Skip to content

Latest commit

 

History

History
322 lines (232 loc) · 8.21 KB

README.md

File metadata and controls

322 lines (232 loc) · 8.21 KB

ESA Commons

Build codecov Maven Central GitHub license

ESA Commons is the common lib of ESA Stack.

Features

  • SPI Enhancement: Allows loading SPI by name, group, tag and so on..
  • Logger: Detect the slf4j automatically.
  • InternalLogger: Write log without any log library.
  • Lambda Enhancement: lambda interfaces receiving multiple parameters, lambda interfaces allows to throw an exception.
  • MultiValueMap: Implementation of Map<K, List<V>.
  • UnsafeUtils, UnsafeArrayUtils: Unity classes for unsafe options.
  • Reflection support: AnnotationUtils, ClassUtils...
  • more features...

Maven Dependency

<dependency>
    <groupId>io.esastack</groupId>
    <artifactId>commons</artifactId>
    <version>${mvn.version}</version>
</dependency>

Reference Guide

Logger & LoggerFactory

There are many excellent logging frameworks such as log4j, log4j2, logback and so on... Although the slf4j framework serves as a simple facade or abstraction for various logging frameworks, maybe you still do not want to use slf4j as the facade in your own framework which will be used by the end user. So we provides a couple of APIs( Logger, LoggerFactory ) to serve as a facade of slf4j and JDK logger(may be we will support log4j and logback in the future).

eg.

just declare the logger as usual

private static final Logger logger = LoggerFactory.getLogger("foo");

also use it as usual

logger.info("hell world!");
logger.debug("hello {}!", "world");
logger.warn("{} {}!", "hello", "world");
logger.error("oops!", new Exception("there's something wrong with it"));

Please do not forget that we are using the Logger and LoggerFactory that are provided in esa-commons instead of the org.slf4j.Logger.

InternalLogger

InternalLogger allows you to write logs without any thirdparty log dependency such as log4j, logback.

In some thirdparty log frameworks, you may need a xxx.xml to configure your logs to the target file which depends on your end user.

eg.

final InternalLogger logger = InternalLoggers.logger("foo", new File("foo.log")).build();
logger.info("hello world!");

and you will see the appended log in the file named foo.log

2020-12-10 11:42:10.141 INFO [main] foo : hello world!

log pattern

you can configure the log patterns below

  • %d, %date: log date
  • %l, %level:log level
  • %t, %thread: caller thread
  • %logger: logger name
  • %m, %msg, %message: log content
  • %n: new line
  • %ex, %exception, %thrown: error detail if present

eg.

the default log pattern: %date %level [%thread] %logger : %msg%n%thrown

and the output of logger.info("hello world!")

2020-12-10 11:42:10.141 INFO [main] foo : hello world!

eg.

log pattern: %msg%n

and the output of

logger.info("hello");
logger.info("world");

is

hello
world

File Rolling

  • SIZE_BASED rolling
  • TIME_BASED rolling
  • TIME_AND_SIZE_BASED rolling

eg.

SIZE_BASED rolling with max size of 10M, and the max history log file's number of 3

final long maxSize = 10 * 1024 * 1024;
final int maxHistory = 3;
final InternalLogger logger = InternalLoggers.logger("foo", new File("foo.log"))
         .useSizeBasedRolling(maxSize, maxHistory)
         .build();

TIME_BASED rolling everyday

final InternalLogger logger = InternalLoggers.logger("foo", new File("foo.log"))
         .useTimeBasedRolling("yyyy-MM-dd")
         .build();

TIME_BASED rolling every hour

final InternalLogger logger = InternalLoggers.logger("foo", new File("foo.log"))
         .useTimeBasedRolling("yyyy-MM-dd_HH")
         .build();

patter of date only support: Hour and Day level

TIME_AND_SIZE_BASED rolling

final long maxSize = 10 * 1024 * 1024;
final int maxHistory = 3;
final InternalLogger logger = InternalLoggers.logger("foo", new File("foo.log"))
         .useTimeAndSizeBasedRolling("yyyy-MM-dd", maxSize, maxHistory)
         .build();

Lambda Enhancement

Arguments Enhancement

JDK only supports a part of lambda interfaces, and here we declare more interfaces like

  • Consumer3, Consumer4, Consumer5: Consumer with 3(4 or5) arguments
  • Function3, Function4, Function5: Function with 3(4 or5) arguments
  • Predicate3, Predicate4, Predicate5: Predicate with 3(4 or5) arguments
  • and so on...

eg.

