This project builds a waterfall service within the university course Advance Full Stack Development. This part of the project is only the backend code. It provides an API for storing and rating of waterfalls. Part of the requirements is to use the backend framework hapi.
- Download the source code as zip or clone it via ssh.
- If you downloaded the zip, then unzip it
- Next go into the folder placemark. It's the root directory of the project.
- Here you need to run the following command in the command prompt (Make sure that you got an up to date version of
node and npm):
npm install
To start the backend use the following command: npm start. Now the application runs on you local machine at
port 3000. The base url of the service is http://localhost:3000.
In this section three different deployment methods for this project are explained. Two of them (Glitch & Heroku) are for the application and Cloud Atlas is for the mongoDB.
To deploy to glitch you just need to connect you glitch account with your github account. Then import it by using the gihub repo link. It always uses the main branch by default. When directly importing is not working try other possiblities suggested in the glitch help forum. This project is deployed with following link: waterfalls-backend-service (This link is the base route for the Glitch deployment).
To deploy the mongo db to Cloud Atlas you need to change the value for the key db in the .env file. Cloud Atlas provides a connection link where you only need to add your password. Important: Do not upload or share this link as because it contains your credentials and might turn into a security breach.
To deploy the application to Heroku you need to either use the command line tool heroku or do it visually in the webapp.
First you need to log in: heroku login.
Then you have to create a new app: heroku create.
The next step is to define a git remote via: heroku git:remote -a <you app name in heroku> (You can find this command in the heroku webapp in the section deploy of you app).
The last step is to call git push heroku main.
Now you should get offered links to your application.
Go to your dashboard in Heroku after signing up/in.
Create a new app and select it.
Next go to the section Deploy.
You can now connect to your Github.
Select the right repository and the branch you want to deploy.
You can Enable Automatic Deploys, to always get the latest main version.
Make sure you select the main branch.
Another option is to deploy manually the main branch.
Here you have to make sure, to always look out for updates.
Since we do not commit the .env file, we need to define the properties in another way.
Go for that into the app's settings and press under the section Config Vars the button Reveal Config Vars.
Now you can set all properties from you local .env file.
Make sure to use secure passwords.
Remember that the .env properties are case sensitive, COOKIE_NAME and cookie_name are not the same!
When running the application you find a OpenAPI (Swagger) Documentation of the API at <urlOfDeployment>/documentation.
I.e. if you run on localhost you will find the documentation at http://localhost:3000/documentation.
There are several API routes available and can be looked up in the file api-routes.js or in the OpenApi documentation at <urlOfDeployment>/documentation.
(I.e. if you run on localhost you will find the documentation at http://localhost:3000/documentation)
The base route for the APIs are http://my.url.yz/api where my.url.yz is the exposed url. When accessing users you append the
API base route with /users. We can either get (HTTP GET), create (HTTP POST), update (HTTP PUT) or
delete (HTTP DELETE) users. The REST Pattern is followed. The same applies to the waterfalls by using /waterfalls
instead of the user route.
The user object has the following structure:
{
"firstName": "Homer",
"lastName": "Simpson",
"email": "[email protected]",
"password": "aSecretYouMustNotShare!",
"isAdmin": false,
"_id": "62b44b300d0ff1c3bcc55d64"
}And a sample waterfall object looks as follows:
{
"_id": "62b45022a85eff476f5d073e",
"name": "Niagara Falls",
"location": {
"lat": 43.0799,
"long": -79.0747
},
"description": "Amazing falls",
"categories": {
"continent": "North America",
"size": "large"
},
"userid": "62b44b300d0ff1c3bcc55d64",
"__v": 0
}The _id is created by the backend.
Later a more detailed API documentation will be provided with swagger.
In-Memory and JSON memory model are deprecated starting with the version 0.3.0. They are no longer supported, since the new features are only tested and developed in the mongodb.
When running the user api we need to change the auth for the create method:
create: {
// auth: {
// strategy: "jwt",
// scope: "admin",
// },
auth: false,
...
Explanation: Due to securing the api the test cannot create an admin user without having already admin privileges. Since we have an empty database it gets tough to achieve then admin privileges which we need to test the whole api.