Mobile Food Facility API

The purpose of this api is to manage mobile food truck stations in a given city. The API can be used to add, find a food truck by location id, and list all food trucks by block. To test out the API, we use mobile food truck stations' data of the city of San Francisco.

Assumptions and consideration

The API should be deployed in a micro-service architecture on the cloud. To that goal a Dockerfile is provided to an image.
The API uses in-memory database which can be replaced easily with a database as the repository implements an interface Repository.
locationId is considered as a primary field.

Technologies

The following technologies are used:

The api is implemented using nodejs with typescript, and express. nodejs is one of the ideal choices to quickly set up an api for micro-service environment.
node-cache is used for in-memory database
For now only applicant, and facilityType are made required when inserting food truck.
logging is provided by winston along with morgan to log http logs.

Usage

To run the app in development run npm run dev. Use REST client or curl to test it.

To add a new food truck

  curl --header "Content-Type: application/json"  \
       --request POST   --data '{"applicant":"binyam","facilityType":"truck"}' \ 
         http://localhost:3000/food-truck

To retrieve a food truck based on locationId field

  curl http://localhost:3000/food-truck/1514023

To get all food trucks around a block. By default, the api returns 10 trucks which can be changed by providing query params limit and offset.
```
  curl http://localhost:3000/food-truck?block=3507
```

Test

To run a unit test do,

  npm run test

To run an integration test do

  npm run integration_test

Deployment

As we mentioned above the app will be deployed in a micro-service environment, which mean we need an image to spin off containers. In the current effort we provided a dockerfile to create an image based on node:alpine base image. We also added docker-compose files to test the container in a local setting.

The production deployment can be tested locally using docker-compose as follows.

   docker-compose -f docker-compose.yml -f docker-compose-prod.yml up -d

TODO

Add the additional REST http verbs such as PATCH, and DELETE.
Add a custom token to track api requests in logs.
Implement token based authorization, and authentication.
Add more validation and test on post data.
Include appropriate API documentation using tools such as swagger.
Use a database with appropriate optimization which handles millions of records. Nosql databases such as mongodb can be a good candidate.

bdarge / mff