Blog App Project

Scenario

You have been hired to complete an API for a blog app. The previous developer hasn't shown up for work for a few days. The talk around this office is that he won the lottery but no one really knows for sure. It's kinda a mystery. Unfortunately Sherlock, that's not the puzzle you were hired to solve. Your challenge is to complete the blog app in time for your manager's demo to the CEO next week.

Here's what you know:

• The previous developer created a front-end client using jQuery and Fetch. And started a API using Node and Express.
• The API currently uses dummy data which is just an array of objects in Javascript. It is not wired up to a database yet.
• The company uses PostgreSQL as their database. And hosts their applications on Heroku.
• Your manager says the specs for the API were documented using Postman's documenter and can be found here: Blog App API.
• And you found the previous developer's todo list which list the tasks and notes that need to be completed.

Tasks:

• Create Skeleton endpoints using dummy data

  • GET list all stories - should return status 200 and an array of objects
  • GET get a story by id - should return status 200 and an object
  • POST (create) a story - should return status 201 with a location header and the new object
  • PUT (update) a story - should return status 200 and the updated object
  • DELETE a story - should return status 204 with no content

• Create database

  • Create local Postgres database named blog-app
  • Create a stories table with the following 3 columns:
    • id: an auto-incrementing integer
    • title: regular text. Required.
    • content: regular text. Can be blank.

  Note: save the CREATE TABLE and INSERT INTO queries in a file so they can be easily run again later. Examples:

  • Local DB: psql -f ./query.sql -U <username> -d blog-app
  • ElephantSQL: psql -f ./query.sql postgres://<USERNAME>:<PASSWORD>@<SERVER:PORT>/<DATABASE>

• Wire-up database to the endpoints. IOW, replace dummy data with real database calls.

  • Add knex and pg to the project
  • Update config.js with DB connection info
  • Import knex and database config into the router file
  • Update GET /api/v1/stories endpoint to use knex.select()...
  • Update GET /api/v1/stories/:id endpoint to use knex.select()...
  • Update POST /api/v1/stories/ endpoint to use knex.insert()...
  • Update PUT /api/v1/stories/:id endpoint to use knex.update()...
  • Update DELETE /api/v1/stories/:id endpoint to use knex.del()...

• Deploy to Heroku (see Deploying to Heroku)

  • Install Heroku CLI app, if necessary
  • Create app on Heroku
  • Update git remote to point to Heroku
  • Push app to Heroku. (Note: app won't work yet)
  • Create a production database on Elephant SQL. Use query.sql from earlier to easily create the database
  • Configure Heroku config vars to use DATABASE_URL and the Elephant SQL connection string

Add Authors

Congrats! The boss is very impressed with your work and the quick turn-around. She is so impressed in fact that she is wondering if you could add authors to the stories. You think about for a bit and decide it should be relatively easy to just update the api. But there is a wrinkle, that sales team is already using the first version and the changes might break their app. After pondering your options for a bit you remember you are using express router so you decide to create a /v2 route.

You create the following high-level tasks:

Tasks:

• Update the Database:
  • Create an authors table
    • 3 Columns
      • id: an autoincrementing integer
      • username: regular text. Required.
      • email: regular text. Required.
  • Add author_id foreign key column the stories table that references the authors table.

  Note, when we delete a story, we do not want to delete the associate author, so choose the correct ON DELETE contraint.

  • Add a dummy users to the db and update stories to have authors.
• Create a V2 router file
  • Create a /router/stories-router-v2.js router file
  • Require the router file in server.js and "mount" using app.use()
• Update the queries in the V2 endpoints to use author info.