A RESTful API for managing car models using Python with an object-oriented approach. The API allows users to create, retrieve, update, and delete car models, as well as search and filter car models based on their properties.
- Python Car Management Challenge
For installing the App it is necessary to have Python 3 installed on your System. In the first Step clone the repository onto your machine by running:
git clone git@github.com:FredNoonienSingh/Challenge.gitThere is no environment file included in the repository, therefore it has to be created
cd Challenge
vim .flaskenvthen paste the following into the file
FLASK_APP='PyCaMa'
FLASK_DEBUG=True
FLASK_RUN_PORT=8080
SQLALCHEMY_COMMIT_ON_TEARDOWN=True
SQLALCHEMY_DATABASE_URI = "sqlite:///database.db"More information about the flask env file can be found here.
Close the file by pressing escape follow by :wq Now run the included setup file
pip install -e .Now create the database file by running:
touch database.dbOPTIONAL:
To fill the database with Mock data run
python3 utils/fill_db.pyAfter this we can start the app by running
flask run To verify everything is running correctly
python -m unittest discover -vFollow this Guide and return after you fished the process, an easier but more costly alternative solution can be found here.
class Car(Base):
__tablename__ = 'cars'
id: int = Column(Integer, primary_key=True, autoincrement=True)
make: str = Column(String(15), nullable=False, unique=False)
model: str = Column(String(15), nullable=False, unique=False)
year: int = Column(Numeric(precision=4, scale=0), nullable=False, unique=False)
color: str = Column(String(15), nullable=False, unique=False)
price: float = Column(Numeric(precision=10, scale=2), nullable=False, unique=False)
created_at: datetime = Column(DateTime, default=datetime.now, unique=False)The car object is a representation of a car in the way outlined in the Challenge. It holds the following methods.
def __init__(self, model: str, make: str, year: int, color: str, price: float) -> None:
""" Initializes a Car object
Args:
model (str): Model of the Car
make (str): Make of the Car
year (int): year in which the car was build
color (str): color of the car
price (float): price of the car
"""
self.model = model
self.make = make
self.year = year
self.color = color
self.price = priceThis method instantiates an object, it does not return a new instance.
def update(self, data:dict) -> None:
""" Overwrites Fields of the object with parsed values
Args:
data (dict): dictionary containing the field and new data
"""
for key, value in data.items():
if hasattr(self, key):
setattr(self, key, value)Works by checking for each key:value pair in the dict data, if the class has the attribute and if so overwrites the old value of this attribute with the new value.
def as_dict(self) -> dict:
""" returns fields of the object as dict
Returns:
dict: fields of the object
"""
return {
'ID': self.id,
'model': self.model,
'make': self.make,
'year': self.year,
'color': self.color,
'price': self.price
}Return the a dict with containing the attributes of the class as key with their respective values as value.
def __repr__(self) -> str:
""" returns string representing the object
Returns:
str: string representing the object
"""
return f'<Car {self.make}{self.model!r}>'Returns a string representation of the car, it is usually a standard method but I have overwritten it for debugging purposes.
{
"id":Integer,
} {'Data':
[
{
'ID': 29,
'color': 'red',
'make': 'Porsche',
'model': '911',
'price': '12322.32',
'year': '1999'
}
],
'Success': 'returned single row'}{
"offset": None,
"limit": 12,
"filters": {
"year": {
"operator": 'lt',
"value": 2000
}
},
"model": {
"operator": 'eq',
"value": '911'
}
}- 200
- 404
- 422
{'Data':
[
{'ID': 3, 'color': 'silver', 'make': 'Porsche', 'model': '911', 'price': '2849.00', 'year': '1946'},
{'ID': 5, 'color': 'blue', 'make': 'VW', 'model': 'Astra', 'price': '2448.00', 'year': '1913'},
{'ID': 8, 'color': 'black', 'make': 'Mercedes', 'model': 'Astra', 'price': '1397.00', 'year': '1904'},
{'ID': 9, 'color': 'blue', 'make': 'Mercedes', 'model': 'A3', 'price': '1379.00', 'year': '1910'},
{'ID': 10, 'color': 'blue', 'make': 'Porsche', 'model': 'R8', 'price': '2424.00', 'year': '1952'},
{'ID': 11, 'color': 'black', 'make': 'BMW', 'model': 'R8', 'price': '2317.00', 'year': '1952'},
{'ID': 12, 'color': 'silver', 'make': 'BMW', 'model': '911', 'price': '2868.00', 'year': '1919'},
{'ID': 13, 'color': 'white', 'make': 'Audi', 'model': 'R8', 'price': '2907.00', 'year': '1991'},
{'ID': 14, 'color': 'red', 'make': 'BMW', 'model': 'Astra', 'price': '1819.00', 'year': '1961'},
{'ID': 15, 'color': 'blue', 'make': 'VW', 'model': 'A3', 'price': '1380.00', 'year': '1991'},
{'ID': 16, 'color': 'silver', 'make': 'Porsche', 'model': 'Astra', 'price': '2233.00', 'year': '1961'},
{'ID': 17, 'color': 'silver', 'make': 'Mercedes', 'model': 'Astra', 'price': '1048.00', 'year': '1919'}
], 'Success': 'returned 12 values from 0'}{
"model": String,
"make": String,
"year": Integer,
"color": String,
"price": Float
} {'Success': 'Resource created'}- 201
- 422
{
'id': 29,
'data':{
"model": String,
"make": String,
"year": Integer,
"color": String,
"price": Float
}
} {'Success': 'Resource updated'}- 200
- 404
- 405
- 422
{
"id":Integer,
} {'Failure': 'id 52 is not in Resource'}- 204
- 404
- 422
There are only the four basic HTTP Verbs implemented, use of any other verb such as HEAD, PATCH or similar will result in an 501 Error.
{'Failure': 'Method not Implemented'}- 501
One big drawback of being self-taught is that there are some conventions that one is not necessarily aware of, therefore i have to assume that i broke conventions in this project more then once. In the case that i would be hired for the position I would familiarize myself better with the relevant conventions. I commented on a few places within the code why i made certain choices, never the less i want to provide a more detailed explanation for a few of them.
Inspiration for the Structure of the App and the Endpoint object was the way modern telephone systems work, calls need to be switched between different endpoints. Assuming that the callers don't not need to be connect to a specific end, just to a 'function', like for as example in a call center.
To visualize how that is relevant to our API, we think about our endpoints like /get/cars/ as Agents and our Endpoints class a trunk connecting calls. A call would come in through the trunk before it moves through a switch to an extension that handles the before specified problem of the Customer. If we now think about resources on the database like departments in our center, it becomes apparent, that it would be wise to layout the center in a way that each department is served by its own trunk and switch.
Similarly to this Idea is the App layed out, each resource gets its own Endpoints object containing a switch.
Because there is only on resource handled in the App, i decided against building a parent class and wrote an implementation that is specific to the resource at hand. If the need to handle multiple resources would arise, it would be possible to abstract the Endpoint class further so that it can be used as a parent class for resource specific Endpoints Objects to inherit from.
To Implement the Trunk/switch idea i wrote a __call__ () method containing a switch case, which also allowed me to return a 501Error if needed.The switch case moves requests to a staticmethod that handles the request, after the request got validated by the Validator
The Validator moves trough the requirements of the method that is validating and returns False when a requirement is not met or True if none of the checks failed.