We ♥ contributors! By participating in this project, you agree to abide by the Ruby for Good code of conduct.
If you're new here, here are some things you should know:
- A great introductory overview of the application is available at the wiki.
- Issues tagged "Help Wanted" are self-contained and great for new contributors
- Pull Requests are reviewed within a week or so
- Ensure your build passes linting and tests and addresses the issue requirements
- This project relies entirely on volunteers, so please be patient with communication
If you have any questions about an issue, comment on the issue, open a new issue, or ask in the RubyForGood slack. human-essentials has a #human-essentials
channel in the Slack. Our channel in slack also contains a zoom link for office hours every day office hours are held.
Many helpful members are available to answer your questions. Just ask, and someone will be there to help you!
You won't be yelled at for giving your best effort. The worst that can happen is that you'll be politely asked to change something. We appreciate any sort of contributions, and don't want a wall of rules to get in the way of that.
Follow documentation from Microsoft for enabling and installing Windows Subsystem For Linux 2 on your machine.
Make sure to install Ubuntu as your Linux distribution. (This should be default.)
Note: If you run into any issues with a command not running, restart your machine.
- Install Ruby
- Install the version specified in
.ruby-version
. - Visit the Install Ruby on Rails guide by GoRails for Ubuntu, Windows, and macOSX setup.
⚠️ Follow only the Installing Ruby step, as our project setup differs⚠️ It is highly recommended you use a ruby version manager such as rbenv, asdf, or rvm. - Verify that your Ruby installation works by running
ruby -v
.
- Install the version specified in
- Install Postgres
- Follow one of these guides: MacOSX, Ubuntu.
- Create a
database.yml
file onconfig/
directory with your database configurations. You can also copy the existing files calleddatabase.yml.example
and.env.example
and change the credentials.
- Clone the project and switch to its directory
- Run
bin/setup
- Run
bin/start
and visit http://localhost:3000/ to see the human essentials page. - Log in as a sample user with the default credentials.
These credentials also work for staging:
Super Users 🦸🏽♀️
username: [email protected]
password: password!
Bank Users 🏦
Organization Admin
Email: [email protected]
Password: password!
User
Email: [email protected]
Password: password!
Partner Users 👥
Verified Partner
Email: [email protected]
Password: password!
Invited Partner
Email: [email protected]
Password: password!
Unverified Partner
Email: [email protected]
Password: password!
Recertification Required Partner
Email: [email protected]
Password: password!
Waiting Approval Partner
Email: [email protected]
Password: password!
Another approved partner (with all groups):
Email: [email protected]
Pasword: password!
- Create the container:
- To run the container on a Github VM, follow the Codespace link above. You can connect to the Codespace using VSCode or the VSCode web editor.
- Or follow instructions to create a new Codespace.
- To clone this repo and run the container locally, follow instructions to install VSCode and Docker. Click the Dev Container link above. Don't forget to add a git remote pointing to your fork once the container is setup and you want to push changes.
- Wait for the container to start. This will take a few (10-15) minutes since Ruby needs to be installed, the database needs to be created, and the
bin/setup
script needs to run - Run
bin/start
. On the Ports tab, visit the forwarded port 3000 URL marked as Application to see the human essentials page. - Login as a sample user with the default credentials.
Please let us know by opening up an issue! We have many new contributors come through and it is likely what you experienced will happen to them as well.
- "My RBENV installation didn't work!" - The rbenv repository provides a rbenv-doctor script to verify the installation and check if a ruby version is installed
- Follow this SO post to force push the main repo's Wiki to your fork's Wiki.
- Make edits to your fork's Wiki.
- Create a documentation issue about your changes. Make sure to note which pages you changed and link to your fork's Wiki.
- Someone will review and approve your changes and merge them into the main Wiki following this SO post
-
Identify an unassigned issue. Read more here about how to pick a good issue.
-
Assign it to avoid duplicated efforts (or request assignment by adding a comment).
-
Fork the repo if you're not a contributor yet. Read about becoming a contributor here.
-
Create a new branch for the issue using the format
XXX-brief-description-of-feature
, whereXXX
is the issue number. -
Commit fixes locally using descriptive messages that indicate the affected parts of the app. Read debugging tips here.
-
If you create a new model run
bundle exec annotate
from the root of the app -
Create RSpec tests to validate that your work fixes the issue (if you need help with this, please reach out!). Read guidelines here.
-
Run the tests and make sure all tests pass successfully; if any fail, fix the issues causing the failures. Read guidelines here.
-
Final commit if tests needed fixing.
-
Squash smaller commits. Read guidelines here.
-
Push up the branch
-
Create a pull request and indicate the addressed issue (e.g.
