Java 8 & Streams · Chapter 07

Collectors

Turning streams into results — toList, toSet, toMap (and duplicate keys), joining, counting, summing, averaging and summarizing.

◷ 7 min read§ 14 sections? 5 interview questions{ } 19 code examples

#Introduction

Up until now you've learned how to filter, transform, sort, de-duplicate and reduce data. Now you'll learn how to collect the results — which is what you'll do in 90% of real Spring Boot applications. Almost every REST API pipeline ends with .collect(...).

By the end of this chapter you'll master collect(), Collectors.toList(), toSet(), toMap(), joining(), counting(), summingInt(), averagingInt() and summarizingInt().

#What is collect()?

Everything you've learned so far produced a Stream:

Java

numbers.stream()
       .filter(n -> n % 2 == 0)   // still a Stream

If you want a result like a List, Set, Map, String or statistics, you use collect().

Stream

Collector

Final container

#Collectors.toList()

The most commonly used collector.

Java

List<Integer> even = numbers.stream()
        .filter(n -> n % 2 == 0)
        .collect(Collectors.toList());

Output

[2, 4, 6, 8]

In Java 16+ you'll also see:

Java

List<Integer> even = numbers.stream()
        .filter(n -> n % 2 == 0)
        .toList();

#Collectors.toSet()

Removes duplicates automatically.

Java

List<Integer> numbers = Arrays.asList(2, 5, 2, 8, 5);

Set<Integer> set = numbers.stream()
        .collect(Collectors.toSet());

System.out.println(set);

Output

[2, 5, 8]

#Collectors.toMap()

One of the most asked interview topics.

Java

class Employee {
    int id;
    String name;
}

Create 1 → John, 2 → Alex, 3 → David:

Java

Map<Integer, String> map = employees.stream()
        .collect(Collectors.toMap(
                Employee::getId,
                Employee::getName
        ));

Output

{1=John, 2=Alex, 3=David}

Duplicate Key Problem

Suppose two employees share id 1 (John and Alex):

Java

Collectors.toMap(Employee::getId, Employee::getName)

Output

IllegalStateException: Duplicate key 1 (attempted merging values John and Alex)

Solution — provide a merge function:

Java

// Keep first
Collectors.toMap(Employee::getId, Employee::getName, (existing, newValue) -> existing)

// Keep last
Collectors.toMap(Employee::getId, Employee::getName, (existing, newValue) -> newValue)

#Collectors.joining()

Join strings.

Java

List<String> names = Arrays.asList("John", "Alex", "David");

names.stream().collect(Collectors.joining());               // JohnAlexDavid
names.stream().collect(Collectors.joining(", "));           // John, Alex, David
names.stream().collect(Collectors.joining(", ", "[", "]")); // [John, Alex, David]

Form	Output
`joining()`	`JohnAlexDavid`
`joining(", ")`	`John, Alex, David`
`joining(", ", "[", "]")`	`[John, Alex, David]`

#Collectors.counting()

Instead of numbers.stream().count(), you can do:

Java

numbers.stream()
       .collect(Collectors.counting());

Output

Usually you'll see counting() inside grouping operations, covered in the next chapter.

#Summing, Averaging and Summarizing

summingInt()

Java

int total = employees.stream()
        .collect(Collectors.summingInt(Employee::getAge));

Output

averagingInt()

Java

double avg = employees.stream()
        .collect(Collectors.averagingInt(Employee::getAge));

Output

26.4

summarizingInt()

Java

IntSummaryStatistics stats = employees.stream()
        .collect(Collectors.summarizingInt(Employee::getAge));

System.out.println(stats);

Output

IntSummaryStatistics{count=5, sum=132, min=20, average=26.400000, max=35}

Instead of separate sum(), average(), max(), min() and count() calls, you get everything in one pass (stats.getMax(), stats.getAverage(), …).

Product Example

Java

class Product {
    String name;
    double price;
}

double total = products.stream()
        .collect(Collectors.summingDouble(Product::getPrice));

double avg = products.stream()
        .collect(Collectors.averagingDouble(Product::getPrice));

DoubleSummaryStatistics stats = products.stream()
        .collect(Collectors.summarizingDouble(Product::getPrice));

#Full Pipeline Example

Requirement: salary > 50k → sort descending → get names → join by comma.

Java

String names = employees.stream()
        .filter(e -> e.getSalary() > 50000)
        .sorted(Comparator.comparing(Employee::getSalary).reversed())
        .map(Employee::getName)
        .collect(Collectors.joining(", "));

With Alex 70000, David 60000, John 50000:

Output

Alex, David

John is excluded because 50000 > 50000 is false. This is production-level code.

#Real Spring Boot Examples

Java

// Employee IDs
List<Integer> ids = employees.stream()
        .map(Employee::getId)
        .toList();

// Employee names
Set<String> names = employees.stream()
        .map(Employee::getName)
        .collect(Collectors.toSet());

// ID → Employee map
Map<Integer, Employee> map = employees.stream()
        .collect(Collectors.toMap(Employee::getId, Function.identity()));

// CSV export
String csv = employees.stream()
        .map(Employee::getName)
        .collect(Collectors.joining(","));

What is Function.identity()?

Instead of writing e -> e, you can write Function.identity() — both mean "use the object itself as the value." So Collectors.toMap(Employee::getId, Function.identity()) creates:

Output

1 -> Employee{id=1, name=John}
2 -> Employee{id=2, name=Alex}

A very common interview question.

#collect() vs reduce()

collect()	reduce()
Produces a container	Produces one value
List, Set, Map, statistics	Sum, product, max, min

Remember: collect() builds a collection; reduce() builds a single result.

#Interview Questions

Q1Why use Collectors.toMap() instead of a loop?

Cleaner, declarative, and integrates naturally into Stream pipelines.

Q2What happens if duplicate keys exist?

IllegalStateException, unless you provide a merge function.

Q3Difference between count() and Collectors.counting()?

count() is a terminal Stream operation.
counting() is a collector and is especially useful inside groupingBy().

Q4When should you use summarizingInt()?

When you need count, sum, average, min and max in one traversal.

Q5What does Function.identity() do?

Returns the object itself. It's equivalent to x -> x.

#Practice Problems

Java

List<Integer> numbers = Arrays.asList(2, 5, 8, 2, 10, 5, 12);

Collect even numbers into a List.
Collect unique numbers into a Set.
Join all numbers into a comma-separated String.
Count the elements using Collectors.counting().
Get summary statistics using summarizingInt().

Java

class Employee {
    int id;
    String name;
    String department;
    int age;
    double salary;
}

Get a List<String> of employee names.
Get a Set<String> of departments.
Create a Map<Integer, Employee> using ID as the key.
Create a Map<Integer, String> of ID → name.
Find the total salary.
Find the average salary.
Find salary statistics.
Join all employee names with " | ".
Create a map with duplicate IDs handled by keeping the latest employee.
Create a map using Function.identity().

Senior Interview Challenge (5+ Years)

Filter employees with salary > 60,000, sort by salary descending, map to names, and join with " -> ":

Java

String result = employees.stream()
        .filter(e -> e.getSalary() > 60_000)
        .sorted(Comparator.comparing(Employee::getSalary).reversed())
        .map(Employee::getName)
        .collect(Collectors.joining(" -> "));

#Chapter Summary

collect() and the Collector concept
toList() / toSet() (and Java 16 Stream.toList())
toMap() and handling duplicate keys
joining() with separators, prefix and suffix
counting(), summingInt(), averagingInt(), summarizingInt()
Function.identity()
collect() vs reduce()