Transaction Fee Calculator ## The Challenge You need to build a system that calculates fees for a payment platform. You will get transaction data as a long CSV string. You need to read this data and

Transaction Fee Calculator ## The Challenge You need to build a system that calculates fees for a payment platform. You will get transaction data as a long CSV string. You need to read this data and calculate the fee for each payment based on specific rules. There are three parts to this problem. We start with simple math and add harder rules later. 1.

Basic Fees: Read the data and calculate cost based on the payment method. 2.

New Rules: Only charge fees for successful payments. Use different rates for different countries. 3.

Discounts: Give discounts to merchants who process a lot of payments.

Input Data The input is a single string in CSV format. The first row lists the column names. The rows after that are the actual transaction records.

id,reference,amount,currency,date,merchant_id,buyer_country,transaction_type,payment_provider,status

What the columns mean: * id: The unique ID for the transaction. * reference: Reference number. * amount: Money involved in cents (e.g., 1000 = $10.00). * currency: The money type (e.g., "eur", "usd"). * date: When it happened (YYYY-MM-DD). * merchant_id: ID for the seller. * buyer_country: Country code of the buyer. * transaction_type: What kind of action it is (e.g., "payment"). * payment_provider: How they paid (e.g., "card", "klarna"). * status: Did it work? (e.g., "payment_completed", "payment_failed"). * * * ## Part 1: Basic Fee Calculation ### The Task Write a function to read the CSV string. Calculate the fee for each transaction based on the payment_provider. Each provider charges a percentage of the amount plus a fixed fee.

Fee Table: Payment Provider Fee Rate card 2.9% + 30 cents klarna 3.5% + 50 cents bank_transfer 0.8% flat You must round the final fee down to the nearest whole number (integer).

Example Data

Input:

id,reference,amount,currency,date,merchant_id,buyer_country,transaction_type,payment_provider,status py_1,1,1000,eur,2024-12-24,acct_1,ie,payment,card,payment_completed py_2,2,2500,eur,2024-12-24,acct_2,de,payment,card,payment_failed py_3,3,3400,eur,2024-12-25,acct_2,ie,payment,klarna,payment_completed py_4,4,5000,eur,2024-12-25,acct_1,fr,payment,bank_transfer,payment_completed

Output:

id,transaction_type,payment_provider,fee py_1,payment,card,59 py_2,payment,card,102 py_3,payment,klarna,169 py_4,payment,bank_transfer,40

How we got these numbers: * py_1: 1000 × 0.029 + 30 = 59 * py_2: 2500 × 0.029 + 30 = 102.5. Round down to 102. * py_3: 3400 × 0.035 + 50 = 169 * py_4: 5000 × 0.008 = 40 ### What You Need To Do * Read the CSV string properly (ignore the header row after reading it). * Use the right math formula for each provider. *

Return the result as a CSV string. * Round fees down (use integer conversion). * * * ## Part 2: Conditional Fee Rules ### New Rules Now, update your code to handle status checks and countries. 1.

Check Status: Only calculate a fee if status is "payment_completed". 2.

Failed Payments: If the status is failed or pending, the fee is 0. 3.

Regional Rates: Ireland ("ie") has special, cheaper rates.

Special Rates for Ireland (ie): Payment Provider Fee Rate card 1.9% + 20 cents klarna 2.5% + 40 cents All other countries use the standard rates from Part 1.

Example Data

Input:

id,reference,amount,currency,date,merchant_id,buyer_country,transaction_type,payment_provider,status py_1,1,1000,eur,2024-12-24,acct_1,ie,payment,card,payment_completed py_2,2,2500,eur,2024-12-24,acct_2,de,payment,card,payment_failed py_3,3,3400,eur,2024-12-25,acct_2,ie,payment,klarna,payment_completed py_4,4,5000,eur,2024-12-25,acct_1,fr,payment,bank_transfer,payment_completed py_5,5,2000,eur,2024-12-26,acct_1,ie,payment,card,payment_pending

Output:

id,transaction_type,payment_provider,fee py_1,payment,card,39 py_2,payment,card,0 py_3,payment,klarna,125 py_4,payment,bank_transfer,40 py_5,payment,card,0

How we got these numbers: * py_1: Ireland rate used. (Cheaper than standard). * py_2: Failed, so fee is 0. * py_3: Ireland rate used. * py_4: France uses standard rate. * py_5: Pending, so fee is 0. ### What You Need To Do * Look at the status column before doing math. * Check the buyer_country. If it is "ie", use the special rates. * Keep your code clean. Separate the logic for checking status and calculating fees. * * * ## Part 3: Volume-Based Discounts ### Adding Discounts Now we add volume discounts. Merchants who sell more get cheaper fees. You need to count how many successful transactions a merchant has made.

Discount Table: Total Transactions so far Discount 1-10 0% (Normal price) 11-50 10% off 51-100 15% off 101+ 20% off

Configuration Data: You will get a dictionary with fees for many countries. It looks like this:

python country_fees = { "ie": {"card": (0.019, 20), "klarna": (0.025, 40), "bank_transfer": (0.006, 0)}, "de": {"card": (0.025, 25), "klarna": (0.030, 45), "bank_transfer": (0.007, 0)}, "fr": {"card": (0.027, 28), "klarna": (0.032, 48), "bank_transfer": (0.008, 0)}, "default": {"card": (0.029, 30), "klarna": (0.035, 50), "bank_transfer": (0.008, 0)} }

Example Data

Input: Imagine acct_1 has done 10 transactions already. The 11th transaction (py_11) happens. The 12th transaction (py_12) happens.

