MCQ API

The Multiple Choice Question API offers the users a set of questions based on a subject and a use (type)

Data

The questions.csv file is the database that keeps all the questions with the following fields:

• question: the question itself
• subject: the question subject
• correct: the correct answer to the question
• use: MCQ type
• responseA: response A
• responseB: response B
• responseC: response C
• responseD: response D (if exists)

API Specifications

1. The user must choose a test use as well as one or more test subject.
2. The API should be able to return quizzes with 5, 10 or 20 questions.
3. Order of the questions must be random when returned.
4. The API must check the identity of users depending on route used.
5. It uses a Basic username:password authentication process
6. API must have a route for the API status check
7. An admin user (username : admin, password: 4dm1N)should be able to add questions to the database
8. Finally, the API should return errors when it is wrongly used

Following are the registered users

{
  "alice": "wonderland",
  "bob": "builder",
  "clementine": "mandarine"
}

Files

The API is made from the following files:

• mcq_api.py : the main app with the actual endpoints
• mcq_utils.py : user database and various modules used in the main app.
• mcq_requests.py : script with requests made to test the API
• questions.csv : question database
• requirements.txt : list of packages to install for the API to run

Endpoints

Endpoints	Authentication	Description
GET /mcq	USER	returns a multiple choice question quizz based on subjects (default : BDD), use (default : Test de positionnement) and nb (default : 5)
GET /api_status	ADMIN	returns API status
POST /add_mcq	ADMIN	add new question in the database (questions.csv) from JSON/Dict Format

Setup

Install the required packages :

pip install -r requirements.txt

Launch the uvicorn server :

uvicorn mcq_api:api --reload

The API is accessible in a browser at http://127.0.0.1:8000

To test the API with the mcq_requests.py script :

python mcq_requests.py

Results with API Interface

API Interface

MCQ parameters

MCQ response

Add MCQ parameters

Add MCQ response

API Status parameters

API Status response

Results for mcq_requests.py

# GET /mcq
response without parameters :
 {'detail': [{'loc': ['query', 'subjects'], 'msg': 'value is not a valid list', 'type': 'type_error.list'}]}

response without credentials :
 {'detail': 'Not authenticated'}

response with wrong parameter :
 {'detail': 'Bad Request : One or more parameter not valid'}

response with all needed :

{'username': 'clementine',
 'authorization': 'Basic',
 'Number of questions': 5,
 'Use': 'Test de positionnement',
 'Subjects': ['BDD', 'Systèmes distribués', 'Docker'],
 'Questions': {'19': {'question': 'DockerHub est',
                      'subject': 'Docker',
                      'use': 'Test de positionnement',
                      'correct': 'C',
                      'responseA': 'un système qui permet de lancer plusieurs '
                                   "containers d'un coup",
                      'responseB': "un système d'orchestration de containers",
                      'responseC': "un répertoire d'images Docker",
                      'responseD': None,
                      'remark': None},
               '17': {'question': 'Docker permet de persister des changements',
                      'subject': 'Docker',
                      'use': 'Test de positionnement',
                      'correct': 'C',
                      'responseA': 'Oui',
                      'responseB': 'Non',
                      'responseC': "Oui à condition d'utiliser des volumes",
                      'responseD': None,
                      'remark': None},
               '0': {'question': 'Que signifie le sigle No-SQL ?',
                     'subject': 'BDD',
                     'use': 'Test de positionnement',
                     'correct': 'A',
                     'responseA': 'Pas seulement SQL',
                     'responseB': 'Pas de SQL',
                     'responseC': 'Pas tout SQL',
                     'responseD': None,
                     'remark': None},
               '3': {'question': 'OrientDB et Neo4J sont des bases de données',
                     'subject': 'BDD',
                     'use': 'Test de positionnement',
                     'correct': 'D',
                     'responseA': 'relationnelles',
                     'responseB': 'orientées objet',
                     'responseC': 'orientées colonne',
                     'responseD': 'orientées graphe',
                     'remark': None},
               '18': {'question': 'Des containers Docker peuvent communiquer '
                                  'entre eux grâce à',
                      'subject': 'Docker',
                      'use': 'Test de positionnement',
                      'correct': 'B',
                      'responseA': 'des volumes',
                      'responseB': 'des networks',
                      'responseC': 'des communications ',
                      'responseD': None,
                      'remark': None}}}

# POST /add_mcq

response with data missing required fields :

{'detail': [{'loc': ['body', 'correct'],
             'msg': 'field required',
             'type': 'value_error.missing'},
            {'loc': ['body', 'responseA'],
             'msg': 'field required',
             'type': 'value_error.missing'}]}

response without credentials :
 {'detail': 'Not authenticated'}

response with wrong credentials :
 {'detail': 'Incorrect username or password'}

response with all needed :

{'username': 'admin',
 'authorization': 'Basic',
 'Question inserted': True,
 'Question': {'question': "Quelle est la couleur du cheval blanc d Henri IV ?",
              'subject': 'Classification',
              'use': 'Test de positionnement',
              'correct': 'B',
              'responseA': 'jaune',
              'responseB': 'blanc',
              'responseC': 'noir',
              'responseD': 'la reponse D',
              'remark': ''}}

la dernière ligne de questions.csv est maintenant:
Quelle est la couleur du cheval blanc d Henri IV ?,Classification,Test de positionnement,B,jaune,blanc,noir,la reponse D,


# GET /api_status

response with wrong credentials :
 {'detail': 'Incorrect username or password'}

response with all needed :
 {'status': 'API working'}