Filter After Sort¶

Rule details


Rule ID	`filter-after-sort`
Severity	INFO — worth knowing, may not matter at your scale
Confidence	HIGH — structurally proven
Category	Sort abuse
Complexity	O(n log n + k) → O(k log k + n)

Description¶

Detects stream or collection pipelines where filter() comes after sorted(). Sorting first processes all N elements at O(n log n), then filter discards some. Filtering first reduces the input to k elements, making the sort O(k log k) where k ≤ n.

Typical¶

List<Order> paidByTotal = orders.stream()
    .sorted(Comparator.comparing(Order::getTotal).reversed())
    .filter(o -> "PAID".equals(o.getStatus()))  // Wasted effort: sorted items that get filtered out
    .collect(Collectors.toList());

After¶

List<Order> paidByTotal = orders.stream()
    .filter(o -> "PAID".equals(o.getStatus()))  // Filter first: fewer items to sort
    .sorted(Comparator.comparing(Order::getTotal).reversed())
    .collect(Collectors.toList());

How Detection Works¶

Looks for SortCall nodes followed by LookupCall(FILTER) in the same pipeline chain
Works for both Java streams and collection method chains

Suggestion¶

Move filter() before sorted() to sort fewer elements