加入收藏 | 设为首页 | 会员中心 | 我要投稿 安卓应用网 (https://www.0791zz.com/)- 科技、建站、经验、云计算、5G、大数据,站长网!
当前位置: 首页 > 编程开发 > Java > 正文

聊一聊Java反射

发布时间:2020-05-27 11:30:58 所属栏目:Java 来源:互联网
导读:这次提到的Java反射涉及的代码比较多。因为工作中经常用到反射,对代码做了很多抽象以及过滤器。虽然代码量很多,但是简单易用,过滤插件也易修改。

这次提到的Java反射涉及的代码比较多。因为工作中经常用到反射,对代码做了很多抽象以及过滤器。虽然代码量很多,但是简单易用,过滤插件也易修改。

下面介绍下工作中哪些地方比较容易用到反射。比如插件或者过滤器,如果抽象的子类比较少,配置成XML等结构也是可以达到同样的效果。如果希望灵活一些,添加了插件或者过滤器代码子类后希望可以直接使用。可能反射会比较好点,通过扫描所有class或者jar文件,得到所有继承的子类。如果每次调用都扫描所有的文件会比较影响性能。所以在实现里面加入反射缓存,对所要获取反射子类时涉及的所有参数作为一个key缓存所有的反射结果。下次如果是同样的key,就不在重新扫描。

代码示例如下:

public static void main(String[] args) {
 //设置扫描范围,可以是class文件所在位置例如bin下或者是mysql开头或者mysql结尾的jar,//设置为""为全部都扫描,这种比较耗时
 ReflectUtils.createSharedReflections("classes","bin","mysql");
 try {
  //调试阶段可以设置每次都全扫描
  //Beans.setDesignTime(true);
  final Collection<String> subTypes = ReflectUtils.listSubClass(IA.class);//
  for (final String subType : subTypes) {
  //这里获取的是所有继承IA的子类
  System.out.println(subType);
  final IA impl = ReflectUtils.initClass(subType,IA.class);
  if (null == impl)
   continue;
  //通过该方式,可以统一做操作,
  impl.print();
  }
 } catch (Exception e) {
  e.printStackTrace();
 }
 }

代码执行结果:

//缓存文件,避免每次调用反射都重新扫描
//如果删除该文件,再次调用反射时,会重新扫描,一般会在代码里面有添加子类的时候会删除该文件
XmlUtils.readXml failure:.configuration.REF (系统找不到指定的文件。)
net.simple.reflect.test.B
net.simple.reflect.test.B
net.simple.reflect.test.D
net.simple.reflect.test.V

具体的类里面如何实现的大家就看下源码吧,这里贴出两个核心类的代码。源码地址:https://git.oschina.net/eliyanfei/api_tools

package net.simple.reflect;

import java.io.File;
import java.io.IOException;
import java.net.JarURLConnection;
import java.net.URL;
import java.net.URLDecoder;
import java.util.ArrayList;
import java.util.Collection;
import java.util.Enumeration;
import java.util.LinkedHashMap;
import java.util.List;
import java.util.Map;
import java.util.concurrent.TimeUnit;
import java.util.jar.JarEntry;
import java.util.jar.JarFile;
import java.util.zip.ZipEntry;

import net.simple.reflect.filter.IPathURLFilter;
import net.simple.reflect.filter.ISubTypeFilter;
import net.simple.reflect.filter.ITypeFilter;

import org.w3c.dom.Document;
import org.w3c.dom.Element;


/**
 * 
 * @author 李岩飞
 * @email eliyanfei@126.com 
 * 2016年11月2日 下午3:23:49
 *
 */
public final class Reflections {
 private final Collection<URL> pathUrls;
 private final Collection<IPathURLFilter> pathURLfilters;
 private final Collection<ITypeFilter> typeFilters;
 private ISubTypeFilter subTypeFilter;

 public Reflections() {
 typeFilters = new ArrayList<ITypeFilter>();
 pathURLfilters = new ArrayList<IPathURLFilter>();
 this.pathUrls = ClasspathHelper.getUrlsForCurrentClasspath();
 }

 public Reflections(final Collection<URL> pathUrls) {
 this.pathUrls = pathUrls;
 typeFilters = new ArrayList<ITypeFilter>();
 pathURLfilters = new ArrayList<IPathURLFilter>();
 }

 /**
 * @param subTypeFilter
 *      the subTypeFilter to set
 */
 public void setSubTypeFilter(final ISubTypeFilter subTypeFilter) {
 this.subTypeFilter = subTypeFilter;
 }

