Java SE 8 for Java EE developers

@Delabassee @JosePaumard#SE84EE
Java EE devs
Java SE 8 for

@Delabassee @JosePaumard#Devoxx #SE84EE
Agenda
Java SE 8 has been released ~1.5 year ago
Java EE 7 Application Servers
- GlassFish, WildFly, Liberty Profile, WebLogic, etc.

Agenda
Java SE 8 has been released ~1.5 year ago
Java EE 7 Application Servers
- GlassFish, WildFly, Liberty Profile, WebLogic, etc.
Check some important Java SE 8 new features
for Java EE developers... with patterns!

The following is intended to outline our general product direction.
It is intended for information purposes only, and may not be
incorporated into any contract. It is not a commitment to deliver
any material, code, or functionality, and should not be relied upon
in making purchasing decisions. The development, release, and
timing of any features or functionality described for Oracle’s
products remains at the sole discretion of Oracle.

Questions?
#SE84EE
http://slido.com

New Date & Time API
Date & Time API is a new introduction in Java 8
Initiated by Stephen Colebourne, based on Joda Time
Complete replacement of Date & Calendar

Date: ZonedTime
Useful for localized times
Set<String> allZonesIds = ZoneId.getAvailableZoneIds() ;
String ukTZ = ZoneId.of("Europe/Paris") ;

Date: ZonedTime
System.out.println(
ZonedDateTime.of(
1998, Month.JULY.getValue(), 12, // year / month / day
22, 0, 0, 0, // h / mn / s / nanos
ZoneId.of("Europe/Paris"))
); // prints 1998-07-22T22:00-00:01:15[Europe/London]

Date: ZonedTime
System.out.println(
ZonedDateTime.of(
22, 0, 0, 0, // h / mn / s / nanos
);
ZonedDateTime nextWorldCup =
currentWorldCup.plus(Period.ofYear(4));

Date: ZonedTime
System.out.println(
ZonedDateTime.of(
22, 0, 0, 0, // h / mn / s / nanos
);
ZonedDateTime nextWorldCup =
currentWorldCup.plus(Period.ofYear(4));
ZonedDateTime nexWorldCupJapan=
nextWorldCup.withZoneSameInstant(ZoneId.of("Asia/Tokyo")) ;

Date: bridges with java.util.Date
Conversions between j.u.Date and Date & Time
Time time = Time.from(localTime); // API -> legacy
LocalTime localTime = time.toLocalTime(); // legacy -> new API
Date date = Date.from(localDate); // API -> legacy
LocalDate localDate = date.toLocalDate(); // legacy -> new API
TimeStamp time = TimeStamp.from(instant); // API -> legacy
Instant instant = time.toInstant(); // legacy -> new API
Date date = Date.from(instant); // API -> legacy
Instant instant = date.toInstant(); // legacy -> new API

Usable in JPA
JPA does not support this Date & Time API (yet)
But with converters, we can still use it
@Entity
public abstract class AbstractPersistent {
@Convert(converter=DateConverter.class)
private Instant instant;
}

Usable in JPA
The converter is a classical object
public class DateConverter
implements AttributeConverter<Instant, Date> {
public Date convertToDatabaseColumn(Instant instant) {
return Date.from(instant);
}
public Instant convertToEntityAttribute(Date date) {
return date.toInstant();
}
}

Usable in JPA
The converter is a classical object
public class DateFormatConverter
implements AttributeConverter<ZonedDateTime, String> {
public String convertToDatabaseColumn(ZonedDateTime time) {
return DateTimeFormatter.ISO_DATE_TIME.format(time);
}
public ZonedDateTime convertToEntityAttribute(String formated) {
return DateTimeFormatter.ISO_DATE_TIME
.parse(formated, ZonedDateTime::from);
}
}

Usable in JPA
With JPA converters we can use the types
from Date & Time API and map in j.u.l.Date or String

Usable in JSF
With a Custom JSF Converter
@FacesConverter("instantConverter")
public class TimeConverter implements javax.faces.convert.Converter {
@Override
public Object getAsObject(FacesContext ctx, … , String value) {
return Instant.parse(value);
}
}

