test if a user already exists

user.spec.js

import chai from 'chai';
import chaiHttp from "chai-http";
import { app as server } from './app.js';

// configure chai to send/receive HTTP REQUEST
chai.use(chaiHttp);

describe("TDD for Users", () => {
  after(() => {
    server.close();
  });
  context("Check if a User already Exist", () => {
    it("username should exist", (done) => {
      chai.request(server)
        .get('api/v1/getuser/emmsdan')
        .end((error, response) => {
          response.should.be.status(200);
          response.should.be.an('object');
          done();
        })
    })
    
    it("user should not exist", (done) => {
      chai.request(appServer)
      .get('api/v1/getuser/fakename')
      .end((error, response) => {
        response.should.be.status(404);
        response.should.be.an('string');
        done();
      })
    })
  })
})

Thank you for that,
for readability using context in a describe block is better than using a describe in a describe.

Simple and straight-forward test.
Do you mind checking your grammatical construction the test descriptions?

Simple and neatly constructed. Noticed a few things:

• When using the it method, it would help if the string you pass into it is easily readable for the developer(s) who will be reading the test descriptions to know what it does. A sample snippet for clarity:

describe('TDD for users')

it('should return a 200 response if user already exists')
it('should return a 404 response if user does not exist')

I think my example seems a little wordy though.

• In the second case, you asserted that the response returned by chai should be a string but it always returns an object like you did in case 1. I think you wanted to drill down to response.body

• Eslint flags accepting a variable or argument you end up not using in your code. e.g. in the end statements, you could prefix the error objects with an underscore i.e. _error.