 /**
 * @return the subTypeFilter
 */
 public ISubTypeFilter getSubTypeFilter() {
 return subTypeFilter;
 }

 public Reflections addPathURLFilter(final IPathURLFilter pathURLFilter) {
 if (null == pathURLFilter)
  return this;
 if (!this.pathURLfilters.contains(pathURLFilter))
  this.pathURLfilters.add(pathURLFilter);
 return this;
 }

 public Reflections addTypeFilter(final ITypeFilter typeFilter) {
 if (null == typeFilter)
  return this;
 if (!this.typeFilters.contains(typeFilter))
  this.typeFilters.add(typeFilter);
 return this;
 }

 private static final String histFile = "./configuration.REF";
 private Document histDom;

 public Collection<String> getSubTypesFast(final Class<?> baseType) {//,final String... typeNames
 //首先过滤出当前允许扫描的路径
 final StringBuilder bufPathsId = new StringBuilder(32);
 final Map<File,URL> fileUrls = new LinkedHashMap<File,URL>(8);
 for (final URL pathUrl : pathUrls) {
  if (!acceptPathUrl(pathUrl))
  continue;
  File file = null;
  try {
  file = new File(URLDecoder.decode(pathUrl.getFile(),"UTF-8"));
  } catch (final Exception e) {
  file = new File(pathUrl.getFile());
  }
  fileUrls.put(file,pathUrl);
  if (!file.exists())//is url file?ignore
  continue;
  bufPathsId.append(file.getName()).append(file.lastModified());
 }
 final String domId = MD5.getHashString(bufPathsId.toString());
 if (null == histDom)
  histDom = W3cUtils.readXml(histFile);
 if (null == histDom)
  histDom = W3cUtils.newDom("R");
 Element rootEle = histDom.getDocumentElement();
 if (null == rootEle)
  histDom.appendChild(rootEle = histDom.createElement("R"));
 if (!domId.equals(rootEle.getAttribute("id"))) {
  rootEle.getParentNode().removeChild(rootEle);
  histDom.appendChild(rootEle = histDom.createElement("R"));
  rootEle.setAttribute("id",domId);
 }
 final String baseTypeId = MD5.getHashString(baseType.getName());
 Element refEle = W3cUtils.firstChildElement(rootEle,"E","id",baseTypeId);
 if (null != refEle) {
  final List<Element> valueEles = W3cUtils.childElementList(refEle,"F");
  final Collection<String> result = new ArrayList<String>(valueEles.size());
  for (final Element valueEle : valueEles) {
  result.add(new String(Base64.decodeFast(valueEle.getAttribute("id"))));
  }
  return result;
 }
 final ThreadPool<ListSubTypes> pool = new ThreadPool<ListSubTypes>();
 for (final File fileKey : fileUrls.keySet()) {
  pool.execute(new ListSubTypes(baseType,fileKey,fileUrls.get(fileKey)));
 }
 try {
  pool.shutdown(3,TimeUnit.MINUTES);
 } catch (final InterruptedException e) {
  e.printStackTrace();//for debug
 }
 final Collection<String> result = new ArrayList<String>();
 for (final ListSubTypes task : pool.getThreadRunables()) {
  result.addAll(task.result);
 }
 refEle = W3cUtils.addEle(rootEle,"E");
 refEle.setAttribute("id",baseTypeId);
 for (final String itm : result) {
  W3cUtils.addEle(refEle,"F").setAttribute("id",Base64.encodeToString(itm.getBytes(),false));
 }
 try {
  W3cUtils.writeXmlDocument(histFile,histDom);
 } catch (final Exception e) {
 }
 return result;
 }

 /**
 * @see {@link ReflectUtils#createSharedReflections(String...)}
 * @see {@link ReflectUtils#setSharedReflections(Reflections)}
 * @see {@link ReflectUtils#listSubClass(Class)}
 * @param baseType
 * @return
 */
 public Collection<String> getSubTypes(final Class<?> baseType,final String... typeNames) {//
 final ThreadPool<ListSubTypes> pool = new ThreadPool<ListSubTypes>();
 for (final URL pathUrl : pathUrls) {
  if (!acceptPathUrl(pathUrl))
  continue;
  File file = null;
  try {
  file = new File(URLDecoder.decode(pathUrl.getFile(),"UTF-8"));
  } catch (final Exception e) {
  file = new File(pathUrl.getFile());
  }
  pool.execute(new ListSubTypes(baseType,file,pathUrl,typeNames));
 }
 try {
  pool.shutdown(3,TimeUnit.MINUTES);
 } catch (final InterruptedException e) {
  e.printStackTrace();//for debug
 }
 final Collection<String> result = new ArrayList<String>();
 for (final ListSubTypes task : pool.getThreadRunables()) {
  result.addAll(task.result);
 }
 return result;
 }

 class ListSubTypes implements Runnable {
 final File file;
 final Class<?> baseType;
 final URL pathUrl;
 final String[] typeNames;

 public ListSubTypes(final Class<?> baseType,final File file,final URL pathUrl,final String... typeNames) {
  this.baseType = baseType;
  this.file = file;
  this.pathUrl = pathUrl;
  this.typeNames = typeNames;
 }

