Spring: @WebMvcTest + @ComponentScan

If you need to use @ComponentScan in your spring boot application, be aware that it will break your @WebMvcTest….unless you add defaults filters from the @SpringBootApplication annotation declaration

    basePackages = {"x.y.z"},
    excludeFilters = {
        @Filter(type = FilterType.CUSTOM, classes = TypeExcludeFilter.class),
        @Filter(type = FilterType.CUSTOM, classes = AutoConfigurationExcludeFilter.class)}
public class Application {

  public static void main(String[] args) {
    SpringApplication.run(Application.class, args);

Angular: amount input field


As a user I would like to have a possibility to enter numbers with my localized separators.

Example of such number regionalization can be found here

var number = 123456.789;

console.log(new Intl.NumberFormat('de-DE', { style: 'currency', currency: 'EUR' }).format(number));
// expected output: "123.456,79 €"

// the Japanese yen doesn't use a minor unit
console.log(new Intl.NumberFormat('ja-JP', { style: 'currency', currency: 'JPY' }).format(number));
// expected output: "¥123,457"

// limit to three significant digits
console.log(new Intl.NumberFormat('en-IN', { maximumSignificantDigits: 3 }).format(number));
// expected output: "1,23,000"<span id="mce_SELREST_start" style="overflow:hidden;line-height:0;">&#65279;</span>


It would be great to have such ‘regionalization‘ applied while typing the numbers in the input field.

Sadly, it is not that easy to do this.

For example, if we would like to use what browser offers – the NumberFormat object that is above (not all browser have it see page section ‘Browser Compatibility’), we need a function that could upon the user localization do the reverse transformation to simple number notation in JavaScript while typing.


The method for displaying a number in right user localization is easy and we saw the example above – so this part is done right away. But doing the opposite conversion is missing!! We don’t have it!!!

What makes matters worse, if we would like to cover such transformation we would have to prepare an implementation for dozens of regions!!! That are compatible with what NubmerFormat implementation offers!

We would also need to allow only valid characters in the input field, where these ‘valid characters’ differ amoung the regions i.e.

Example 1: German number notation  is – 123.456,789

Example 2: India number notation  is – 1,23,456.789

As you can see the COMMA and DOT have diffrent meenings in diffrent countries.

This means we need a specific MASKING for diffrent region….

In JavaScript such masking can be done with: https://www.npmjs.com/package/angular2-text-mask

So far, we saw that there are more and more obstacles…

How can we tackle the problem?

Below is a solution, that in my opinion should do the job.


I will provide example in angular soon.


Java: RX buffer while subscribe

    final class BufferWhileSubscriber extends Subscriber {

        final Subscriber<? super List> actual;

        List buffer = new ArrayList();

        BufferWhileSubscriber(Subscriber<? super List> actual) {
            this.actual = actual;

        public void onNext(T t) {
            if (buffer.size() > 1 && boundaryPredicate.call(buffer)) {
                actual.onNext(buffer.subList(0, buffer.size() - 1));

                T lastElement = buffer.get(buffer.size() - 1);
                buffer = Lists.newLinkedList();

        public void onError(Throwable e) {
            buffer = null;

        public void onCompleted() {
            List b = buffer;
            buffer = null;
            if (!b.isEmpty()) {

Java: awaiting termination


        final ExecutorService executor = Executors.newCachedThreadPool();
        try {
            Future task = executor.submit(() -> {
            executor.awaitTermination(timeoutInMinutes, TimeUnit.MINUTES);

            SomeResult result = task.get();

            return Optional.of(result);
        } catch (InterruptedException | ExecutionException e) {
            if (!executor.isShutdown()) {
        } finally {
            log.debug("Shutting down the service executor");
            if (!executor.isShutdown()) {

Java: streams with zipping

With the appearence of Java 8 there are more and more APIs that return Stream object instead of simple Collections.

For example, with JPA you can just say that as a return type you want to have a Stream<Something>…. and you will have it 😉

Repository returning Stream of some objects… how this can be helpful? You may ask…

Well, in some cases it might be helpful, actually. Especially, when there is a constraint on the memory, and we need to process things on the “fly”.

Below is a sample code that uses 2 repositories that return streams, combines these 2 streams into 1 stream of pairs… does some computation and sends the results in the batches.


import com.google.common.collect.Iterators;
import com.google.common.collect.Maps;
import io.vavr.Function1;
import io.vavr.Function2;
import io.vavr.Tuple;
import io.vavr.Tuple2;

import org.slf4j.MDC;
import java.util.List;
import java.util.Map;
import java.util.Optional;
import java.util.stream.Collectors;
import java.util.stream.Stream;

public class SomeService {

    public void doIt(final String currentProcessId) {
        MDC.put(MDC_PROCESS_CORRELATION_ID, currentProcessId);

        log.debug("Fetch current domain objects");
        Stream previousStream = null;
        Stream currentStream =

        log.debug("Combine the current and previous values into currentPreviousPictures");
        Stream currentPreviousPictures;
        if (someCondition1) {
            log.debug("Fetch previous domain objects");
            previousStream = repository.findByDomainObjectBySomePropertyOrderByIdAsc(someProperty, previous);

            currentPreviousPictures = zip(currentStream, previousStream);

            if (someCondition2) {
        } else {
            currentPreviousPictures = zip(currentStream, Stream.empty());

        currentPreviousPictures = currentPreviousPictures

        log.debug("Process previous and current domain objects in batch");
        Iterators.partition(currentPreviousPictures.iterator(), processBatchSize)
            .forEachRemaining(currentPreviousList -> {


                log.info("Approximate progress: {}%", 100 * allProcessedDomainObjects.get() / maxDomainObjects);



Java: zipping sorted collections

Let’s say we need to combine two sorted lists of numbers into a list of pairs, where a pair contains numbers that are equal.

