IATI is a living project with a number of tools that are owned by the IATI Secretariat. We are an open-source project and we encourage code contributions.
Please follow this guide if you wish your pull requests to be reviewed by us. This guide was created to ensure your contributions have a meaningful long-term impact on the IATI code base by ensuring your code can be maintained over the long-term. It is also beneficial to open a Github Issue about any significant features you may wish to add as we reserve the right to reject pull requests that clash with, or otherwise hinder, existing or ongoing work.
This guide is inspired by the React contribution guides.
It is important to remember that IATI is still a relatively young project and so our current code base has varying standards of code quality and style. We are continuously working towards improving this situation. With this in mind, we ask that you use our pyIATI work as the expected standard to work towards.
pyIATI is part of our current work to improve the stability and maintainability of our infrastructure. We believe it exemplifies our move towards stronger architecture and adoption of industry-standard best practices. Because of this, we wish to be consistent with the coding style used with throughout this project whenever possible.
Generally speaking we expect all contributions to adhere to the following conventions:
- Human readable code:
- Write your code clearly with descriptive variable names.
- Write your code with the intention that any developer could read it.
- You are contributing to a community - write your code for them, not just for you.
- Robustly tested code:
- All core functionality is to be unit tested with py.test and edge cases considered. We strongly suggest you use Test Driven Development. We also require details and evidence of any manual testing to show no existing functionality is unexpectedly broken.
- Documented code:
- We expect a clear docstring per module, class, and function, explaining what it does at a minimum. This is to reduce both developer onboarding time and the barriers to entry for new developers who want to get involved. They should be formatted as per Google-style Sphinx Docstrings.
- No code smells:
- Use pyIATI wherever possible:
- We have built this library with the intention of reducing the need to reinvent the wheel. We want to improve the time it takes for external developers to contribute to our code base. As time goes on pyIATI will have increased functionality for many common tasks needed to interface with IATI data. Keep an eye on the changelog for updates on what pyIATI can do and planned improvements.
We run these locally in editors and as git-hooks to prevent commits from being made when the linters do not pass. These linters also run as part of our automated tests to aid contributors with their linting. We expect all code to pass the linters before we review it.
You may add in-line linting exceptions in these scenarios:
- Where linting rules obscure function or variable names clarity should be prioritised over linting.
- Docstrings are not expected to adhere to low line length limits. Please soft-wrap docstrings in your code editor if you don’t like long lines. Hard-wrapping must never be undertaken mid-sentence.
- What is good code?
- Writing good unit tests:
- Test Driven Development:
- Refactoring unruly code:
Last updated on 2018-02-19