 Collection<String> result = new ArrayList<String>(4);

 @Override
 public void run() {
  if (file.isDirectory()) {
  listSubTypesFromDirectory(file,baseType,result,typeNames);
  } else
  listSubTypesFromJar(baseType,typeNames);
 }
 }

 /**
 * @param baseType
 * @param pathUrl
 * @param result
 */
 public void listSubTypesFromDirectory(final File baseDirectory,final Class<?> baseType,final File directory,final Collection<String> result,final String... typeNames) {
 File[] files = directory.listFiles();
 if (null == files)
  files = new File[] {};
 String clazzPath;
 final int baseDirLen = baseDirectory.getAbsolutePath().length() + 1;
 for (final File file : files) {
  if (file.isDirectory()) {
  listSubTypesFromDirectory(baseDirectory,typeNames);
  } else {
  clazzPath = file.getAbsolutePath().substring(baseDirLen);
  clazzPath = clazzPath.replace('','/');
  doTypesFilter(baseType,clazzPath,typeNames);
  }
 }
 }

 /**
 * @param baseType
 * @param pathUrl
 * @param result
 */
 public void listSubTypesFromJar(final Class<?> baseType,URL pathUrl,final String... typeNames) {
 try {
  // It does not work with the filesystem: we must
  // be in the case of a package contained in a jar file.
  JarFile jarFile = null;
  try {
  if ("file".equals(pathUrl.getProtocol()))
   pathUrl = new URL("jar:" + pathUrl.toExternalForm() + "!/");
  jarFile = ((JarURLConnection) pathUrl.openConnection()).getJarFile();
  } catch (final Exception e) {
  final String filePath = pathUrl.getFile();
  // if on win platform
  if (filePath.indexOf(':') != -1) {
   if (pathUrl.getFile().charAt(0) == '/')
   jarFile = new JarFile(filePath.substring(1));
  }
  if (null == jarFile)
   jarFile = new JarFile(filePath);
  }
  final Enumeration<JarEntry> e = jarFile.entries();
  ZipEntry entry;
  while (e.hasMoreElements()) {
  entry = e.nextElement();
  doTypesFilter(baseType,entry.getName(),typeNames);
  }
 } catch (final IOException ioex) {
 }
 }

 private void doTypesFilter(final Class<?> baseType,final String clazzPath,final String... typeNames) {
 if (!clazzPath.endsWith(".class"))
  return;
 final int lastDotIdx = clazzPath.lastIndexOf('.');
 if (-1 == lastDotIdx)
  return;
 final String typeDef = clazzPath.substring(0,lastDotIdx).replace('/','.');
 if (null != typeNames && typeNames.length > 0) {
  final int lastDot = typeDef.lastIndexOf('.');
  if (lastDot == -1)
  return;
  final String typeName = typeDef.substring(lastDot + 1);
  boolean withLiked = false;
  for (final String tmpTypeName : typeNames) {
  if (!typeName.contains(tmpTypeName))
   continue;
  withLiked = true;
  break;
  }
  if (withLiked == false)
  return;
 }
 if (this.typeFilters.isEmpty()) {
  if (null == this.subTypeFilter || this.subTypeFilter.accept(baseType,clazzPath))
  result.add(typeDef);
 } else {
  for (final ITypeFilter typeFilter : this.typeFilters) {
  if (!typeFilter.accept(clazzPath))
   continue;
  if (null == this.subTypeFilter || this.subTypeFilter.accept(baseType,clazzPath))
   result.add(typeDef);
  }
 }
 }

 /**
 * @param pathUrl
 * @return
 */
 private boolean acceptPathUrl(final URL pathUrl) {
 if (this.pathURLfilters.isEmpty())
  return true;
 for (final IPathURLFilter pathURLFilter : this.pathURLfilters) {
  if (pathURLFilter.accept(pathUrl))
  return true;
 }
 return false;
 }
}

(编辑:安卓应用网)

【声明】本站内容均来自网络,其相关言论仅代表作者个人观点,不代表本站立场。若无意侵犯到您的权利,请及时与联系站长删除相关内容!

    推荐文章
      热点阅读