Consumer3<Integer, String, Foo> c = (num, str, foo) -> {};

Throwing Enhancement

JDK lambda interface is not allowed to throw an exception, and here we declare the interfaces like

  • ThrowingConsumer, ThrowingBiConsumer, ThrowingConsumer3, ThrowingConsumer4

  • ThrowingFunction, ThrowingBiFunction, ThrowingFunction3, ThrowingFunction4

  • ThrowingPredicate, ThrowingBiPredicate, ThrowingPredicate3, ThrowingPredicate4

  • ThrowingSupplier

  • ThrowingRunnable

  • and so on...

eg.

ThrowingConsumer<String> c = str -> {
    mayThrowIOException();
    // ...
}

static void mayThrowIOException() {
    if (condition) {
        throw new IOException();
    }
}

this will transfer your ThrowingConsumer to Consumer which will rethrow the exception(if present).

Consumer<String> c = ThrowingConsumer.rethrow(str -> {
    mayThrowIOException();
    // ...
});

this will just suppress the exception

Consumer<String> c = ThrowingConsumer.suppress(str -> {
    mayThrowIOException();
    // ...
});

this will handle the exception

Function<String, String> c = ThrowingFunction.failover(str -> {
    mayThrowIOException();
    return "foo";
}, (v, t) -> {
    // handle error 't'
    return "bar";
});

Auto-(Un)Box Enhancement

Allows you to use lambda interfaces without Auto-Box or Auto-Unbox.

  • ObjIntFunction, ObjDoubleFunction, ObjLongFunction
  • ObjIntPredicate, ObjDoublePredicate, ObjLongPredicate
  • ThrowingIntConsumer, ThrowingDoubleConsumer, ThrowingLongConsumer
  • ThrowingIntFunction, ThrowingDoubleFunction, ThrowingLongFunction
  • ThrowingIntPredicate, ThrowingDoublePredicate, ThrowingLongPredicate
  • ThrowingIntSupplier, ThrowingDoubleSupplier, ThrowingLongSupplier
  • and so on...

just find what you want.

SPI

usage

use @SPI to annotate it is a SPI interface

@SPI
public interface Shape {
    // ...
}

use @Feature to add properties for implementation

@Feature(groups = "foo", order = 1, tags = "a:1", excludeTags = "b:1")
public class Circle implements Shape {
    // ...
}
@Feature(groups = "bar", order = -1, tags = "a:2", excludeTags = "b:2")
public class Triangle implements Shape {
    // ...
}

add spec file to META-INF/services/, or META-INF/esa/, or META-INF/esa/internal/(any directory is ok)

so we add a spec file

META-INF/services/io.esastack.commons.Shape

io.esastack.commons.Circle
io.esastack.commons.Triangle

use SpiLoader to get SPI extensions

instances should be in sort by @Feature#order

final List<Shape> shapes = SpiLoader.cached(Shape.class).getAll();
assertEquals(2, shapes.size());
assertTrue(shapes.get(0) instanceof Triangle);
assertTrue(shapes.get(1) instanceof Circle);

get by groups

final List<Shape> shapes = SpiLoader.cached(Shape.class).getByGroup("foo");
assertEquals(1, shapes.size());
assertTrue(shapes.get(0) instanceof Circle);

get by tags

final List<Shape> shapes = SpiLoader.cached(Shape.class).getByTags(Collections.singletonMap("a", "2"));
assertEquals(1, shapes.size());
assertTrue(shapes.get(0) instanceof Triangle);

get by features

final List<Shape> shapes = SpiLoader.cached(Shape.class)
        .getByFeature("foo", Collections.singletonMap("a", "1"));
assertEquals(1, shapes.size());
assertTrue(shapes.get(0) instanceof Circle);