Problem

In the Ancestry Marketing mobile app, analytics events arrive as pipe-delimited strings. Implement a function that both parses these events and applies business rules to compute the total revenue from valid purchase events.

Each event string has the format: timestamp|event_type|user_id|campaign|amount

Your task is to process a list of event strings and return the sum of amounts for valid purchase events.

Business Rules

1. Only events with event_type == "purchase" count toward revenue.
2. Ignore malformed records that do not have exactly 5 fields.
3. Ignore records where timestamp or amount is not a valid integer.
4. Ignore records where amount <= 0.
5. Deduplicate purchases by (timestamp, user_id). If the same purchase appears multiple times, count it only once.

Input / Output

• Input: events, a list of strings
• Output: an integer representing total valid revenue

Examples

Example 1 Input: events = ["100|purchase|u1|summer|20", "101|open|u1|summer|0", "100|purchase|u1|summer|20"] Output: 20

Explanation: Only purchase events count, and the duplicate purchase at timestamp 100 for user u1 is counted once.

Example 2 Input: events = ["200|purchase|u2|dna|15", "bad|purchase|u3|dna|10", "201|purchase|u4|dna|-5"] Output: 15

Explanation: The second record has an invalid timestamp, and the third has a non-positive amount.

Constraints

• 1 <= len(events) <= 10^4
• Each event string length is between 1 and 200
• timestamp and amount, when valid, fit in 32-bit signed integers