This is the Java Stream API, introduced in Java 8 - not the I/O package streams.

Not a tutorial - just some examples.

Filter Map Reduce

Example (taken from the package documentation):

Here we use widgets, a Collection<Widget>, as a source for a stream, and then perform a filter-map-reduce on the stream to obtain the sum of the weights of the red widgets:

Java


1
2
3
4


int sum = widgets.stream()
        .filter(b -> b.getColor() == RED)
        .mapToInt(b -> b.getWeight())
        .sum();

Group by a Key

Convert a vararg list of Map<Integer, MyObject>s to Map<Integer, Set<MyObject>>:

Java


 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17


import java.util.Map;
import java.util.Set;
import java.util.stream.Stream;
import java.util.stream.Collectors;

...

Map<Integer, Set<String>> collected = Stream.of(
        Map.entry(123, "foo"),
        Map.entry(123, "bar"),
        Map.entry(234, "baz"))
        .collect(Collectors.groupingBy(
                Map.Entry::getKey,
                Collectors.mapping(
                        Map.Entry::getValue, Collectors.toSet()
                )
        ));

The classifier in the above groupingBy is simply the key of the source map: Map.Entry::getKey. This determines which group each element will belong to.

Then there is a downstream mapping: Collectors.mapping(...). This controls how the elements are transformed before being saved: each input value is placed in a set.

The result:

{234=[baz], 123=[bar, foo]}

The above example used groupingBy(). There is also groupingByConcurrent() which performs the same grouping but collects its data into a ConcurrentHashMap instead of HashMap:

Not at all relevant in my small example above - but could be, for larger data sets where concurrent/parallel execution across multiple threads may make a performance difference.

Group by Two Fields

Nested groupings populate the nested map:

Java


1
2
3
4


final Map<Employee.Gender, Map<Integer, List<Employee>>> groupByGenderAndAge = employees.stream()
                .collect(groupingBy(Employee::getGender,
                         groupingBy(Employee::getAge))
                );

Sort a List/Array

Natural ordering is assumed (no comparator used).

For a list of objects:

Java


1
2
3
4
5
6


public static <T> List<T> sortList(List<T> unsortedList) {
    return unsortedList
            .stream()
            .sorted()
            .collect(Collectors.toList());
}

For an array of primitives:

Java


1
2
3
4


return Arrays.stream(unsortedArray)
        .boxed()
        .sorted()
        .collect(Collectors.toList());

Count Items in a List

This includes distinct(), so it provides a count of unique values:

Java

1

items.stream().distinct().count();

Get List of Fields from a List

We start with a list of objects - in this case courses, where each course has a name and a class room number:

Java


1
2
3
4


List<Course> courses = new ArrayList<>();
courses.add(new Course("Math", "123"));
courses.add(new Course("Physics", "456"));
courses.add(new Course("Chemistry", "789"));

Then we build a new list containing just the room numbers. In this example, we then count the number of distinct room numbers:

Java


1
2
3
4
5
6
7
8


import static java.util.stream.Collectors.toList;

...

long roomsCount = courses.stream()
        .map(Course::getClassRoom)    // get the room name from each course
        .collect(toList())            // build a list of these room names
        .stream().distinct().count(); // count the number of unique room names

Stringify a List

Create a comma-separated list of items in a String, from data in a List:

Java

1
2

String sorted = myList.stream().map(item -> item.getItemName())
            .collect(Collectors.joining(", "));

Check Items in a Collection

Three examples of the same basic thing:

Traditional for loop

The traditional way using a for() loop:

Java


1
2
3
4
5
6
7
8


boolean nameMatch = Boolean.FALSE;
for (ContentType ct : contentTypes) {
    if (ct.getContentTypeName().equals("movie")) {
        nameMatch = Boolean.TRUE;
        break;
    }
}
assertThat(nameMatch).isTrue();

Stream forEachOrdered

Using a filter .forEach():

Java


1
2
3
4
5
6


contentTypes
        .stream()
        .filter(ct -> ct.equals(expectedType))
        .forEachOrdered((ct) -> {
            assertThat(ct.getContentTypeName()).isEqualTo("movie");
        });

Stream ifPresentOrElse

Using a filter .ifPresentOrElse()

Java


 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12


contentTypes
        .stream()
        .filter(ct -> ct.equals(expectedType))
        .findFirst()
        .ifPresentOrElse(
            value -> {
                assertThat(value.getContentTypeName())
                    .isEqualTo("movie");
            },
            () -> {
                assertThat(false);
            });

At this point I am missing the simplicity of the traditional syntax.

