JUnit4-Description.java源代码 解读与分析
作用域
- children 子描述集合,套件测试时,children不为空,套件测试可以由套件测试和原子测试组合而成。
- displayName 正常情况下,displayName格式都是这样的:方法名(类名),方法
formatDisplayName
就是将其方法名,类名格式化成这种形式。 - uniqueId 唯一标识符,从两个私有构造方法可看出,不对其设置的话,默认为displayName
- annotation[] 注解
- testClass 测试类
创建
由于构造方法私有化,Description的创建只能够通过静态方法来创建。
public static Description createSuiteDescription(String name, Annotation... annotations) {
return new Description(null, name, annotations);
}
public static Description createSuiteDescription(String name, Serializable uniqueId, Annotation... annotations) {
return new Description(null, name, uniqueId, annotations);
}
public static Description createTestDescription(String className, String name, Annotation... annotations) {
return new Description(null, formatDisplayName(name, className), annotations);
}
public static Description createTestDescription(Class<?> clazz, String name, Annotation... annotations) {
return new Description(clazz, formatDisplayName(name, clazz.getName()), annotations);
}
public static Description createTestDescription(Class<?> clazz, String name) {
return new Description(clazz, formatDisplayName(name, clazz.getName()));
}
public static Description createTestDescription(String className, String name, Serializable uniqueId) {
return new Description(null, formatDisplayName(name, className), uniqueId);
}
public static Description createSuiteDescription(Class<?> testClass) {
return new Description(testClass, testClass.getName(), testClass.getAnnotations());
}
方法解读
源码中,有些方法有点不好理解,这次说明一下:
1.从当前注解数组中,获取某种类型的注解.
这个方法在JUnit中涉及到注解时,就常常出现。如作用域上的注解,方法上的注解啊,测试类上的注解啊…泛型T, 继承Annotation,即T是注解类型。Class.cast(Annotation)能够将Annotation转化为T类型
public <T extends Annotation> T getAnnotation(Class<T> annotationType) {
for (Annotation each : fAnnotations) {
if (each.annotationType().equals(annotationType)) {
return annotationType.cast(each); //相当于(T)each
}
}
return null;
}
2.获取测试类,测试类名,方法名时,这些信息都可为空,而实际上可能不为空,创建时,没有传进这些参数;但有一个信息不为空,那就是displayName, 它的格式是 方法名(类全名)
因此,在一些信息为空时,采取正则表达式匹配的方法,从displayName获取信息。
正则表达式中, 组是用括号划分的正则表达式,([\\s\\S]*)\\((.*)\\)
中,第0组是整个表达式,第1组是[\\s\\S]*
表示方法名, 第二组是.*
,表示类全名。在下面方法中,如果匹配的话(matcher.matches()
), 那么就会根据组数group来获取对应的信息。
private static final Pattern METHOD_AND_CLASS_NAME_PATTERN = Pattern
.compile("([\\s\\S]*)\\((.*)\\)");
private String methodAndClassNamePatternGroupOrDefault(int group,
String defaultString) {
Matcher matcher = METHOD_AND_CLASS_NAME_PATTERN.matcher(toString());
return matcher.matches() ? matcher.group(group) : defaultString;
}
//获取方法名 group = 1
public String getMethodName() {
return methodAndClassNamePatternGroupOrDefault(1, null);
}
//获取类名, testClass不为空时,直接getName()获取,否则group=2
public String getClassName() {
return fTestClass != null ? fTestClass.getName() : methodAndClassNamePatternGroupOrDefault(2, toString());
}
//获取测试类时,testClass为空时,尝试根据className, 使用Class.forName()获取
public Class<?> getTestClass() {
if (fTestClass != null) {
return fTestClass;
}
String name = getClassName();
if (name == null) {
return null;
}
try {
fTestClass = Class.forName(name, false, getClass().getClassLoader());
return fTestClass;
} catch (ClassNotFoundException e) {
return null;
}
}
JUnit4中的使用
首先是抽象类Runner实现了Describable接口,使我们的运行类可被描述
public abstract class Runner implements Describable {
public abstract Description getDescription();
public abstract void run(RunNotifier notifier);
public int testCount() {
return getDescription().testCount();
}
然后是ParentRunner(怎么称呼这个类,是真的不好表达,但它又很重要(=@__@=))。为其子类Runner,提供了许多方法。
而子类需要对自己子节点进行描述。(这里不详细讲,后面会有补充)
/**
* Provides most of the functionality specific to a Runner that implements a
* "parent node" in the test tree, with children defined by objects of some data
* type {@code T}. (For {@link BlockJUnit4ClassRunner}, {@code T} is
* {@link Method} . For {@link Suite}, {@code T} is {@link Class}.) Subclasses
* must implement finding the children of the node, describing each child, and
* running each child. ParentRunner will filter and sort children, handle
* {@code @BeforeClass} and {@code @AfterClass} methods,
* handle annotated {@link ClassRule}s, create a composite
* {@link Description}, and run children sequentially.
