Pantry API

Pantry API is a GraphQL application which is used to aid in managing inventory for a pantry. This application utilizes Postgres and Elasticsearch as its data backend. The application is designed to run in a Docker container.

Running

This application requires the following services to run:

• Keycloak⁠
• OpenSearch⁠
• PostgreSQL⁠

In order to set up connections, check the configurations section below for environment variable configurations.

The docker image is built at https://hub.docker.com/r/mikeyglitz/pantry-api.

docker run -p 8080:8080 -e KEYCLOAK_CLIENT_ID=... mikeyglitz/pantry-api

Configuration

The configuration for this application is managed through environment variables

OpenSearch Role

This application requires a user account and role to be created in OpenSearch with the following permissions:

Cluster-Level Permissions

• cluster:monitor/health
• cluster:monitor/main
• indices:data/read/scroll/clear

Index-Level Permissions

Indices - ingredients, users, and pantries

• create_index
• indices:data/read/search
• indices:data/write/delete/byquery
• indices:admin/get
• indices:admin/aliases/exists
• indices:data/write/update/byquery
• indices:data/read/get
• indices:admin/create
• indices:data/write/index
• indices:admin/mapping/put
• indices:data/read/scroll/clear
• indices:data/write/bulk*

Endpoints

The API provided by this application exposes the following endpoints:

The /playground and /graphql endpoints can be configured by the SVC_PLAYGROUNDPATH and SVC_PATH environment variables respectively

Project Structure

This project utilizes the Go standard project layout⁠.

• test - stores integration tests utilizing testcontainers⁠
• pkg - contains the root application server bootstrapping code
• cmd - contains the main class -- the application entrypoint
• internal - majority of the business logic is here
  • internal/config - configuration functions and interfaces
  • internal/database - database driver, models, and migration logic
  • internal/index - OpenSearch driver and interfaces
  • internal/dataloaders - dataloader⁠ interfaces
  • internal/handlers - gin rest handlers
  • internal/middleware - middleware used to enhance the request context (i.e. authentication)
  • internal/graph - GraphQL resolvers
  • internal/graph/services - database service layer - utilized primarily to supply data to GraphQL resolvers
  • internal/graph/schema - GraphQL schemas -- written in the GraphQL DSL⁠

Roadmap

• Implement configuration
• Implement Elasticsearch driver
• Implement ORM
• Implement GraphQL schema
• Separate GraphQL schema into separate files/resolvers
• Break database/elasticsearch code into services and dataloaders
• Validate auth token with Keycloak
• Refactor to supply services and user using middleware
• Create validation directive
• Create user permissions directive

References

• gormigrate - https://github.com/go-gormigrate/gormigrate⁠
• testcontainers with elasticsearch - https://riferrei.com/using-testcontainers-go-with-elasticsearch/⁠
• Keycloak golang webservices - https://mikebolshakov.medium.com/keycloak-with-go-web-services-why-not-f806c0bc820a⁠
• Opinionated graphql server with go - https://dev.to/cmelgarejo/creating-an-opinionated-graphql-server-with-go-part-1-3g3l⁠
• Custom GraphQL validation - https://david-yappeter.medium.com/gqlgen-custom-data-validation-part-1-7de8ef92de4c⁠
• Custom error types with go - https://klotzandrew.com/blog/error-handling-in-golang⁠