Usable in JSF
With a Custom JSF Converter...
@Override
public String getAsString(FacesContext ctx, … , Object value) {
DateTimeFormatter formatter =
DateTimeFormatter.ofLocalizedDateTime(FormatStyle.SHORT)
.withLocale(Locale.US)
.withZone(ZoneId.systemDefault());
return formatter.format((TemporalAccessor) value);
}
}

Usable in JSF
With a Custom JSF Converter<h:form>
<h:inputText id = "date”
value = "#{order.timestamp}"
size = "20" required="true”
label = "start"
converter = "instantConverter" />
...
@ManagedBean(name="order") @SessionScoped
public class myBean implements Serializable{
Instant timestamp;
…

Annotations in Java 7
Wrapping annotations
An annotation cannot be applied more than once
@NamedQueries({
@NamedQuery(name=SELECT_ALL, query="..."),
@NamedQuery(name=COUNT_ALL, query="...")
})
public class Customer {
}

Annotations in Java 8
Java 8 makes it possible!
@NamedQuery(name=SELECT_ALL, query="..."),
@NamedQuery(name=COUNT_ALL, query="...")
public class Customer {
}

Bean Validation
Suppose we want to validate a parameter
public orderdCar order( @CheckCar("VW") Car aCar ) {
...
}

Bean Validation
Here is the Bean Validation code for @CheckCar
@Target({PARAMETER})
@Retention(RUNTIME)
@Constraint(validatedBy = CarValidator.class)
public @interface CheckCar {
Class<?>[] groups() default {};
Class<? extends Payload>[] payload() default {};
String value();
}

Bean Validation
And for the validator
public class CarValidator
implements ConstraintValidator<CheckCar, Car> {
private String carBrand;
public void initialize(CheckCar constraintAnnotation) {
this.carBrand = constraintAnnotation.value();
}
public boolean isValid(Car obj, ConstraintValidatorContext ctrCtx) {
if (object == null) return true;
return (!obj.getBrand().toUpperCase().contains(carBrand));
}
}

Bean Validation
And the Java 8 trick to make it repeatable
@Retention(RUNTIME)
@Repeatable(CheckCars.class)
@Constraint(validatedBy = CarValidator.class)
public @interface CheckCar {
Class<?>[] groups() default {};
Class<? extends Payload>[] payload() default {};
String value();
}

Bean Validation
And the Java 8 trick to make it repeatable
@Retention(RUNTIME)
public @interface CheckCars {
CheckCar[] value() default{};
}

Bean Validation
Suppose we want to validate a parameter
public orderdCar order( @CheckCar("VW") @ChechCar("Audi") Car aCar ) {
...
}

Type annotations
Annotations can be now put on types
Example 1: tell that a variable should not be null
private @NonNull List<Person> persons = ... ;

Type annotations
Annotations can be now put on types
Example 1: tell that a variable should not be null
Example 2: the content should not be null neither
private @NonNull List<Person> persons = ... ;
private @NonNull List<@NonNull Person> persons = ... ;

Joining strings
New String joiners are coming to the JDK!

Joining strings
StringJoiner class
And it prints:
StringJoiner sj = new StringJoiner(", ") ;
sj.add("one").add("two").add("three") ;
String s = sj.toString() ;
System.out.println(s) ;
> one, two, three

Joining strings
StringJoiner with prefix / postfix
And it prints:
StringJoiner sj = new StringJoiner(", ", "{", "}") ;
sj.add("one").add("two").add("three") ;
String s = sj.toString() ;
System.out.println(s) ;
> {one, two, three}

Joining strings
Also accessible from the String class
And it prints:
// From the String class, with a vararg
String s = String.join(", ", "one", "two", "three");
System.out.println(s);
> one, two, three

Joining strings
And directly from a List
// From a list to a String using the Stream & Collectors API
String s =
listOfStrings.stream()
.collect(Collectors.joining(", "));
> one, two, three

Joining strings
And directly from a List
The result is
// From the String class, with a vararg
String s =
listOfStrings.stream()
.collect(Collectors.joining(", ", "{", "}"));
> {one, two, three}

Support for Base64 decoding
After a few years…
import java.util.Base64;
String txt = "Modern Times!";
String encoded = Base64.getEncoder().encodeToString(
txt.getBytes(StandardCharsets.UTF_8));
String decoded = new String(Base64.getDecoder().decode(encoded),
StandardCharsets.UTF_8);

What is a Stream?
A new API
A typed interface
A new concept

What is a Stream?
A new API
A typed interface
A new concept
Let’s see some code!

