Compute balances, rejections, and platform reserve

You are given a set of transactions in CSV-like format with columns: - `account_name`, `timestamp`, `currency`, `amount` Assume: - Transactions for all accounts can be interleaved; you must process them in **ascending timestamp order** (if input is already ordered, you may process in given order). - All amounts are integers in the smallest currency unit (e.g., cents). - Treat each `(account_name, currency)` as a separate balance unless stated otherwise. ## Part 1 — Non-zero ending balances Compute each account’s ending balance (sum of its amounts). Output all accounts whose ending balance is non-zero, along with the balance. ## Part 2 — Reject transactions that would make balance negative Process transactions in time order while maintaining balances. - If applying a transaction would make the account’s balance negative, **reject** it (do not apply it). - Collect rejected transactions as strings in the original row format (or a clear equivalent such as `"acct,timestamp,currency,amount"`). Output: 1. The rejected transaction list. 2. All accounts with non-zero ending balances after applying only accepted transactions. ## Part 3(a) — Allow borrowing from a platform account You are also given a `platform_id`. Process transactions in time order with this additional rule: - For any **non-platform** account, if after applying a transaction its balance would become negative, the account may “borrow” from the platform to bring it back up to 0 **as long as the platform has sufficient funds**. - Borrow amount = `-(new_balance)`. - Decrease the platform’s balance by the borrowed amount. - The transaction is accepted if borrowing succeeds. - If the platform does not have enough balance to cover the borrow, the transaction is rejected (and not applied). Additionally compute: - `max_reserve`: the **maximum total amount borrowed from the platform at any point in time** (i.e., peak borrowed amount outstanding). You should define and track this consistently based on your borrowing model. Output for Part 3(a): 1. `max_reserve` 2. `rejected_transactions` 3. All accounts with non-zero ending balances after processing. ## Notes - Clearly document how you compute `max_reserve` (e.g., peak cumulative borrowed amount across accounts over time). - If multiple currencies exist, state whether borrowing is allowed only within the same currency (recommended) or across currencies (not recommended). Implement the chosen assumption consistently.