Calculating invoices for customer billing

readme.md

Background

In the past, we provided some raw billing data in JSON format to the finance team, which they used to manually generate monthly invoices for our customers. Recently, they’ve asked us to create some automation to make this process less error-prone.

Instructions

Your goal is to implement the bill_for function to calculate the total monthly bill for a customer.

Customers are billed based on their subscription tier for the month. We charge them a prorated amount for each user who was active during that month.

You talked with the other engineers on the team and decided that the following algorithm would work:

1. Calculate a daily rate for the active subscription tier
2. For each day of the month, identify which users were active that day
3. Multiply the number of active users for the day by the daily rate to calculate the total for the day
4. Return the running total for the month at the end, rounded to 2 decimal places

Parameters

This billing function accepts the following parameters:

• month

Always present. Has the following structure: "2019-01" # January 2019 in YYYY-MM format

• active_subscription May be None. If present, has the following structure:

{
  'id': 1,
  'customer_id': 1,
  'monthly_price_in_dollars': 4  # price per active user per month
}

• users May be empty, but not None. Has the following structure:

[
  {
    'id': 1,
    'name': 'Employee #1',
    'customer_id': 1,

    # when this user started
    'activated_on': datetime.date(2018, 11, 4),

    # last day to bill for user
    # should bill up to and including this date
    # since user had some access on this date
    'deactivated_on': datetime.date(2019, 1, 10)
  },
  {
    'id': 2,
    'name': 'Employee #2',
    'customer_id': 1,

    # when this user started
    'activated_on': datetime.date(2018, 12, 4),

    # hasn't been deactivated yet
    'deactivated_on': None
  }
]

• Return value This function should return the total monthly bill for the customer, rounded to 2 decimal places.

Calculation Examples

Here is an example of calculating the amount to bill each customer using the algorithm described above. This example is captured by the "works when a user is activated during the month" test.

Month	2019-01
Subscription tier	$4 / month
User activations	2018-11-04
2018-12-04
2019-01-10 (new this month)
User deactivations	None
Daily rate is $0.129032258   (this is the monthly rate of $4.00 / 31 days in January) 

2019-01-01  2 active users * $0.129032258 = $0.258064516  (subtotal: $0.258064516)
2019-01-02  2 active users * $0.129032258 = $0.258064516  (subtotal: $0.516129032)
...
2019-01-09  2 active users * $0.129032258 = $0.258064516  (subtotal: $2.322580645)
2019-01-10  3 active users * $0.129032258 = $0.387096774  (subtotal: $2.709677419)
2019-01-11  3 active users * $0.129032258 = $0.387096774  (subtotal: $4.387096772)
...
2019-01-30  3 active users * $0.129032258 = $0.387096774  (subtotal: $10.451612903)
2019-01-31  3 active users * $0.129032258 = $0.387096774  (subtotal: $10.838709677)

Total = $10.84 (round subtotal to nearest cent)

Testing

The automated tests we provide only cover a few key cases, so you should plan to add some of your own tests or modify the existing ones to ensure that your solution handles any edge cases. You should be able to follow the existing patterns for naming and constructing tests to add your own.

Notes / Edge cases

It’s more important for the return value to be correct than it is for the algorithm to be highly optimized. You can store intermediate results as any kind of decimal type (e.g. float, double). You do not need to round values until the last step. You should bill for all days between and including both the activation and deactivation dates since the user still had some access on the last day. You should not change function names or return types of the provided functions since our test cases depend on those not changing. Time guidance Aim to take between 25-45 minutes for this challenge. Set a timer now for 45 minutes to gauge the time you have spent on this challenge.

If you have spent 45 or more minutes:

Stop working Document where you are at in the "Your Notes" section in the upper right corner Go on to the next challenge. It is independent from any solution you got on this challenge. You can return to any incomplete challenges if you have time left over at the end.

Solution.py

import datetime
import calendar

def bill_for(month, active_subscription, users):
    # your code here!
    month = datetime.datetime.strptime(month, '%Y-%m')
    month = datetime.date(month.year, month.month, 1)
    # first day of month
    first_day = first_day_of_month(month)
    # last day of month
    last_day = last_day_of_month(month)

    # get the number of days in the month
    days_in_month = (last_day - first_day).days + 1

    if active_subscription is None:
        return 0
        
    daily_rate = 0
    cost = active_subscription["monthly_price_in_dollars"]
    daily_rate += float(cost) / days_in_month
    total_cost = 0

    for user in users:
        d_date = user["deactivated_on"]
        a_date = user["activated_on"] 
        
        if d_date is not None:
            if d_date < month:
            # get the number of days between the deactivated date and the first day of the month
              total_cost += 0
            else:
              no_of_days_active = abs((d_date - first_day)).days + 1
              total_cost += no_of_days_active * daily_rate


        if d_date is None:
          # if the user is active in the month
          if a_date >= first_day and a_date <= last_day:
            # get the number of days between the activated date and the first day of the month
            no_of_days_active = abs((a_date - last_day)).days + 1
            total_cost += no_of_days_active * daily_rate
            
          # if the user didn't cancel the subscription
          else:
             total_cost += days_in_month * daily_rate

    return round(total_cost, 2)


####################
# Helper functions #
####################

def first_day_of_month(date):
  """
  Takes a datetime.date object and returns a datetime.date object
  which is the first day of that month. For example:

  >>> first_day_of_month(datetime.date(2019, 2, 7))  # Feb 7
  datetime.date(2019, 2, 1)                          # Feb 1

  Input type: datetime.date
  Output type: datetime.date
  """
  return date.replace(day=1)

def last_day_of_month(date):
  """
  Takes a datetime.date object and returns a datetime.date object
  which is the last day of that month. For example:

  >>> last_day_of_month(datetime.date(2019, 2, 7))  # Feb  7
  datetime.date(2019, 2, 28)                        # Feb 28

  Input type: datetime.date
  Output type: datetime.date
  """
  last_day = calendar.monthrange(date.year, date.month)[1]
  return date.replace(day=last_day)

def next_day(date):
  """
  Takes a datetime.date object and returns a datetime.date object
  which is the next day. For example:

  >>> next_day(datetime.date(2019, 2, 7))   # Feb 7
  datetime.date(2019, 2, 8)                 # Feb 8

  >>> next_day(datetime.date(2019, 2, 28))  # Feb 28
  datetime.date(2019, 3, 1)                 # Mar  1

  Input type: datetime.date
  Output type: datetime.date
  """
  return date + datetime.timedelta(days=1)