Artify

Installation

1. Ensure that Docker, docker-compose and python3 are installed on your machine
2. Run docker-compose up
3. In another shell instance navigate to the api directory
4. To initialize the Elasticsearch Indices run python3 init_db.py wait until the script outputs that the indices have been created.
5. The services should all be up and running

Running After Installation

1. Run docker-compose up and wait until the logging outputs that the: Cluster health has changed from [RED] to [GREEN] as shown below

   elasticsearch    | {"type": "server", "timestamp": "XXX", "level": "INFO", "component": "o.e.c.r.a.AllocationService", "cluster.name": "docker-cluster", "node.name": "XXX", "message": "Cluster health status changed from [RED] to [GREEN] (reason: [shards started [[image-products][0]]]).", "cluster.uuid": "XXX", "node.id": "XXX"  }
   

   XXX used denote extra information

Description of Services Running

API - http://localhost:5000

The api can also be accessed on http://locahost/api which is just proxied to http://localhost:5000

Endpoints

• Examples use cURL with -i to provide the HTTP response headers in the output

Register

• Used to create an account for use with the image repository returns 200 and a session cookie if the request is successful

Example usage:

curl -i 'http://localhost/api/register' -H 'Content-Type: application/json' --data-raw '{"email":"test@user.ca","password":"supersecret","name":"Test User"}' -c cookies.txt

Example error:

{
    "error_message": "Password is required"
}

Login

• Used to login to a registered account returns 200 and a session cookie if the request is successful

Example usage:

curl -i 'http://localhost/api/login' -H 'Content-Type: application/json' --data-raw '{"email":"test@user.ca","password":"supersecret"}' -c cookies.txt

Example error:

{
    "error_message": "Password is required"
}

Create Product

• Used to create a new product. Returns 200 if successful otherwise returns 400 and an error message or a 403 if you are not authenticated

Example usage:

 curl -i 'http://localhost/api/create' -F "title=ABC" -F "visibility=Public" -F "price=1234" -F "description=Test123" -F "files=@/PATH_TO_FILE.png" -b cookies.txt -c cookies.txt

Example error:

{
    "error_message": "A title is required"
}

Search

• Used to search for visible products. Supports limiting searches by price and by query text. Returns an array of matching products.

Example usage:

curl -i 'http://localhost/api/search?query=Logo&price_min=5&price_max=12'

Get Image

• Used to retrieve images from the storage system.

Example usage (change the UUID):

curl 'http://localhost/api/image/1b640668-ab4c-4db1-84f9-9698e2068dcc/' > out

Frontend - http://localhost

• This hosts a static set of pages to enable testing of the API in a more easy fashion

• Upon loading the page for the first time there will be an empty white screen. This is as there are no products added to the database. To add products:

  • Using the top navbar navigate to Account > Create Account fill in the signup form and submit
  • After submitting the page you should be redirected to the Create Product page
  • Fill in the form to create a product
  • Afterwards, navigate to http://localhost
  • Your newly added item should be visible (assuming visibility was set to Public)

• Additionally, there is a search feature allowing for filtering the products by price and by title and description

Elasticsearch - http://localhost:9200

• All the data about the users and products are stored and indexed in Elasticsearch

Things to Improve

• Web is mainly just for testing the API. As a result, it doesn't dynamically respond to the user being logged in

• Support for chunking uploads to allow very large files to be uploaded

• Use JWT to handle authentication instead of session cookies

• Pagination of results

• Add better documentation of the API

• Initialize database indices with a different method

• Add support to display multiple images on web app (backend supports multiple images per product)