Resolves #1
) in the title, which will ensure the issue gets closed automatically when the pull request gets merged. Read PR guidelines here. -
Code review: At this point, someone will work with you on doing a code review. The automated tests will run linting, rspec, and brakeman tests. If the automated tests give 👍 to the PR merging, we can then do any additional (staging) testing as needed.
-
Merge: Finally if all looks good the core team will merge your code in; if your feature branch was in this main repository, the branch will be deleted after the PR is merged.
-
Deploys are currently done about once a week! Read the deployment process here.
Please feel free to contribute! While we welcome all contributions to this app, pull-requests that address outstanding Issues and have appropriate test coverage for them will be strongly prioritized. In particular, addressing issues that are tagged with the next milestone should be prioritized higher.
All work is organized by issues. Find issues here.
If you would like to contribute, please ask for an issue to be assigned to you. If you would like to contribute something that is not represented by an issue, please make an issue and assign yourself. Only take multiple issues if they are related and you can solve all of them at the same time with the same pull request.
Users that are frequent contributors and are involved in discussion (join the slack channel! :)) may be given direct Contributor access to the Repo so they can submit Pull Requests directly instead of Forking first.
If starting server directly, via rail s
or rail console
, or built-in debugger in RubyMine, or running bundle exec rspec path/to/spec.rb:line_no
, then you can use binding.pry
to debug. Drop the pry where you want the execution to pause.
If starting via Procfile with bin/start
, then drop a binding.remote_pry
into the line where you want execution to pause at. Then run pry-remote
in the terminal to connect to it.
https://github.com/Mon-Ouie/pry-remote
If you want to connect via Shopify Ruby LSP VSCode extension or rdbg, start the server with bundle exec rdbg -O -n -c -- bin/rails server -p 3000
When running tests in browser, visit the forwarded port 6080 URL to see the browser in Codespaces. You can also visit this port to access the GUI desktop in Codespaces.
In VSCode Run and Debug view, there are some helpful defaults for running RSpec tests in browser at your cursor as well as attaching to a live server. Make sure the Ruby LSP server is started before debugging.
Consider the balance of "polluting the git log with commit messages" vs. "providing useful detail about the history of changes in the git log". If you have several smaller commits that serve a one purpose, you are encouraged to squash them into a single commit. There's no hard and fast rule here about this (for now), just use your best judgement. Please don't squash other people's commits. Everyone who contributes here deserves credit for their work! :)
Only commit the schema.rb only if you have committed anything that would change the DB schema (i.e. a migration).
Try to keep your PRs limited to one particular issue, and don't make changes that are out of scope for that issue. If you notice something that needs attention but is out of scope, please create a new issue.
If you are so inclined, you can open a draft PR as you continue to work on it. Sometimes we want to get a PR up there and going so that other people can review it or provide feedback, but maybe it's incomplete. This is OK, but if you do it, please tag your PR with in-progress
label so that we know not to review / merge it.
Add a test for your change. If you are adding functionality or fixing a bug, you should add a test!
If you are inexperienced in writing tests or get stuck on one, please reach out for help :)
- Prefer request tests over system tests (which run much slower) unless you need to test Javascript or other interactivity
- When creating factories, in each RSpec test, hard code all values that you check with a RSpec matcher. Don't check FactoryBot default values. See #4217 for why.
- Keep individual tests tightly scoped, only test the endpoint that you want to test. E.g. create inventory directly using
TestInventory
rather than using an additional endpoint. - You probably don't need to write new tests when simple re-stylings are done (ie. the page may look slightly different but the Test suite is unaffected by those changes).
- If you need to see a browser/system spec run in the browser, you can use the following env variable
NOT_HEADLESS=true bundle exec rspec
- We've added magic_test which makes creating browser specs much easier. It allows you to record actions on the browser running the specs and easily paste them into the spec. You can do this by adding
magic_test
within your system spec:and run the spec using this command:it "does some browser stuff" do magic_test end
See videos of it in action hereMAGIC_TEST=1 NOT_HEADLESS=true bundle exec rspec <path_to_spec>`
- Helpful classes for viewing and modifying inventory include
View::Inventory
,TestInventory
and variousCreateService
services, see the Event Sourcing wiki page.
Before submitting a pull request, run all tests and lints. Fix any broken tests and lints before submitting a pull request.
- There are Github Actions workflows which will run all tests in parallel using Knapsack and lints whenever you push a commit to your fork.
- Once your first PR has been merged, all commits pushed to an open PR will also run these workflows.
- Run all lints with
bin/lint
. - Run all tests with
bundle exec rspec
- You can run a single test with
bundle exec rspec {path_to_test_name}_spec.rb
or on a specific line by appending:LineNumber
- If you need to skip a failing test, place
pending("Reason you are skipping the test")
into theit
block rather than skipping withxit
. This will allow rspec to deliver the error message without causing the test suite to fail.
it "works!" do
pending("Need to implement this")
expect(my_code).to be_valid
end