合成注解属性调整为同时根据名称和类型进行覆盖;

This commit is contained in:
huangchengxing 2022-06-14 17:25:57 +08:00
parent ff30bdf1ea
commit 0e2e894a3e
2 changed files with 40 additions and 16 deletions

View File

@ -2,6 +2,8 @@ package cn.hutool.core.annotation;
import cn.hutool.core.annotation.scanner.MateAnnotationScanner;
import cn.hutool.core.lang.Opt;
import cn.hutool.core.map.MapUtil;
import cn.hutool.core.util.ClassUtil;
import cn.hutool.core.util.ObjectUtil;
import cn.hutool.core.util.ReflectUtil;
import cn.hutool.core.util.StrUtil;
@ -21,7 +23,7 @@ import java.util.stream.Stream;
/**
* 表示一个根注解与根注解上的多层元注解合成的注解
*
* <p>假设现有注解AA上存在元注解BB上存在元注解C则对A解析得到的合成注解XX作为新的根注解则CBA都是X的元注解<br />
* <p>假设现有注解AA上存在元注解BB上存在元注解C则对A解析得到的合成注解X则CBA都是X的元注解X为根注解<br />
* 通过{@link #isAnnotationPresent(Class)}可确定指定类型是注解是否是该合成注解的元注解即是否为当前实例的父类
* 若指定注解是当前实例的元注解则通过{@link #getAnnotation(Class)}可获得动态代理生成的对应的注解实例<br />
* 需要注意的是由于认为合并注解X以最初的根注解A作为元注解因此{@link #getAnnotations()}{@link #getDeclaredAnnotations()}
@ -34,6 +36,8 @@ import java.util.stream.Stream;
* 若两相同注解处于同一层级则按照从其上一级子注解{@link AnnotatedElement#getAnnotations()}的调用顺序排序<br />
* {@link #getAnnotation(Class)}获得的代理类实例的属性值遵循该规则
*
* <p>同名属性将根据类型彼此隔离即当不同层级的元注解存在同名的属性但是属性类型不同时此时低层级的属性并不会覆盖高层级注解的属性
*
* <p>别名在合成注解中仍然有效若注解X中任意属性上存在{@link Alias}注解{@link Alias#value()}指定的属性值将会覆盖注解属性的本身的值<br />
* {@link Alias}注解仅能指定注解X中存在的属性作为别名不允许指定元注解或子类注解的属性
*
@ -55,7 +59,7 @@ public class SyntheticAnnotation<A extends Annotation> implements Annotation, An
/**
* 属性值缓存
*/
private final Map<String, Object> attributeCaches;
private final Map<String, Map<Class<?>, Object>> attributeCaches;
/**
* 构造
@ -108,15 +112,16 @@ public class SyntheticAnnotation<A extends Annotation> implements Annotation, An
}
/**
* 获取属性值若存在{@link Alias}则获取{@link Alias#value()}指定的别名属性的值
* <p>当不同层级的注解之间存在同名属性时将优先获取更接近根注解的属性
* 根据指定的属性名与属性类型获取对应的属性值若存在{@link Alias}则获取{@link Alias#value()}指定的别名属性的值
* <p>当不同层级的注解之间存在同名同类型属性时将优先获取更接近根注解的属性
*
* @param attributeName 属性名
*/
public Object getAttribute(String attributeName) {
return attributeCaches.computeIfAbsent(attributeName, a -> metaAnnotationMap.values()
public Object getAttribute(String attributeName, Class<?> attributeType) {
Map<Class<?>, Object> values = attributeCaches.computeIfAbsent(attributeName, t -> MapUtil.newHashMap());
return values.computeIfAbsent(attributeType, a -> metaAnnotationMap.values()
.stream()
.filter(ma -> ma.hasAttribute(attributeName)) // 集合默认是根据distance有序的故此处无需再排序
.filter(ma -> ma.hasAttribute(attributeName, attributeType)) // 集合默认是根据distance有序的故此处无需再排序
.findFirst()
.map(ma -> ma.getAttribute(attributeName))
.orElse(null)
@ -154,7 +159,7 @@ public class SyntheticAnnotation<A extends Annotation> implements Annotation, An
}
/**
* 获取根注解直接声明注解
* 获取根注解直接声明的注解该方法正常情况下当只返回原注解
*
* @return 直接声明注解
*/
@ -238,6 +243,19 @@ public class SyntheticAnnotation<A extends Annotation> implements Annotation, An
return attributeMethodCaches.containsKey(attributeName);
}
/**
* 元注解是否存在该属性且该属性的值类型是指定类型或其子类
*
* @param attributeName 属性名
* @param returnType 返回值类型
* @return 是否存在该属性
*/
public boolean hasAttribute(String attributeName, Class<?> returnType) {
return Opt.ofNullable(attributeMethodCaches.get(attributeName))
.filter(method -> ClassUtil.isAssignable(returnType, method.getReturnType()))
.isPresent();
}
/**
* 获取元注解的属性值
*
@ -304,7 +322,7 @@ public class SyntheticAnnotation<A extends Annotation> implements Annotation, An
return getToString();
}
return ObjectUtil.defaultIfNull(
syntheticAnnotation.getAttribute(method.getName()),
syntheticAnnotation.getAttribute(method.getName(), method.getReturnType()),
() -> ReflectUtil.invoke(this, method, args)
);
}
@ -317,7 +335,7 @@ public class SyntheticAnnotation<A extends Annotation> implements Annotation, An
private String getToString() {
String attributes = Stream.of(annotationType().getDeclaredMethods())
.filter(AnnotationUtil::isAttributeMethod)
.map(method -> StrUtil.format("{}={}", method.getName(), syntheticAnnotation.getAttribute(method.getName())))
.map(method -> StrUtil.format("{}={}", method.getName(), syntheticAnnotation.getAttribute(method.getName(), method.getReturnType())))
.collect(Collectors.joining(", "));
return StrUtil.format("@{}({})", annotationType().getName(), attributes);
}