Map of Lists to Set

Convert a Map<String, List<String>> to Set<String>.

Traditional forEach()

Java

1

map.values().forEach(l -> l.forEach(stringValues::add));

Stream flatMap()

Java


1
2
3
4


Set<String> stringValues = map.values()
   .stream()
   .flatMap(List::stream)
   .collect(Collectors.toCollection(HashSet::new));

Here, flatMap() flattens the nested list inside the map.

Process Map of Maps

Convert text in the following structure to upper case:

Java


 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16


Map<String,Map<String,String>> input = Map
        .of("abc",
            Map.of("Def", "Ghi"),
            "jkl",
            Map.of("MNO", "PQR", "stu", "vwx")
        );

Map<String,Map<String,String>> output = input.entrySet().stream()
        .collect(Collectors.toMap(
                e1 -> e1.getKey().toUpperCase(),
                e1 -> e1.getValue().entrySet().stream().collect(Collectors.toMap(
                        e2 -> e2.getKey().toUpperCase(),
                        e2 -> e2.getValue().toUpperCase()))));

System.out.println(input);
System.out.println(output);

Results:

{abc={Def=Ghi}, jkl={stu=vwx, MNO=PQR}}

{ABC={DEF=GHI}, JKL={STU=VWX, MNO=PQR}}

Note: The code will fail with IllegalStateException: Duplicate key if there are 2 keys that will become the same when uppercased.

Custom Sorting

Start with this:

Java


 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13


List<Map<String,String>> list = new ArrayList<>();
Map<String,String> map = new HashMap<>();
Map<String,String> map1 = new HashMap<>();
Map<String,String> map2 = new HashMap<>();
map.put("productNumber", "107-001");
map1.put("productNumber", "108-001");
map2.put("productNumber", "109-001");
map.put("price", "1.99");
map1.put("price", "1.02");
map2.put("price", "1.99");
list.add(map);
list.add(map1);
list.add(map2);

Objective: Sort by key1 in descending order, key2 ascending order.

price	productNumber
1.99	107-001
1.99	109-001
1.02	108-001

The following uses comparator chaining:

Java


1
2
3
4


Comparator<Map<String, String>> comp = Comparator
    .comparing(m ->Double.parseDouble(m.get("price"))
         ,Comparator.reverseOrder());
comp = comp.thenComparing(m -> m.get("productNumber"));

Apply it as follows:

Java


1
2
3
4
5
6


List<Map<String, String>> formattedResult =
                list.stream().sorted(comp)  
                        .collect(Collectors.toList())

        formattedResult.forEach(m -> System.out.println(
                m.get("price") + " : " + m.get("productNumber")));

Compute If Absent

Assume we have a set of client IDs. Each ID can have many invoices, each invoice generated at a specific datetime.

Java


 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25


private void doComputeIfAbsent() {
    int id = 0;
    Invoice invoice = new Invoice();
    LocalDateTime date = LocalDateTime.now();

    HashMap<Integer, HashMap<LocalDateTime, Invoice>> allInvoicesAllClients = new HashMap();
    HashMap<LocalDateTime, Invoice> allInvoices = allInvoicesAllClients.get(id);

    // sub-optimal:
    if (allInvoices != null) {
        allInvoices.put(date, invoice);      //<---REPEATED CODE                  
    } else {
        allInvoices = new HashMap<>();
        allInvoices.put(date, invoice);      //<---REPEATED CODE
        allInvoicesAllClients.put(id, allInvoices);
    }

    // If id isn't present as a key in allInvoicesAllClients, then
    // it'll create mapping from id to a new HashMap and return the
    // new HashMap. If id is present as a key, then it'll return the
    // existing HashMap. See:
    // https://docs.oracle.com/en/java/javase/13/docs/api/java.base/java/util/Map.html#computeIfAbsent(K,java.util.function.Function)
    allInvoicesAllClients.computeIfAbsent( id, k -> new HashMap() ).put(date, invoice);

