Exercise 2: Reconstructing an Order Count Over a Threshold, Purely From the Log — Possible Solution
====================================================================

THE NEW REPLAY FUNCTION
------------------------------
    def replay_orders_over(event_log, threshold):
        count = 0
        for event_type, payload in event_log.events:
            if event_type == 'OrderPlaced' and payload['items_total'] > threshold:
                count += 1
        return count

Follows this chapter's own replay_total_revenue() shape exactly - a
single pass over event_log.events, filtering by event_type, with no
separate counter maintained anywhere outside this one function call.

VERIFYING AGAINST THIS CHAPTER'S OWN THREE-EVENT LOG
------------------------------
Using the same three OrderPlaced events this chapter's own
replay_total_revenue() example used (items_total of 100, 50, and 75),
called with threshold=60:

    orders over 60: 2
    manual check: 100>60 and 75>60, 50 is not -> 2

100 and 75 both exceed 60; 50 does not - the function correctly counts
2, matching a manual check of the same three values.

WHY THIS CONFIRMS EVENT SOURCING'S OWN GENERALITY
------------------------------
This chapter's own replay_total_revenue() answered one specific
question (what's the total?) by replaying the log. This exercise
answers a COMPLETELY DIFFERENT question (how many orders exceeded a
threshold?) using the exact same log, with no changes to how events
were originally published or stored. This is the real payoff of event
sourcing: the log doesn't need to anticipate every question that will
ever be asked of it in advance - new replay functions can answer new
questions against data that's already there.

WHY THIS WORKS AS AN ANSWER
------------------------------
The new function mirrors this chapter's own established replay
pattern exactly, and it's verified against the identical event data
this chapter's own example used, with the result cross-checked by a
manual count rather than only asserted correct.