Thinking Ahead: Preconditions, Caching and Reuse

An input can be user-entered data, a file, a sensor reading or another routine’s result. An output can be displayed text, a stored record, a returned value or an action signal. For a class-average calculation, specify which scores are included, their valid range and how missing results are represented before deciding how to divide the total.

A precondition is something assumed to hold before an operation, such as a sorted input array for binary search or a non-empty collection for a simple mean. Validation can enforce a condition at a boundary, but the condition itself is part of the contract. Decide what happens when it does not hold: reject, transform or use a different operation under a stated rule.

Caching retains a result or data copy so a later request can avoid repeated work or slower access. A useful cache key includes everything that determines the result. If price depends on item and catalogue version, caching by item alone can return a stale price after the catalogue changes. A higher hit rate is useful only if returned answers still meet the requirements.

The demonstration deliberately caches a lookup into mutable data. After the source price changes, the cached result stays old until cleared. Real policies can use versioned keys, expiry or explicit invalidation. A cache consumes space, has lookup and maintenance costs and provides no benefit for a result never reused. Explain whether stale data are acceptable in the scenario.

Reusable components can reduce duplicate work and concentrate testing around a stable interface. Specify parameter types, allowed ranges, outputs, failure responses and side effects. A component tested for positive integers is not automatically suitable for arbitrary signed decimals without further design and tests.

Reuse introduces dependencies and compatibility obligations. A library update can change behaviour or supported platforms, so retain suitable version and test evidence. Do not equate reuse with copying an unexplained block into a project. Understanding and adapting an interface is necessary for maintenance, and independently assessed work still needs proper acknowledgement of external material and the learner’s own contribution.

from functools import lru_cache
prices = {"A": 12}

@lru_cache(maxsize=16)
def lookup_price(item):
    return prices[item]

print(lookup_price("A"))
prices["A"] = 15
print(lookup_price("A"))
lookup_price.cache_clear()
print(lookup_price("A"))

Attempt the task before opening its worked response. Record each change of state, explain the rule used, and check the result against the original requirements. Code examples labelled Python 3 use that language’s conventions; pseudocode states its own assumptions.