Spring MVC 源码学习札记（四）HandlerMapping和HandlerAdapter(3)

上一节主要阅读了BeanNameUrlHandlerMapping类，并理清了它的父类关系，值得注意的是它的爷爷的爸爸也就是AbstractHandlerMapping实现了HandlerMapping接口，而继承了WebApplicationObjectSupport类，在AbstractDetectingUrlHandlerMapping类中，有这样一个方法：

/**
	 * Calls the {@link #detectHandlers()} method in addition to the
	 * superclass's initialization.
	 */
	@Override
	public void initApplicationContext() throws ApplicationContextException {
		super.initApplicationContext();
		detectHandlers();
	}

 上一节没有说明这个，经过我翻WebApplicationObjectSupport类和它的父类ApplicationObjectSupport发现有一个setApplicationContext(ApplicationContext context)方法，在其中调用了initApplicationContext方法，前面的setter肯定是在bean构造时用到的，更上层的以后再去阅读理解吧，到这里就追溯到了handler探测的本源。

 DefaultAnnoationHandlerMapping类也是继承自AbstractDetectingUrlHandlerMapping类，则必然实现了它的抽象方法：

/**
	 * Determine the URLs for the given handler bean.
	 * @param beanName the name of the candidate bean
	 * @return the URLs determined for the bean,
	 * or <code>null</code> or an empty array if none
	 */
	protected abstract String[] determineUrlsForHandler(String beanName);

  那么detectHandler接下来的执行顺序就和上一节的顺序相同，这时我们跳到了DefaultAnnoationHandlerMapping类中的该方法：

/**
	 * Checks for presence of the {@link org.springframework.web.bind.annotation.RequestMapping}
	 * annotation on the handler class and on any of its methods.
	 */
	@Override
	protected String[] determineUrlsForHandler(String beanName) {
		ApplicationContext context = getApplicationContext();
		Class<?> handlerType = context.getType(beanName);
		RequestMapping mapping = context.findAnnotationOnBean(beanName, RequestMapping.class);
		if (mapping != null) {
			// @RequestMapping found at type level
			this.cachedMappings.put(handlerType, mapping);
			Set<String> urls = new LinkedHashSet<String>();
			String[] typeLevelPatterns = mapping.value();
			if (typeLevelPatterns.length > 0) {
				// @RequestMapping specifies paths at type level
				String[] methodLevelPatterns = determineUrlsForHandlerMethods(handlerType);
				for (String typeLevelPattern : typeLevelPatterns) {
					if (!typeLevelPattern.startsWith("/")) {
						typeLevelPattern = "/" + typeLevelPattern;
					}
					for (String methodLevelPattern : methodLevelPatterns) {
						String combinedPattern = getPathMatcher().combine(typeLevelPattern, methodLevelPattern);
						addUrlsForPath(urls, combinedPattern);
					}
					addUrlsForPath(urls, typeLevelPattern);
				}
				return StringUtils.toStringArray(urls);
			}
			else {
				// actual paths specified by @RequestMapping at method level
				return determineUrlsForHandlerMethods(handlerType);
			}
		}
		else if (AnnotationUtils.findAnnotation(handlerType, Controller.class) != null) {
			// @RequestMapping to be introspected at method level
			return determineUrlsForHandlerMethods(handlerType);
		}
		else {
			return null;
		}
	}

 既然是默认注解支持的HandlerMapping，那么它必然是去探测@RequestMapping所注解的Handler和Method。这里就有必要来看看RequestMapping接口的定义：

@Target({ElementType.METHOD, ElementType.TYPE})
@Retention(RetentionPolicy.RUNTIME)
@Documented
@Mapping
public @interface RequestMapping {

	String[] value() default {};

	RequestMethod[] method() default {};

	String[] params() default {};

	String[] headers() default {}; 

}

 value数组显然是存储所有该Handler中的url，method则是POST GET等方法，RequestMethod是一个enum类型

public enum RequestMethod {

	GET, HEAD, POST, PUT, DELETE, OPTIONS, TRACE

}

 context.findAnnotationOnBean(beanName, RequestMapping.class)显然是查找了该Handler上所有使用了@RequestMapping的注解，将它们的value,method取出，再封装到RequestMapping类对象中。接下来的加入缓存语句就不用解释了。

我们来看看determineUrlsForHandlerMethods方法的源码：

/**
	 * Derive URL mappings from the handler's method-level mappings.
	 * @param handlerType the handler type to introspect
	 * @return the array of mapped URLs
	 */
	protected String[] determineUrlsForHandlerMethods(Class<?> handlerType) {
		final Set<String> urls = new LinkedHashSet<String>();
		Class<?>[] handlerTypes =
				Proxy.isProxyClass(handlerType) ? handlerType.getInterfaces() : new Class<?>[]{handlerType};
		for (Class<?> currentHandlerType : handlerTypes) {
			ReflectionUtils.doWithMethods(currentHandlerType, new ReflectionUtils.MethodCallback() {
				public void doWith(Method method) {
					RequestMapping mapping = AnnotationUtils.findAnnotation(method, RequestMapping.class);
					if (mapping != null) {
						String[] mappedPaths = mapping.value();
						for (String mappedPath : mappedPaths) {
							addUrlsForPath(urls, mappedPath);
						}
					}
				}
			});
		}
		return StringUtils.toStringArray(urls);
	}

 这里就比较麻烦了，先从字面意思理解就是查询类中的方法，找到方法对应的RequestMapping的所包含的value，组成一个url数组。

接下来看它的实现，首先是检查了handlerType是否是一个代理类

/**
     * Returns true if and only if the specified class was dynamically
     * generated to be a proxy class.
*/
public static boolean isProxyClass(Class<?> cl) {
	if (cl == null) {
	    throw new NullPointerException();
	}
       
	return proxyClasses.containsKey(cl);
    }

/** set of all generated proxy classes, for isProxyClass implementation */
    private static Map proxyClasses =
	Collections.synchronizedMap(new WeakHashMap());

 接下来是反射和回调，我也看到云里雾里的。。。。理解不来，以后再想吧，直到它最后是得到的url就行了。

handlerType.getInterfaces();

用于得到所有handlerType实现的接口的Class对象

new Class<?>[]{handlerType}

得到的还只是handlerType的Class对象

所以这里就有些不懂为什么要这么做，希望有高手看到我的文章 帮我解答下疑惑