hilyafadhilah / bnmofin-api

REST API Seleksi Tim Laboratorium Programming 2020 Tahap IIA

bnmofin-api

13520024 Hilya Fadhilah Imania

Stack

• Node.js >=16
• yarn
• Express
• PostgreSQL
• Redis
• Firebase Storage
• ApiLayer Exchange Rate API

Relevant Tools

• TypeORM 0.3.x
• TypeORM Seeding (fork by jorgebodega) 5.x
• jsonwebtoken
• bcryptjs
• Axios

Other Tools

• routing-controllers
• class-validator
• class-transformer
• fakerjs

Environment Variables

Notes

In this section, these values are defined as

• Truthy value: TRUE, T, YES, Y, 1, ON
• Falsy value: FALSE, F, NO, N, 0, OFF

Case insensitive

Also, if you are not using docker, you need to set up the environment yourself. This app does NOT use dotenv.

App

Most important environment variable for setup is PORT. This needs to be the same port the client(s) will be connecting to. Set it up to whatever value available on your host. Default value: 3030.

Stack-Related

Important variables that need to be configured by yourself:

• APILAYER_KEY: ApiLayer Key (for Exchange API)
• GOOGLE_APPLICATION_CREDENTIALS: Google service account key file
• GOOGLE_STORAGE_BUCKET: Google storage bucket url

Without Docker Compose

You can set up database and redis connection options for development or other cases where you need to connect to database and redis server other than local docker compose services.

• DATABASE_URL: PostgreSQL connection URL
• DATABASE_SSL: If set to truthy value will enable SSL connection
• REDIS_URL: Redis connection URL

Seeding

It is recommended to enable seeding the first time you run the app, then disable it afterwards.

Specify SEED_DATA to truthy value alongside NODE_ENV=development (or unset). Beware that the seeder will drop the database schema alongside all data inside it.

If NODE_ENV is set to production the app will not run the seeder.

The SEED_INFO variable is useful to get information for testing purposes. It will print out the user info needed for authentication. This feature is tightly related to the seeder because the seeder will make accounts have the same password as their username. (Aside for easier input, this is due to the password being one-way encrypted.)

Sample output when SEED_INFO is set to truthy value:

--- Start Seeder Info ---

[Admin Accounts]
- Cydneyrnmg4
- Santa63860
- Timmyzm2yb
- Samanthabperq
- Kylee9o5q9

[Customer Accounts (positive balance)]
- Alverta8aj2o
- Peggieul7pj
- Helena1ohdo
- Gabet438q
- Demetriuselzn6

[Customer Accounts (negative balance)]
- Hobart69yoe
- Brendaubu5v
- Dustino6mw9
- Antwonc49cn
- Richardhxate

P.S. Username = Password


--- End Seeder Info ---

Output of seeder info is console-only and not logged.

Logging

Error-level log file is enabled by default and can be disabled with falsy value. All-level log file can be enabled with truthy value. A custom path file can also be specified. By default, the logs will be on logs/errors.log and logs/logs.log.

Others

For other variables and more information see .env.example.

Build & Run

Docker compose

$ docker compose [--env-file path-to-env-file] build
$ docker compose [--env-file path-to-env-file] up [-d]

Without Docker

$ yarn
$ yarn build
$ yarn start

Don't forget to set up the DATABASE_URL and REDIS_URL env variables.

Development (recommended)

With db & redis env settings:

DATABASE_HOSTPORT=5433
DATABASE_URL=postgresql://postgres:postgres@localhost:5433/bnmofin
REDIS_HOSTPORT=6380
REDIS_URL=redis://localhost:6380

Then run:

$ docker compose [--env-file path-to-dev-env-file] up db redis -d
$ yarn dev

Scripts

• seed:run: Run the seeder & seeder info, regardless of SEED_DATA and SEED_INFO environment settings. Beware when running on production.
• seed:info: Only run seeder info.
• storage:clear: Delete all files from google storage "directory" (which can be defined with GOOGLE_STORAGE_PREFIX)

For these scripts to run properly the related environment variables need to be set

Documentation

API Docs is available on bnmofin swagger.

Database Schema

Design Patterns

1. Singleton

Used for database & redis connection and logging mechanism (file opening, etc.). All of which only need to be instantiated once. The implementation does not use objects, but rather ES6 exports, using const to prevent reassignment.

See: src/data-source.ts, src/logger.ts, src/cacher.ts

2. Chain of responsibility

This design pattern is implemented through middlewares and interceptors. Each middleware deals with specific task, but all work upon the same Request object that is instantiated for each request the server receives.

See: src/middlewares/

3. Factory Method

This pattern is used for seeding. Creation of each entity object with random values is "uniform" (have specific rules) for each entity type. So the code that handles the seeding uses the factories and doesn't need to worry about the specific details of generating random entity objects.

See: src/seed/factories/

Deployment

The master branch is deployed and live on https://bnmofin-api.herokuapp.com using Heroku automatic deployment.