Skip to content

Latest commit

 

History

History
221 lines (158 loc) · 6.48 KB

CONTRIBUTING.md

File metadata and controls

221 lines (158 loc) · 6.48 KB
Raw

Contributing

There is more than one way of contributing to GRASS GIS. Here we will focus on contributions centered around the main GRASS GIS source code. You can also report issues, plan new features, or explore https://grass.osgeo.org/get-involved/.

Changing code and documentation

This guide covers contributing to the main version of GRASS GIS source code which is in the branch called main. It assumes that you have some very basic knowledge of Git and GitHub, but if you don't just go through some tutorial online or ask on the GRASS GIS developer mailing list.

First time setup

  • Create an account on GitHub.
  • Install Git on your computer.
  • Set up Git with your name and email.
  • Fork the repository (by clicking the Fork button in the upper right corner of the GitHub interface).
  • Clone your fork (use HTTPS or SSH URL, here we will use HTTPS):
git clone git@github.com:your_GH_account/grass.git
  • Enter the directory
cd grass/
  • Add main GRASS GIS repository as "upstream" (use HTTPS URL):
git remote add upstream https://github.com/OSGeo/grass
  • Your remotes now should be "origin" which is your fork and "upstream" which is this main GRASS GIS repository. You can confirm that using:
git remote -v
  • You should see something like:
origin  git@github.com:your_GH_account/grass.git (fetch)
origin  git@github.com:your_GH_account/grass.git (push)

For the following workflow, it is important that "upstream" points to the OSGeo/grass repository and "origin" to your fork (although generally, the naming is up to you).

Update before creating a feature branch

  • Make sure your are using the main branch to create the new branch:
git checkout main
  • Download updates from all branches from the upstream remote:
git fetch upstream
  • Update your local main branch to match the main branch in the upstream repository:
git rebase upstream/main

Notably, you should not make commits to your local main branch, so the above is then just a simple update (and no actual rebase or merge happens).

Update if you have local changes

If rebase fails with "error: cannot rebase: You have unstaged changes...", then move your uncommitted local changes to "stash" using:

git stash
  • Now you can rebase:
git rebase upstream/main
  • Get the changes back from stash:
git stash pop

Creating a new feature branch

Now you have updated your local main branch, you can create a feature branch based on it.

  • Create a new feature branch and switch to it:
git checkout -b new-feature

Making changes

You can use your favorite tools to change source code or other files in the local copy of the code. When making changes, please follow the Submitting Guidelines.

Testing changes

Testing helps to ensure that the changes work well with the rest of the project. While there are many different ways to test, usually you will want to compile the source code (see below), add test code (using grass.gunittest or pytest), and run code linters (automated code quality checks).

There is a series of automated checks which will run on your pull request after you create one. You don't need to run all these checks locally and, indeed, some of them may fail for your code. This is a part of the standard iterative process of integrating changes into the main code, so if that happens, just see the error messages, go back to your code and try again. If you are not sure what to do, let others know in a pull request comment.

Note that there are some steps you can do locally to improve your code. For Python, run black . to apply standardized formatting. You can also run linter tools such as Pylint which will suggest different improvements to your code.

Committing

  • Add files to the commit (changed ones or new ones):
git add file1
git add file2
  • Commit the change (first word is the module name):
git commit -m "module: added a new feature"

Pushing changes to GitHub

  • Push your local feature branch to your fork:
git push origin new-feature

Pull request

When you push, GitHub will respond back in the command line to tell you what URL to use to create a pull request. You can follow that URL or you can go any time later to your fork on GitHub, display the branch new-feature, and GitHub will show you a button to create a pull request.

Alternatively, you can explore GitHub CLI tool (gh) which allows you to do git push and create a pull request in one step with gh pr create -fw.

After creating a pull request

GRASS GIS maintainers will now review your pull request. If needed, the maintainers will work with you to improve your changes.

Once the changes in the pull request are ready to be accepted, the maintainers will usually squash all your commits into one commit and merge it to the main branch.

Once the pull request is merged, it is a good time to update your local main branch in order to get the change you just contributed.

Further notes

GRASS GIS maintainers use additional workflows besides the one described above. These are detailed at https://trac.osgeo.org/grass/wiki/HowToGit

Compilation

More often than not, in order to test the changes, you need to create a runnable binary program from the source code, using the so-called "compilation step". While the source code consists of thousands of C and Python files (plus HTML documentation and other files), the included "makefiles" tell the build system to generate binaries from the source code in the correct order, render the manual pages, etc.

The way to install the compiler tools and Python depends on the operating system. To make this easier, we have collected copy-paste instructions to install dependencies and compile GRASS source code for most operating systems. Please see our dedicated wiki:

Compile and install instructions

About source code

GRASS GIS is written in more than one programming language, but you need to know only the language relevant to your contribution. While much of the source code is written in C, a significant portion is written in Python. A compiler is needed to convert the C and C++ source code into executable files ("binaries"). In contrast, Python is an interpreted language that can only be executed with Python software. There is also documentation in HTML files and other files in the GRASS GIS source code.