Contribution Guidelines#

Label Sleuth is an open source project committed to bringing NLP model creation to domain experts. The guide below provides information on how you can join the Label Sleuth community in this goal.

How can I contribute?#

There are multiple ways to contribute to the project:

Reporting bugs#

To report a bug, please open a bug issue in the issue tracker. To make it easier for the community to address the issue, please provide a detailed description of it and include steps to reproduce it.

Requesting new features#

If you have an idea about an enhancement or new feature, please open a feature request/enhancement issue in the issue tracker.

Contributing code#

Working on Issues#

While not required, it is good practice to ensure before you start working on a code change that what you will be working on is captured in an issue. This allows the community to be aware of what everyone is working on and provide input before you spend time on the actual change.

Prerequisites#

Before contributing to Label Sleuth, you should make sure you have the following tools installed:

  • Node.js v16 or above. You can install it directly or through a package manager.

    • If you are on macOS, we recommend using nvm to help manage different versions of Node.js

  • Git

  • Anaconda

Note: Node.js is only required if you contributing code to the frontend of Label Sleuth.

1. Fork the repo#

Go to Label Sleuth’s GitHub repo and click the Fork button on the top-right corner. This will create a copy of the repo associated with your account.

2. Clone your fork#

  1. Click on [your_github_username]/label-sleuth.

  2. Click on the Clone or Download button and copy the URL from the Clone with SSH option. It should start with git@github.com...

In your terminal, run:

git clone git@github.com:[your_github_username]/label-sleuth.git
cd label-sleuth

See GitHub docs for more details.

3. Add upstream remotes#

When you clone your forked repo, running git remote -v will show that the origin is pointing to your forked repo by default.

Now you also need to add the label-sleuth/label-sleuth repo as your upstream remote branch:

# Add the upstream remote to your repo
git remote add upstream git@github.com:label-sleuth/label-sleuth.git

# Verify that the remote was added
git remote -v

Your terminal should output something like this:

origin  [your forked repo] (fetch)
origin  [your forked repo] (push)
upstream    git@github.com:label-sleuth/label-sleuth.git (fetch)
upstream    git@github.com:label-sleuth/label-sleuth.git (push)

4. Work in a branch#

When contributing to Label Sleuth, your work should always be done in a separate branch. This is also how you will be submitting your pull request when your work is done.

To create a new branch, ensure you are on your forked branch in your terminal and run:

git pull origin main
git checkout -b {your-branch-name}

5. Build and start the development environment#

From the root directory of your fork, run:

#To install backend dependencies
conda create --yes -n label-sleuth python=3.8
conda activate label-sleuth
pip install label-sleuth

#To run the backend 
python -m label_sleuth.start_label_sleuth


#To install frontend dependencies
cd frontend
npm install

#To run the frontend in development mode
npm start

The last version of the frontend is statically served by the backend, so if you are not making any frontend contributions, the frontend-related commands above are not required.

6. Test the application#

To test the backend, run:

python -m unittest

To test the frontend, run:

npm start --prefix frontend

6. Create a Pull Request#

We use GitHub pull requests to accept contributions. Once submitted, the pull request will be reviewed by the community as explained below.

Developer Certificate of Origin#

Before contributing to the project, all contributors must sign a Developer Certificate of Origin (DCO). By signing the DCO, you are attesting that you are the author of the contribution and you are freely contributing it under the terms of the Apache-2.0 license.

To sign the DCO, include a copy of the Developer’s Certificate of Origin 1.1 in a pull request comment.

Code Review#

Code review of pull requests is open to anyone. While only maintainers can merge pull requests, community feedback on pull requests is very valuable. It can also be helpful in becoming acquainted with the codebase. You can find a list of open pull requests here.

Asking for Help/Feedback#

If you are working on the code and need help finishing it or would want to get input from the community on the approach that you are following, you can create a Work In Progress pull request. To this end, when creating the pull request, select ‘Create draft pull request’ by clicking on the arrow next to ‘Create pull request’. This will inform the reviewer that the code is not final. It also means that the pull request will not be merged. When the pull request is ready to be reviewed for merging, the reviewer or you can mark it as “Ready for review”.

9. Updating a pull request#

Stay up to date with the activity in your pull request. Contributors of Label Sleuth will be reviewing your work and making comments, asking questions and suggesting changes to be made before they merge your code.

Your branch has to be updated with the main branch before merging it. Thus, anytime new changes are added to the main branch, you will have to update the branch you are working on. To do so, run git merge origin main or git pull --rebase origin main. The former will include a merge commit while the latter will re-write history by adding the new commits of the main branch to your branch, while maintaining a cleaner and linear commit history. If you think that your changes should be added into main without being squashed into a single commit, use the rebase approach.

Once all revisions to your pull request are complete, a maintainer of Label Sleuth will merge your commits for you.

Good First Issues#

If you want to contribute to Label Sleuth and do not know where to start, take a look at the issues tagged with the good first issue label. These have been reviewed by other contributors and identified as self-contained issues that are suitable for a first contribution. You can find a list of currently open issues with this label here.