it should be local. (on fallback? then maybe ..)
- Other than that, pigar recommends using Pipenv (pipenv has serious issues)
- Sheer curiousity. "I wonder if I can create a project that has potential of thosands of stars and most importantly, hundreds of contributors under 24 hours"
- Manually ?
- Are you serious right now ?
Installation
pip install --user rqmts
Contribute
The major challenge of this project is to extract the required metadata from modules which are first extracted from the input script.
Challenges
- Version numbers in python can be in very different places depending on the case
- Package name in the package index is independent of the module name we import
and these quirks make this project interesting. There's a funny comment in the source which reflects the diversity between us and it goes like :
# module_name.__version__ sucks, because we suck (PEP 0396)
This project aims to combine the best existing strategies to cover the broadest possible set of cases (if not all). The project was built keeping in mind the modular programming paradigms and so other than being readable it's easily extensible making it possible to add new strategies/algorithms quickly.
If you have any issues or suggestions, please do not hesitate to open an issue or a pull request!
License
This software is licensed under BSD 3-Clause "New" or "Revised" License. To view a copy of this license, visit BSD 3-Clause.
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.