When creating a pull-request you should:
- Open an issue first: Confirm that the change or feature will be accepted
- gofmt and vet the code: Use
gofmt
,golint
,govet
andgoimports
to clean up your code. - vendor dependencies with dep: Use
dep ensure --update
if you have added to or updated dependencies, so that they get added to the dependency manifest. - Update api documentation: If your pull-request adding/modifying an API request, make sure you update the swagger documentation (
api-docs.yml
) - Run Api Tests: If your pull request modifies the API make sure you run the integration tests using dredd.
- Check out the
develop
branch - Install Go. Go must be >= v1.10 for all the tools we use to work
- Install MySQL / MariaDB (OPTIONAL!!!)
- Install node.js
- Set up GOPATH, GOBIN and Workspace.
mkdir -p $GOPATH/src/github.com/ansible-semaphore && cd $GOPATH/src/github.com/ansible-semaphore
- Clone semaphore (with submodules)
git clone --recursive git@github.com:ansible-semaphore/semaphore.git && cd semaphore
- Install dev dependencies
go install github.com/go-task/task/v3/cmd/task@latest
task deps
Windows users will additionally need to manually install goreleaser from https://github.com/goreleaser/goreleaser/releases
- Create database if you want to use MySQL (Semaphore also supports bbolt, it doesn't require additional action)
echo "create database semaphore;" | mysql -uroot -p
- Compile, set up & run
task compile
go run cli/main.go setup
go run cli/main.go service --config ./config.json
Open localhost:3000
Note: for Windows, you may need Cygwin to run certain commands because the reflex package probably doesn't work on Windows.
You may encounter issues when running task watch
, but running task build
etc... will still be OK.
Dredd is used for API integration tests, if you alter the API in any way you must make sure that the information in the api docs matches the responses.
As Dredd and the application database config may differ it expects it's own config.json in the .dredd folder. The most basic configuration for this using a local docker container to run the database would be
{
"mysql": {
"host": "0.0.0.0:3306",
"user": "semaphore",
"pass": "semaphore",
"name": "semaphore"
}
}