}

Get Last n Entries

Sort a linked hash map in reverse order by value, in order to get the last (i.e. now the first) n entries:

Java

1

Map<String, Double> indDistance = new LinkedHashMap<String, Double>();

The comparator:

Java


1
2
3
4
5


public class ReverseDoubleComparator implements java.util.Comparator<Double> {
    public int compare(Double d1, Double d2) {
        return (d1.compareTo(d2)) * (-1);
    }
}

Now use this Comparator in sorted() method, i.e.:

Java


1
2
3
4


Map sortedInds =
        indDistance.entrySet().stream()
        .sorted(Entry.comparingByValue(new ReverseDoubleComparator()))
        .collect(Collectors.toMap(Entry::getKey, Entry::getValue, (e1, e2) -> e1, LinkedHashMap::new));

Frequency Count (1)

Java


1
2
3
4
5
6
7


List<String> wordsList = Arrays.asList("foo", "bar", "foo", "baz", "bar");
Map<String, Long> freqMap = wordsList.stream()
        .collect(Collectors.groupingBy(
                Function.identity(),
                Collectors.counting())
        );
System.out.println(freqMap);

Output: {bar=2, foo=2, baz=1}

Notes:

In the above groupingBy:

the first parameter is a function - in this case identity(), so that the input is returned as the output. In other words, each word in the list is what we group by.
the second parameter is the “classifier” - it maps the input elements to keys.

The classifier here is Collectors.counting(), which is a special case of Collectors.reducing() - specifically:

Java

1

Collectors.reducing(0L, e -> 1L, Long::sum)

This maps each element e to the number 1 and then sums it (with 0 as the default).

Frequency Count (2)

Java


1
2
3
4
5
6


List<String> wordsList = Arrays.asList("foo", "bar", "foo", "baz", "bar");
Map<String, Integer> freqMap2 = new HashMap<>();
wordsList.forEach(word
        -> freqMap2.compute(word, (k, v) -> v != null ? v + 1 : 1)
);
System.out.println(freqMap2);

This uses a more generic way of counting, using Map.compute(). In this example, the following…

Java

1

(k, v) -> v != null ? v + 1 : 1

…is the “remapping” function, used to compute each value in the end-result map.

The output is the same as for the previous frequency count example: {bar=2, foo=2, baz=1}

Group a List, Get the Min of Each Group

Test data:

Java


 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14


import java.time.Instant;

public class CarDetails {

    private final String model;
    private final Instant createdDate;

    public CarDetails(String model, Instant createdDate) {
        this.model = model;
        this.createdDate = createdDate;
    }

    // getters
}

Java


1
2
3
4
5
6
7
8
9


List<CarDetails> cars = new ArrayList<>();
CarDetails cd1 = new CarDetails("Ford", Instant.ofEpochSecond(1680742800));
CarDetails cd2 = new CarDetails("Ford", Instant.ofEpochSecond(1680746400));
CarDetails cd3 = new CarDetails("Honda", Instant.ofEpochSecond(1680746400));
CarDetails cd4 = new CarDetails("Honda", Instant.ofEpochSecond(1680742800));
cars.add(cd1);
cars.add(cd2);
cars.add(cd3);
cars.add(cd4);

The streams:

Java


 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13


Map<String, CarDetails> reducedCars = cars.stream()
        .collect(Collectors.groupingBy(CarDetails::getModel))
        .entrySet().stream().collect(
                Collectors.toMap(
                        (entry) -> entry.getKey(),
                        (entry) -> entry.getValue().stream()
                                .min(Comparator.comparing(CarDetails::getCreatedDate))
                                .orElseThrow(NoSuchElementException::new)
                ));

for (Map.Entry<String, CarDetails> entry : reducedCars.entrySet()) {
    System.out.println(entry.getKey() + " - " + entry.getValue().getCreatedDate());
}

In the above code, we use orElseThrow(...). If we did not include this step, we would need to use Optional:

Java

1

Map<String, Optional<CarDetails>> reducedCars