Sorting Before limit
You'll learn why the order of sorted() and limit() in your stream pipeline dramatically affects both correctness and performance.
Why Sorted Must Complete Before Limit
When you place sorted() before limit() in a pipeline, the sorting operation must examine and order the entire stream before limit() can select the top N elements. This is unavoidable because:
- To know which elements are the "first 3" after sorting, you must sort everything first
sorted()is a stateful operation that buffers all elements internally- Unlike
filterormap, sorting can't produce output until it's seen all inputs
The Performance Impact
Consider finding the 5 smallest numbers from a million-element stream:
stream()
.sorted() // Processes all 1,000,000 elements
.limit(5) // Then takes first 5
Even though you only want 5 results, sorted() must:
- Consume the entire stream (1 million elements)
- Sort them all (expensive: O(n log n))
- Only then can
limit()take the first 5
The Analogy: Imagine you need the 3 shortest pencils from a box of 1,000. If you fully sort all pencils by length first, you're doing far more work than necessary. (A partial sort or priority queue would be smarter, but Stream's sorted() doesn't optimize this automatically.)
When This Pattern Is Correct
This ordering is necessary when you want "top N after sorting":
employees.stream()
.sorted(Comparator.comparing(Employee::getSalary))
.limit(10) // Lowest 10 salaries
Key Takeaway: sorted() before limit() forces full sorting of the entire stream, which is correct for "top N" queries but expensive—always consider whether you truly need sorting before limiting.