*
* @since 4.5
*/
public abstract class ParentRunner<T> extends Runner{
@Override
public Description getDescription() {
Description description = Description.createSuiteDescription(getName(),
getRunnerAnnotations());
for (T child : getFilteredChildren()) {
description.addChild(describeChild(child));
}
return description;
}
protected abstract Description describeChild(T child);
}
如下是它的两个节点类型不同的子类:
以方法为测试节点 BlockJUnit4ClassRunner
public class BlockJUnit4ClassRunner extends ParentRunner<FrameworkMethod> {
@Override
protected Description describeChild(FrameworkMethod method) {
Description description = methodDescriptions.get(method);
if (description == null) {
description = Description.createTestDescription(getTestClass().getJavaClass(),
testName(method), method.getAnnotations());
methodDescriptions.putIfAbsent(method, description);
}
return description;
}
}
以Runner为节点的 Suite套件测试
public class Suite extends ParentRunner<Runner> {
@Override
protected Description describeChild(Runner child) {
return child.getDescription();
}
}
完整代码
package org.junit.runner;
import java.io.Serializable;
import java.lang.annotation.Annotation;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.Collection;
import java.util.concurrent.ConcurrentLinkedQueue;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
/**
* A <code>Description</code> describes a test which is to be run or has been run. <code>Descriptions</code>
* can be atomic (a single test) or compound (containing children tests). <code>Descriptions</code> are used
* to provide feedback about the tests that are about to run (for example, the tree view
* visible in many IDEs) or tests that have been run (for example, the failures view).
* <p>
* <code>Descriptions</code> are implemented as a single class rather than a Composite because
* they are entirely informational. They contain no logic aside from counting their tests.
* <p>
* In the past, we used the raw {@link junit.framework.TestCase}s and {@link junit.framework.TestSuite}s
* to display the tree of tests. This was no longer viable in JUnit 4 because atomic tests no longer have
* a superclass below {@link Object}. We needed a way to pass a class and name together. Description
* emerged from this.
*
* @see Request
* @see Runner
* @since 4.0
*/
public class Description implements Serializable {
private static final long serialVersionUID = 1L;
private static final Pattern METHOD_AND_CLASS_NAME_PATTERN = Pattern
.compile("([\\s\\S]*)\\((.*)\\)");
/**
* Create a <code>Description</code> named <code>name</code>.
* Generally, you will add children to this <code>Description</code>.
*
* @param name the name of the <code>Description</code>
* @param annotations meta-data about the test, for downstream interpreters
* @return a <code>Description</code> named <code>name</code>
*/
public static Description createSuiteDescription(String name, Annotation... annotations) {
return new Description(null, name, annotations);
}
/**
* Create a <code>Description</code> named <code>name</code>.
* Generally, you will add children to this <code>Description</code>.
*
* @param name the name of the <code>Description</code>
* @param uniqueId an arbitrary object used to define uniqueness (in {@link #equals(Object)}
* @param annotations meta-data about the test, for downstream interpreters
* @return a <code>Description</code> named <code>name</code>
*/
public static Description createSuiteDescription(String name, Serializable uniqueId, Annotation... annotations) {
return new Description(null, name, uniqueId, annotations);
}
/**
* Create a <code>Description</code> of a single test named <code>name</code> in the 'class' named
* <code>className</code>. Generally, this will be a leaf <code>Description</code>. This method is a better choice
* than {@link #createTestDescription(Class, String, Annotation...)} for test runners whose test cases are not
* defined in an actual Java <code>Class</code>.
*
* @param className the class name of the test
* @param name the name of the test (a method name for test annotated with {@link org.junit.Test})
* @param annotations meta-data about the test, for downstream interpreters
* @return a <code>Description</code> named <code>name</code>
*/
public static Description createTestDescription(String className, String name, Annotation... annotations) {
return new Description(null, formatDisplayName(name, className), annotations);
}
/**
* Create a <code>Description</code> of a single test named <code>name</code> in the class <code>clazz</code>.
* Generally, this will be a leaf <code>Description</code>.
*
* @param clazz the class of the test
* @param name the name of the test (a method name for test annotated with {@link org.junit.Test})
* @param annotations meta-data about the test, for downstream interpreters
* @return a <code>Description</code> named <code>name</code>
*/
public static Description createTestDescription(Class<?> clazz, String name, Annotation... annotations) {
return new Description(clazz, formatDisplayName(name, clazz.getName()), annotations);
}
/**
* Create a <code>Description</code> of a single test named <code>name</code> in the class <code>clazz</code>.
* Generally, this will be a leaf <code>Description</code>.
* (This remains for binary compatibility with clients of JUnit 4.3)
*
* @param clazz the class of the test
* @param name the name of the test (a method name for test annotated with {@link org.junit.Test})
* @return a <code>Description</code> named <code>name</code>
*/
public static Description createTestDescription(Class<?> clazz, String name) {
return new Description(clazz, formatDisplayName(name, clazz.getName()));
}
/**
* Create a <code>Description</code> of a single test named <code>name</code> in the class <code>clazz</code>.
* Generally, this will be a leaf <code>Description</code>.
