sinardyas / comuline-api

The API for Indonesian public transportation data, built using Elysia

@comuline/api

An API to get the schedule of KRL commuter line in Jakarta and Yogyakarta using Elsyia and Bun, deployed to Render. This API is primarily used on the web app (source code).

How does it work?

This API uses a daily cron job (at 00:00) to fetch the schedule of KRL commuter line in Jakarta and Yogyakarta from the official website of PT. KAI. The data is then processed and stored in a PostgreSQL database and cached in a Redis (for every once read request). All endpoints can be found in the docs.

Technology stacks

1. Elsyia API framework
2. Bun runtime
3. PostgresSQL (Neon)
4. Redis (Upstash)
5. Render deployment platform
6. Drizzle ORM

Getting Started

Development

1. Clone the repository

git clone https://github.com/comuline/api.git

2. Install the dependencies

bun install

3. Run database locally

docker-compose up -d

4. Copy the .env.example to .env

cp .env.example .env

5. Run the database migration

bun db:generate && bun db:migrate

6. Sync the data and populate it into your local database (once only as you needed)

# Please do this in order
# 1. Sync station data and wait until it's done
curl --request POST --url http://localhost:3001/v1/station/
# 2. Sync schedule data
curl --request POST --url http://localhost:3001/v1/schedule/

Deployment

1. Create a new PostgreSQL database in Neon and copy the connection string value as DATABASE_URL in your .env file

2. Run the database migration

bun db:generate && bun db:migrate

3. Sync the data and populate it into your remote database (once only as you needed)

# Please do this in order
# 1. Sync station data and wait until it's done
curl --request POST --url http://localhost:3001/v1/station/
# 2. Sync schedule data
curl --request POST --url http://localhost:3001/v1/schedule/

4. Generate SYNC_TOKEN (This is used in production level only to prevent unauthorized access to your POST /v1/station and POST /v1/schedule endpoint)

openssl rand -base64 32
# Copy the output value as a `SYNC_TOKEN`

2. Create a new Redis database in Upstash and copy the connection string value as REDIS_URL

3. Create a Web Service in Render, copy the DATABASE_URL, REDIS_URL, and SYNC_TOKEN as environment variables, and deploy the application.

4. Set the cron job to fetch the schedule data using Cron-Job. Don't forget to set the SYNC_TOKEN as a header in your request. Add the ?from_cron=true query parameter to flag the request as a cron job request.

# Example
curl --request POST --url https://your-service-name.onrender.com/v1/station?from_cron=true -H "Authorization: Bearer ${SYNC_TOKEN}"
curl --request POST --url https://your-service-name.onrender.com/v1/schedule?from_cron=true -H "Authorization: Bearer ${SYNC_TOKEN}"

Database schema

Station

Column Name	Data Type	Description
id	TEXT	Primary key (Station ID)
name	TEXT	Station name
daop	INTEGER	Station regional operation code
fgEnable	BOOLEAN	-
haveSchedule	BOOLEAN	Schedule availability status
updatedAt	TEXT	Last updated date

Schedule

Column Name	Data Type	Description
id	TEXT	Primary key (Station ID + Train ID)
stationId	TEXT	Station ID
trainId	TEXT	Train ID
line	TEXT	Train commuter line
route	TEXT	Train route
color	TEXT	Commuter line color
destination	TEXT	Train destination
timeEstimated	TIME	Estimated time
destinationTime	TIME	Destination time
updatedAt	TEXT	Last updated date

Sync

Column Name	Data Type	Description
id	TEXT	Primary key (Sync ID)
n	BIGINT	n of sync
type	ENUM	Sync type (manual, cron)
status	ENUM	Sync status (PENDING, SUCCESS, FAILED)
item	ENUM	Sync item (station, schedule)
duration	BIGINT	Sync duration
message	TEXT	Sync message (if status failed)
startedAt	TEXT	Sync started date
endedAt	TEXT	Sync ended date
createdAt	TEXT	Sync created date