Readable and efficient patterns
Extract histograms from data
List<Person> people = Arrays.asList();
Map<City, Double> agesByCity =
people.stream()
.filter(p -> p.getAge() > 20)
.collect(
Collectors.groupingBy(
Person::getCity,
Collectors.averagingDouble(Person::getAge)
)
);

Building a Stream on a String
Building a Stream on the letters of a String:
String s = "hello";
IntStream stream = s.chars(); // stream on the letters of s

Building a Stream on a String
Building a Stream on the letters of a String:
String s = "hello";
IntStream stream = s.chars(); // stream on the letters of s
s.chars() // IntStream
.mapToObj(letter -> (char)letter) // Stream<Character>
.map(Character::toUpperCase)
.forEach(System.out::println); // Prints a Character
> HELLO

Build a Stream from a text file
Find the first error line from a log file
// Java 7 : try with resources and use of Paths
Path path = Paths.get("d:", "tmp", "debug.log");
try (Stream<String> stream = Files.lines(path)) {
stream.filter(line -> line.contains("ERROR"))
.findFirst()
.ifPresent(System.out::println);
} catch (IOException ioe) {
// handle the exception
}

Build a Stream from a regex
Building a Stream from a regexp
// book is a looooooooong String
Stream<String> words =
Pattern
.compile(" ")
.splitAsStream(book) ;

Build a Stream from a regex
Building a Stream from a regexp
More efficient than:
Pattern
.compile(" ")
.splitAsStream(book) ;
Stream.of(book.split(" "));

Flatmap: Files.lines + regex
Splitting a text files into words
Stream<String> streamOfLines = Files.lines(path);
Function<String, Stream<String>> splitter =
line -> Pattern.compile(" ").splitAsStream(line);
long numberOfWords =
streamOfLines.flatMap(splitter)
.map(String::toLowerCase)
.distinct()
.count();

Why is it important for Java EE?
Given this JSON file
[
{
"name":"Duke",
"gender":"m",
"phones":[
"home":"650‐123‐4567",
"mobile":"650‐111‐2222"
]
},
{
"name":"Jane", …
]
JsonArray contacts =
Json.createArrayBuilder()
.add(…

And some JSON-P magic
Map<String, Long> names =
contacts.getValuesAs(JsonObject.class).stream()
.filter(x -> "f".equals(x.getString("gender")))
.map(x -> (x.getString("name")))
.collect(
Collectors.groupingBy(
Function.identity(),
Collectors.counting()
)
);

// Today
JsonArray names =
.collect(
Collector.of(
() -> Json.createArrayBuilder(),
(builder, value) -> builder.add(value),
(builder1, builder2) -> builder1.add(builder2),
builder -> builder.build()
)
);

// Today
JsonArray names =
.collect(
Collector.of(
Json::createArrayBuilder,
JsonArrayBuilder::add,
JsonArrayBuilder::build
)
);

// Today
JsonArray names =
.collect(
Collectors.collectingAndThen(
Collector.of(
Json::createArrayBuilder,
JsonArrayBuilder::add
),
JsonArrayBuilder::build
)
);

Collect into JSON
Avaiblable through the JsonCollectors factory
// Tomorrow, with JSON-P 1.1
JsonArray names =
.collect(
JsonCollectors.toJsonArray()
);

A warning on parallelism
All parallel operations (Streams, ConcurrentHashMap)
take place in the common ForkJoinPool
The common ForkJoinPool takes all the
available cores / CPUs