*
* @param name the name of the test (a method name for test annotated with {@link org.junit.Test})
* @return a <code>Description</code> named <code>name</code>
*/
public static Description createTestDescription(String className, String name, Serializable uniqueId) {
return new Description(null, formatDisplayName(name, className), uniqueId);
}
private static String formatDisplayName(String name, String className) {
return String.format("%s(%s)", name, className);
}
/**
* Create a <code>Description</code> named after <code>testClass</code>
*
* @param testClass A {@link Class} containing tests
* @return a <code>Description</code> of <code>testClass</code>
*/
public static Description createSuiteDescription(Class<?> testClass) {
return new Description(testClass, testClass.getName(), testClass.getAnnotations());
}
/**
* Describes a Runner which runs no tests
*/
public static final Description EMPTY = new Description(null, "No Tests");
/**
* Describes a step in the test-running mechanism that goes so wrong no
* other description can be used (for example, an exception thrown from a Runner's
* constructor
*/
public static final Description TEST_MECHANISM = new Description(null, "Test mechanism");
/*
* We have to use the f prefix until the next major release to ensure
* serialization compatibility.
* See https://github.com/junit-team/junit/issues/976
*/
private final Collection<Description> fChildren = new ConcurrentLinkedQueue<Description>();
private final String fDisplayName;
private final Serializable fUniqueId;
private final Annotation[] fAnnotations;
private volatile /* write-once */ Class<?> fTestClass;
private Description(Class<?> clazz, String displayName, Annotation... annotations) {
this(clazz, displayName, displayName, annotations);
}
private Description(Class<?> testClass, String displayName, Serializable uniqueId, Annotation... annotations) {
if ((displayName == null) || (displayName.length() == 0)) {
throw new IllegalArgumentException(
"The display name must not be empty.");
}
if ((uniqueId == null)) {
throw new IllegalArgumentException(
"The unique id must not be null.");
}
this.fTestClass = testClass;
this.fDisplayName = displayName;
this.fUniqueId = uniqueId;
this.fAnnotations = annotations;
}
/**
* @return a user-understandable label
*/
public String getDisplayName() {
return fDisplayName;
}
/**
* Add <code>Description</code> as a child of the receiver.
*
* @param description the soon-to-be child.
*/
public void addChild(Description description) {
fChildren.add(description);
}
/**
* Gets the copy of the children of this {@code Description}.
* Returns an empty list if there are no children.
*/
public ArrayList<Description> getChildren() {
return new ArrayList<Description>(fChildren);
}
/**
* @return <code>true</code> if the receiver is a suite
*/
public boolean isSuite() {
return !isTest();
}
/**
* @return <code>true</code> if the receiver is an atomic test
*/
public boolean isTest() {
return fChildren.isEmpty();
}
/**
* @return the total number of atomic tests in the receiver
*/
public int testCount() {
if (isTest()) {
return 1;
}
int result = 0;
for (Description child : fChildren) {
result += child.testCount();
}
return result;
}
@Override
public int hashCode() {
return fUniqueId.hashCode();
}
@Override
public boolean equals(Object obj) {
if (!(obj instanceof Description)) {
return false;
}
Description d = (Description) obj;
return fUniqueId.equals(d.fUniqueId);
}
@Override
public String toString() {
return getDisplayName();
}
/**
* @return true if this is a description of a Runner that runs no tests
*/
public boolean isEmpty() {
return equals(EMPTY);
}
/**
* @return a copy of this description, with no children (on the assumption that some of the
* children will be added back)
*/
public Description childlessCopy() {
return new Description(fTestClass, fDisplayName, fAnnotations);
}
/**
* @return the annotation of type annotationType that is attached to this description node,
* or null if none exists
*/
public <T extends Annotation> T getAnnotation(Class<T> annotationType) {
for (Annotation each : fAnnotations) {
if (each.annotationType().equals(annotationType)) {
return annotationType.cast(each);
}
}
return null;
}
/**
* @return all of the annotations attached to this description node
*/
public Collection<Annotation> getAnnotations() {
return Arrays.asList(fAnnotations);
}
/**
* @return If this describes a method invocation,
* the class of the test instance.
*/
public Class<?> getTestClass() {
if (fTestClass != null) {
return fTestClass;
}
String name = getClassName();
if (name == null) {
return null;
}
try {
fTestClass = Class.forName(name, false, getClass().getClassLoader());
return fTestClass;
} catch (ClassNotFoundException e) {
return null;
}
}
/**
* @return If this describes a method invocation,
* the name of the class of the test instance
*/
public String getClassName() {
return fTestClass != null ? fTestClass.getName() : methodAndClassNamePatternGroupOrDefault(2, toString());
}
/**
* @return If this describes a method invocation,
* the name of the method (or null if not)
*/
public String getMethodName() {
return methodAndClassNamePatternGroupOrDefault(1, null);
}
private String methodAndClassNamePatternGroupOrDefault(int group,
String defaultString) {
Matcher matcher = METHOD_AND_CLASS_NAME_PATTERN.matcher(toString());
return matcher.matches() ? matcher.group(group) : defaultString;
}
}