Output: *

Transaction 11: Base fee is calculated. Then, apply a 10% discount. *

Transaction 12: Base fee is calculated. Apply a 10% discount.

Math Detail: For py_11 (Ireland Card): Base = 1500 × 0.019 + 20 = 48.5. Discount = 10%. New Fee = 48.5 × 0.9 = 43.65. Round down -> 43. ### What You Need To Do * Keep a count of transactions for every merchant. * Check the merchant's count before the current transaction to find the discount. * Only increase the count if the transaction is "payment_completed". * Process rows in order. The order in the CSV matters. * Use the provided dictionary for country fees. If the country isn't there, use "default". ### Questions to Ask the Interviewer * Do I apply the discount before rounding or after? * What if a merchant sells in different countries? Does that count towards the same volume total? * Does the volume count ever reset (like every month)? * * * ## How to Solve It ### Part 1 Solution

Plan: 1. Split the CSV string by lines () and commas (,). 2. Ignore the first line (headers). 3. Look at the payment_provider column. 4. Do the math and save the result.

Complexity: O(n) (linear time). We look at each line once.

Code:

python def calculate_fees(csv_data): lines = csv_data.strip().split('
') header = lines[0].split(',') # Map column names to indices col_idx = {name: i for i, name in enumerate(header)} fee_config = { 'card': (0.029, 30), 'klarna': (0.035, 50), 'bank_transfer': (0.008, 0) } results = ['id,transaction_type,payment_provider,fee'] for line in lines[1:]: fields = line.split(',') tx_id = fields[col_idx['id']] tx_type = fields[col_idx['transaction_type']] provider = fields[col_idx['payment_provider']] amount = int(fields[col_idx['amount']]) rate, fixed = fee_config[provider] fee = int(amount * rate + fixed) # floor by int conversion results.append(f'{tx_id},{tx_type},{provider},{fee}')

**return** '
'.join(results)

Watch Out For: * Empty inputs. * Providers that are not in your list. * Bad data in the CSV. * * * ### Part 2 Solution

Plan: 1. Check if status is "payment_completed". If not, fee is 0. 2. Create a function to pick the right fees. It checks the country first. 3. If the country is missing, use default fees.

Code:

python def calculate_fees(csv_data): lines = csv_data.strip().split('
') header = lines[0].split(',') col_idx = {name: i for i, name in enumerate(header)} # Default fees default_fees = { 'card': (0.029, 30), 'klarna': (0.035, 50), 'bank_transfer': (0.008, 0) } # Regional overrides regional_fees = { 'ie': { 'card': (0.019, 20), 'klarna': (0.025, 40) } } def get_fee_config(country, provider): if country in regional_fees and provider in regional_fees[country]:

**return** regional_fees[country][provider]

**return** default_fees[provider] results = ['id,transaction_type,payment_provider,fee'] for line in lines[1:]: fields = line.split(',') tx_id = fields[col_idx['id']] tx_type = fields[col_idx['transaction_type']] provider = fields[col_idx['payment_provider']] amount = int(fields[col_idx['amount']]) status = fields[col_idx['status']] country = fields[col_idx['buyer_country']] if status != 'payment_completed': fee = 0 else: rate, fixed = get_fee_config(country, provider) fee = int(amount * rate + fixed) results.append(f'{tx_id},{tx_type},{provider},{fee}')

**return** '
'.join(results)

• • • Part 3 Solution

Plan: 1. Use a dictionary (defaultdict) to count how many times each merchant appears. 2. Before doing math for a row, check the merchant's current count. 3. Find the discount percent based on that count. 4. Calculate the fee, apply the discount, then round. 5. If the transaction was successful, add 1 to the merchant's count.

Code:

python from collections import defaultdict def calculate_fees(csv_data, country_fees): lines = csv_data.strip().split('
') header = lines[0].split(',') col_idx = {name: i for i, name in enumerate(header)} # Volume discount tiers: (min_transactions, discount_rate) discount_tiers = [(101, 0.20), (51, 0.15), (11, 0.10), (1, 0.00)] def get_discount(tx_count): for threshold, discount in discount_tiers: if tx_count >= threshold:

**return** discount return 0.0 def get_fee_config(country, provider): if country in country_fees:

**return** country_fees[country].get(provider, country_fees['default'][provider])

**return** country_fees['default'][provider] merchant_volume = defaultdict(int) results = ['id,transaction_type,payment_provider,fee'] for line in lines[1:]: fields = line.split(',') tx_id = fields[col_idx['id']] tx_type = fields[col_idx['transaction_type']] provider = fields[col_idx['payment_provider']] amount = int(fields[col_idx['amount']]) status = fields[col_idx['status']] country = fields[col_idx['buyer_country']] merchant_id = fields[col_idx['merchant_id']] if status != 'payment_completed': fee = 0 else: # Get current volume (before incrementing) current_volume = merchant_volume[merchant_id] + 1 discount = get_discount(current_volume) rate, fixed = get_fee_config(country, provider) base_fee = amount * rate + fixed fee = int(base_fee * (1 - discount)) # Increment volume count for completed transactions merchant_volume[merchant_id] += 1 results.append(f'{tx_id},{tx_type},{provider},{fee}')

**return** '
'.join(results)

Performance: *

Time Complexity: O(n). We still process each line once. *

Space Complexity: O(n + m). We store the output string (n) and the count for each merchant (m).