A warning on parallelism
All parallel operations (Streams, ConcurrentHashMap)
take place in the common ForkJoinPool
The common ForkJoinPool takes all the
available cores / CPUs
To preserve your Application Server:
System.setProperty(
"java.util.concurrent.ForkJoinPool.common.parallelism", "1") ;

CompletableFuture
New addition to j.u.concurrent
Allow to chain asynchronous tasks
Use case: test the asynchronous calls in the Servlet API,
EJB, JAX-RS, …

Creation of an asynchronous task
The Jersey way to create an asynchronous call
@Path("/resource")
public class AsyncResource {
@GET
public void asyncGet(@Suspended final AsyncResponse asyncResponse) {
new Thread(new Runnable() {
public void run() {
String result = longOperation();
asyncResponse.resume(result);
}
}).start();
}
}

Creation of an asynchronous task
(let us fix this code with Java 8)
@Path("/resource")
@GET
executor.submit(() -> {
});
}
}

How to test it?
We want to check if the result object
is passed to the resume() method of the asyncResponse
It is a very basic test, but tricky to write since we are in an
asynchronous world
We have mocks for that!

How to test it?
We can inject a mock AsyncResponse, even mock the result
Then verify the correct interaction:
But we need to verify this once the run() method has been
called…
Mockito.verify(mockAsyncResponse).resume(result);

How to test it?
This is where CompletionStage come to the rescue!
@Path("/resource")
@Inject ExecutorService executor;
@GET
executor.submit(() -> {
});
}
}

How to test it?
@Path("/resource")
@Inject ExecutorService executor;
@GET
executeAsync(asyncResponse);
}
public CompletableFuture<Void> executeAsync(final AsyncResponse asyncResponse) {
return CompletableFuture.runAsync(() -> {
asyncResponse.resume(longOperation());
}, executor);
}
}

How to test it?
AsyncResource asyncResource = new AsyncResource();
asyncResource.executeAsync(mockAsyncResponse); // returns the CompletableFuture
.thenRun(() -> { // then execute this Runnable
Mockito.verify(mockAsyncResponse).resume(result);
}
);

How to test it?
Be careful of visibility issues
1) It’s better to run this in the same thread
2) Since the mocks are used and checked in this thread,
create them in this thread too

How to test it?
So the complete pattern becomes this one
1) First we create our mocks
String result = Mockito.mock(String.class);
AsyncResponse response = Mockito.mock(AsyncResponse.class);
Runnable train = () -> Mockito.doReturn(result).when(response.longOperation());
Runnable verify = () -> Mockito.verify(response).resume(result);

How to test it?
2) Then we create the call & verify
Runnable callAndVerify = () -> {
asyncResource.executeAsync(response).thenRun(verify);
}

How to test it?
3) Then we create the task
ExecutorService executor = Executors.newSingleThreadExecutor();
AsynResource asyncResource = new AsyncResource();
asyncResource.setExecutorService(executor);
CompletableFuture
.runAsync(train, executor) // this trains our mocks
.thenRun(callAndVerify); // this verifies our mocks

How to test it?
Since a CompletableFuture is also a Future, we can fail
with a timeout if the test does not complete fast enough
ExecutorService executor = Executors.newSingleThreadExecutor();
AsynResource asyncResource = new AsyncResource();
asyncResource.setExecutorService(executor);
CompletableFuture
.runAsync(train, executor) // this trains our mocks
.thenRun(callAndVerify) // this verifies our mocks
.get(10, TimeUnit.SECONDS);

Conclusion
Java 8 is not just about lambdas
There are also many new and very useful patterns
Ready to be used before becoming a lambda ninja!
Not covered:
- Collection framework
- Concurrency, Concurrent hashmap
- JavaScript on the JVM with Nashorn
- Security

Thank you!

Java SE 8 for Java EE developers

Recommandé

Recommandé

Contenu connexe

Tendances

Tendances (20)

En vedette

En vedette (20)

Similaire à Java SE 8 for Java EE developers

Similaire à Java SE 8 for Java EE developers (20)

Plus de José Paumard

Plus de José Paumard (15)

Dernier

Dernier (20)

Java SE 8 for Java EE developers