• Receives and sends datasets as files and json
• Uses MongoDB to store datasets
• .csv, .json and .arff file formats are supported. Once the file contents have been saved, it can be downloaded as any of the previously listed formats.
• 'testDatasets' folder contains one file for each supported fileformat. They can be used for testing.
• Some features (such as dataset uploading) require an authorization token that can be received by logging in with an account.
• An account can be created by sending a POST-request with credentials to the '/users' endpoint.
• Max upload size at a time: 100 KB
• Unit tests ready for '/datasets/' endpoint

1. Create a MongoDB database
2. Clone this repository
3. Run npm install inside the project folder
4. Set environment variables (More on this in the Configuration section)

To run this project in production mode, run npm start. For development, run npm run watch. And to run tests, run npm test.

In order to use this server, the following environment values should be set:

• PORT: Port in which this server runs. The matching frontend expects this to be set to 8000)
• MONGODB_URI: MongoDB address)
• SECRET: String that is used for user authentication
• (TEST_MONGODB_URI: MongoDB address that is used in test environment)

One way to setup environment variables is to use the dotenv module (included in package.json).

The authentication token needs to be sent as the 'Authorization' header value with the prefix 'bearer '. Example: Header key: 'Authorization', value: 'bearer someValidToken'.

Creates a new account with given credentials. The credentials should be passed in the request body. The account creation will fail if the username is too short (3 letter) or if it is not unique. Succesful requests get an response that contains the username and id.

Example body:

{
 "username": "unique username",
 "password": "not an easy to guess pw"
}

Deletes an account with the same id as the one specified in the URL parameter. All the datasets that the account has posted will also be deleted. Upon successful request, the server will respond with an empty 204 response.

Updates an account with the same id as the one specified in the URL parameter. Only the password can be changed. Upon successful request, the server will respond with an empty 201 response.

Example body:

{
  "password": "new password"
}

Checks the database to see if an account with the given credentials exists. If such an account is found, the response will contain an authentication token along with the username.

Example body:

{
 "username": "unique username",
 "password": "not an easy to guess pw"
}

Returns all of the uploaded datasets as JSON. This endpoint supports query strings with the following options: 'name', 'username' 'fields' and 'limit_instances'. The first two options tell this server to only return datasets that match the corresponding values (same name or username). 'fields' option lets the requester specify which of the dataset fields should be returned. These fields are:

• id
• user
• name
• relation (only appears in datasets uploaded as .arff files)
• headers
• instances

Finally the 'limit_instances' option tells the server to only return the specified amount of data instances per dataset. The request will fail (Status code 400) if the value is not a positive number.

Saves the received dataset json to the database. The dataset contents should be placed inside the request body. Upon successful request, the server will respond with with an 201 response with the saved dataset object.

Example body:

{
  "name": "patient_stats",
  "headers": [
    "height",
    "weight",
    "blood-type"
  ],
  "instances": [
    [
      "160cm",
      "87kg",
      "A"
    ],...
  ],
  "user": "someValidUserId"
}

Returns the dataset with an id that matches the specified URL parameter (:id) as JSON if one exists. This endpoint supports the same query string options as the previous endpoint ('/datasets').

Deletes the dataset with an id that matches the specified URL parameter (:id) if one exists. Users can only delete datasets that they themselves have posted (same account). Upon successful request, the server will respond with an empty 204 response.

Attaches a file with the same id as is specified in the URL parameter (:id) to the response. The file format will is specified in the ':format' URL parameter. Format options are csv, arff and json. Sent files are removed from the disk as soon as the response has been sent.

During user login, the client sends the user's credentials to this server (POST request to '/login'). The credentials are then validated and access token is sent on the response. The client can then send the received token with further requests to gain access to services such as file uploads which would otherwise be unavailable.

It should be noted that this project does not separate the authorization server from the resource server. In a more strict environment, the validation process and token creation should be handled in a dedicated authorization server.

For more information on OAuth2, check this link