View File

@ -23,10 +23,10 @@ public class SyntheticAnnotationTest {
Assert.assertEquals(syntheticAnnotation.getDeclaredAnnotations()[0], rootAnnotation);
Assert.assertEquals(3, syntheticAnnotation.getAnnotations().length);
Assert.assertEquals(syntheticAnnotation.getAttribute("childValue"), "Child!");
Assert.assertEquals(syntheticAnnotation.getAttribute("childValueAlias"), "Child!");
Assert.assertEquals(syntheticAnnotation.getAttribute("parentValue"), "Child's Parent!");
Assert.assertEquals(syntheticAnnotation.getAttribute("grandParentValue"), "Child's GrandParent!");
Assert.assertEquals(syntheticAnnotation.getAttribute("childValue", String.class), "Child!");
Assert.assertEquals(syntheticAnnotation.getAttribute("childValueAlias", String.class), "Child!");
Assert.assertEquals(syntheticAnnotation.getAttribute("parentValue", String.class), "Child's Parent!");
Assert.assertEquals(syntheticAnnotation.getAttribute("grandParentValue", String.class), "Child's GrandParent!");
Map<Class<? extends Annotation>, SyntheticAnnotation.MetaAnnotation> annotationMap = syntheticAnnotation.getMetaAnnotationMap();
ChildAnnotation childAnnotation = syntheticAnnotation.getAnnotation(ChildAnnotation.class);
@ -34,31 +34,35 @@ public class SyntheticAnnotationTest {
Assert.assertNotNull(childAnnotation);
Assert.assertEquals(childAnnotation.childValue(), "Child!");
Assert.assertEquals(childAnnotation.childValueAlias(), "Child!");
Assert.assertEquals(childAnnotation.grandParentType(), Integer.class);
Assert.assertEquals(annotationMap, SyntheticAnnotation.of(childAnnotation).getMetaAnnotationMap());
ParentAnnotation parentAnnotation = syntheticAnnotation.getAnnotation(ParentAnnotation.class);
Assert.assertTrue(syntheticAnnotation.isAnnotationPresent(ParentAnnotation.class));
Assert.assertNotNull(parentAnnotation);
Assert.assertEquals(parentAnnotation.parentValue(), "Child's Parent!");
Assert.assertEquals(parentAnnotation.grandParentType(), "java.lang.Void");
Assert.assertEquals(annotationMap, SyntheticAnnotation.of(parentAnnotation).getMetaAnnotationMap());
GrandParentAnnotation grandParentAnnotation = syntheticAnnotation.getAnnotation(GrandParentAnnotation.class);
Assert.assertTrue(syntheticAnnotation.isAnnotationPresent(GrandParentAnnotation.class));
Assert.assertNotNull(grandParentAnnotation);
Assert.assertEquals(grandParentAnnotation.grandParentValue(), "Child's GrandParent!");
Assert.assertEquals(grandParentAnnotation.grandParentType(), Integer.class);
Assert.assertEquals(annotationMap, SyntheticAnnotation.of(grandParentAnnotation).getMetaAnnotationMap());
}
// 注解结构如下
// AnnotatedClass -> @ChildAnnotation -> @ParentAnnotation -> @GrandParentAnnotation
// -> @GrandParentAnnotation
@ChildAnnotation(childValueAlias = "Child!")
class AnnotatedClass {}
@ChildAnnotation(childValueAlias = "Child!", grandParentType = Integer.class)
static class AnnotatedClass {}
@Retention(RetentionPolicy.RUNTIME)
@Target({ ElementType.ANNOTATION_TYPE })
@interface GrandParentAnnotation {
String grandParentValue() default "";
Class<?> grandParentType() default Void.class;
}
@GrandParentAnnotation(grandParentValue = "Parent's GrandParent!") // 覆盖元注解@GrandParentAnnotation的属性
@ -66,6 +70,7 @@ public class SyntheticAnnotationTest {
@Target({ ElementType.TYPE })
@interface ParentAnnotation {
String parentValue() default "";
String grandParentType() default "java.lang.Void";
}
@GrandParentAnnotation(grandParentValue = "Child's GrandParent!") // 重复的元注解靠近根注解的优先级高
@ -76,6 +81,7 @@ public class SyntheticAnnotationTest {
String childValueAlias() default "";
@Alias("childValueAlias")
String childValue() default "";
Class<?> grandParentType() default Void.class;
}
}