If there is a number missing in one of the two lists, than we create a pair with the number and a ZERO element.


list_1 = [1, 3, 4, 6, 10]

list_2 = [1, 2, 3, 4, 5, 7, 10]

zippedList = [ (1, 1), (0, 2), (3, 3), (4, 4), (0, 5), (6, 0), (0, 7), (10, 10) ]


We can achive it with following implementation:


import io.vavr.Tuple;
import io.vavr.Tuple2;

import java.util.Iterator;
import java.util.stream.Stream;
import java.util.stream.StreamSupport;

public class ZippedIterator implements Iterator<Tuple2> {

    private final Iterator iter1;
    private final Iterator iter2;
    private final T zeroElement;

    private T lastElement1 = null;
    private T lastElement2 = null;

    public static  Iterator<Tuple2>
    zipAllIterators(Iterator iter1, Iterator iter2, T zeroElement) {
        if (!iter1.hasNext() && !iter2.hasNext()) {
            return Stream.<Tuple2>empty().iterator();
        } else {
            return new ZippedIterator(iter1, iter2, zeroElement);

    public static  Stream<Tuple2>
    zipAll(Stream stream1, Stream stream2, T zeroElement, boolean parallel) {
        Iterator<Tuple2> tuple2Iterator =
            zipAllIterators(stream1.iterator(), stream2.iterator(), zeroElement);
        Iterable<Tuple2> iterable = () -> tuple2Iterator;
        return StreamSupport.stream(iterable.spliterator(), parallel);

    public ZippedIterator(Iterator iter1, Iterator iter2, T zeroElement) {
        this.iter1 = iter1;
        this.iter2 = iter2;
        this.zeroElement = zeroElement;

    public boolean hasNext() {
        return hasNext1() || hasNext2();

    public Tuple2 next() {
        T e1 = next1();
        T e2 = next2();
        if (e1 != null && e2 != null) {
            int r = e1.compareTo(e2);
            if (r == 0) {
                lastElement1 = null;
                lastElement2 = null;
                return Tuple.of(e1, e2);
            } else if (r  0
                lastElement2 = null;
                return Tuple.of(zeroElement, e2);
        if (e1 == null) {
            lastElement2 = null;
            return Tuple.of(zeroElement, e2);
        lastElement1 = null;
        return Tuple.of(e1, zeroElement);

    private boolean hasNext1() {
        return lastElement1 != null || iter1.hasNext();

    private boolean hasNext2() {
        return lastElement2 != null || iter2.hasNext();

    private T next1() {
        return lastElement1 != null ? lastElement1 : (!iter1.hasNext() ? null : (lastElement1 = iter1.next()));

    private T next2() {
        return lastElement2 != null ? lastElement2 : (!iter2.hasNext() ? null : (lastElement2 = iter2.next()));



We can check this implementation with a few tests:


import static io.vavr.Tuple.of;
import static org.hamcrest.Matchers.contains;
import static org.junit.Assert.assertThat;

import io.vavr.Tuple2;
import org.junit.Test;

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

public class ZippedIteratorTest {

    private Integer zero = 0;

    public void testZipper() throws Exception {
        Stream s1 = Stream.of(4, 5, 10, 11);
        Stream s2 = Stream.of(1, 2, 4, 6, 8, 10, 15);

        Stream<Tuple2> zipped = zipper(s1, s2, zero);

        assertThat(zipped.collect(Collectors.toList()), contains(
            of(zero, 1),
            of(zero, 2),
            of(4, 4),
            of(5, zero),
            of(zero, 6),
            of(zero, 8),
            of(10, 10),
            of(11, zero),
            of(zero, 15)

    public void testZipperRigthEmpty() throws Exception {
        Stream s1 = Stream.of(4, 5, 10, 11);
        Stream s2 = Stream.empty();

        Stream<Tuple2> zipped = zipper(s1, s2, zero);

        assertThat(zipped.collect(Collectors.toList()), contains(
            of(4, zero),
            of(5, zero),
            of(10, zero),
            of(11, zero)

    public void testZipperLeftEmpty() throws Exception {
        Stream s1 = Stream.empty();
        Stream s2 = Stream.of(4, 5, 10, 11);

        Stream<Tuple2> zipped = zipper(s1, s2, zero);

        assertThat(zipped.collect(Collectors.toList()), contains(
            of(zero, 4),
            of(zero, 5),
            of(zero, 10),
            of(zero, 11)

    private Stream<Tuple2> zipper(Stream s1, Stream s2,
                                                    Integer nullObject) {
        return zipAll(s1, s2, nullObject, false);

Java: JAXB generation

Generating classes from XML schema:

    <!-- The package of your generated sources -->

Additionally, we need a folder with our schema in the project:


Now, we can generate the classes from the schema types…


These classes will land in the request package declared in the plugin.