- API Description
- List of Technologies
- Features
- Installation
- Configuration Setup
- Usage
- Testing
- Design and Architecture
- Deployment Suggestions
- License
- Contact Information
image-labeling-api is an image labelling and handling API. This API manages uploading images, assigning labels (tags) to them and enables retrieving or changing image information. Users have different roles such as administrators with highest access to the system. Every user should login to the system to upload an image and assign a label to it. Labelling can also be done to an existing image in the system separately. All users can upload, change and retrieve information of images. Only administrators can register a new user. But, every user can change his/her own password in the system.
I used the following technologies for creating image-labeling-api.
| Technology | Description | Version |
|---|---|---|
| TypeScript | Main programming language | 4.4.3 |
| Node.js | Runtime JavaScript Environment | 17.2.0 |
| Express.js | Web application framework for Node.js | 4.17.1 |
| TypeORM | Object Relational Mapping | 0.2.41 |
| JEST | Testing Framework | 27.2.1 |
| JWT | JASON Web Token for authentication | 8.5.1 |
| PostgreSQL | Relational Database Management System | 14.1 |
| pgAdmin 4 | PostgreSQL Tools | 6.2 |
| Docker | Containerization of app, database and database admin tools | 20.10.11 |
| Postman and docgen | Testing API endpoints and generating documentation | 9.3.1 |
image-labeling-api has number of good features as follows:
- Documentation
- Custom error handling
- Authentication
- Expressive use of HTTP status codes
- Log activity
- Paginate the results
- Express URLs with nouns rather than verbs
- Consistent code convention
- Scalable designed architecture
- Database migration
- Unit testing
- Development-environment and Production-ready docker images
For installation please consider the following steps:
-
Clone the repository from git link to repository
-
Build docker image (~docker-compose build)
-
Start the docker image (~docker-compose up)
API features authentication and only admin users can create a new user. First admin user should be added through database migration. For migration and creating first user, please run the following command.
~npm run migration:run
To check database schema and access to it, please do the following steps:
pgAdmin:
- open 127.0.0.1:5050 in your browser
- Login to pgAdmin. username: root@root.com, password: root
- in pgAdmin create new server with following parameters. server name: yourchoice, server host: db, port:5432, database: postgres, username: postgres, password: postgres
Alternatively you can access database in terminal:
- ~docker exec -it image-labeling-api_db_1 bash
- ~psql -U postgres
Detailed documentation for usage of API was provided in two formats in docs folder of main branch. HTML and Markdown.
The HTML documentation visualises examples and should be viewed in a browser after a download from git repository. Git does not provide preview for HTMl documents currently.
To run unit tests please run the following command:
~npm test
I added CI through Github Actions on push and pull_requests.
image-labeling-api was designed base on the REST API architecture and features:
Platform independence. Any user is able to call the API, regardless of how the API is implemented internally. It uses standard protocols, and client and server agree on the format of the data (JSON) to exchange.
Service evolution. This API is able to evolve and add functionality independently from client applications. Additionally, during the service evolution, existing client applications can continue to function without modification.
Figure below shows an example of flow of a successful request from API.
image-labeling-api has user and image data models in a schema that provides one to many relationship between user and image models. Figure below demonstrates the entity relation diagram designed for database.
image-labeling-api is scalable and can be used as a microservice. Figure below shows aa suggested architecture for this purpose.
A suggesstion for deployment of API in AWS is shown in figure below.
image-labeling-api is licensed under MIT.
My Github username is hamiddoost.