Java lambda aggregator

Question

I have a list

PLANS:

ID,    A_CODE, COMPARTMENT
10683   163     213 
10683   616     194
10683   163     212
10683   163     211
10683   163     214

and want to produce AGGREGATED_PLANS

ID      A_CODE  COMPARTMENTS
10683   163     211/212/213/214
10683   616     194

How can I do this using java lambda expression ?

I am thinking something like this, but not sure about the aggregateCompartments part ?

plans.stream()
     .collect(groupingBy(Plan::getACode, 
           aggregateCompartments(Plan::getCompartments)));

Any suggestions ?

Please update your question with a minimal, complete, and verifiable example. — giorgiga
– giorgiga, Commented Feb 15, 2018 at 5:47
Does using groupingBy(Plan::getACode, mapping(Plan::getCompartments, joining("/")), help? — Alanpatchi
– Alanpatchi, Commented Feb 15, 2018 at 5:53

Farvardin · Accepted Answer · 2018-02-15 06:21:21Z

1

this may help you:

    List<Plan> plans = new ArrayList<>();
    plans.add(new Plan("10683", "163", "213"));
    plans.add(new Plan("10683", "616", "194"));
    plans.add(new Plan("10683", "163", "212"));
    plans.add(new Plan("10683", "163", "211"));
    plans.add(new Plan("10683", "163", "214"));

    System.out.println(
        plans.stream().collect(
               Collectors.groupingBy(p -> p.id + " " + p.aCode, 
               Collectors.mapping(Plan::getCompartment, Collectors.joining("/")))));

output:

{10683 616=194, 10683 163=213/212/211/214}

answered Feb 15, 2018 at 6:21

Farvardin

5,4226 gold badges36 silver badges54 bronze badges

Sign up to request clarification or add additional context in comments.

Comments

Kirill Simonov · Accepted Answer · 2018-02-15 07:16:00Z

If I get you right, your desired output is a collection of plans, where compartment field of each plan is aggregated from all grouped compartments.

In this case I suggest you to use a toMap collector, which has a special mergeFunction argument:

Collection<Plan> aggregatedPlans = plans
    .stream()
    .collect(toMap(
            Plan::getACode,
            p -> new Plan(p.getId(), p.getACode(), p.getCompartment()),
            (p1, p2) -> {
                p1.compartment += "/" + p2.compartment;
                return p1;
            }))
    .values();

If you add the following two methods to your Plan object

public Plan clone() {
    return new Plan(id, aCode, compartment);
}

public Plan merge(Plan other) {
    compartment += "/" + other.compartment;
    return this;
}

then the code can be simplified like this:

Collection<Plan> aggregatedPlans = plans.stream()
            .collect(toMap(Plan::getACode, Plan::clone, Plan::merge)).values();

fps · Accepted Answer · 2018-02-16 00:36:17Z

0

If you want to grupo by A_CODE:

Map<ACodeType, Set<CompartmentType>> result = plans.stream()
    .collect(Collectors.groupingBy(
        Plan::getACode, 
        Collectors.mapping(
            Plan::getCompartments, 
            Collectors.toCollection(TreeSet::new))));

This groups by A_CODE, with a downstream collector that first adapts each plan to its compartment and then collects each adapted group to a TreeSet, to maintain the compartments sorted and without duplicates. This needs the type of the compartment to be Comparable. If it's not, you'd need to provide a Comparator to the TreeSet's constructor.

answered Feb 16, 2018 at 0:36

fps

34.5k8 gold badges62 silver badges116 bronze badges

Collectives™ on Stack Overflow

Java lambda aggregator

3 Answers 3

Comments

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

Comments

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related