JAVA线程池的实现及优先级设置
来源:优易学  2011-12-9 15:59:55   【优易学:中国教育考试门户网】   资料下载   IT书店

  我们大家都知道,在处理多线程服务并发时,由于创建线程需要占用很多的系统资源,所以为了避免这些不必要的损耗,通常我们采用线程池来解决这些问题。
  线程池的基本原理是,首先创建并保持一定数量的线程,当需要使用线程时,我们从池中取得线程,再将需要运行的任务交给线程进行处理,当任务完成后再将其释放回池中。
  下面,我给出一个很简单的实现模型,仅供参考。
  ThreadPool.java
  package org.loon.framework.util.test;
  import java.util.LinkedList;
  import java.util.List;
  /** */ /**
  *
  * Title: LoonFramework
  *
  *
  * Description:
  *
  *
  * Copyright: Copyright (c) 2007
  *
  *
  * Company: LoonFramework
  *
  *
  * @author chenpeng
  * @email:ceponline@yahoo.com.cn
  * @version 0.1
  */
  public class ThreadPool ... {
  private static ThreadPool instance = null ;
  // 优先级低
  public static final int PRIORITY_LOW = 0 ;
  // 普通
  public static final int PRIORITY_NORMAL = 1 ;
  // 高
  public static final int PRIORITY_HIGH = 2 ;
  // 用以保存空闲连接
  private List[] _idxThreads;
  // 关闭
  private boolean _shutdown = false ;
  // 线程数量
  private int _threadCount = 0 ;
  // debug信息是否输出
  private boolean _debug = false ;
  /** */ /**
  * 返回ThreadPool实例
  *
  * @return
  */
  public static ThreadPool getInstance() ... {
  if (instance == null ) ... {
  instance = new ThreadPool();
  }
  return instance;
  }
  // 初始化线程list
  private ThreadPool() ... {
  this ._idxThreads = new List[] ... { new LinkedList(), new LinkedList(),
  new LinkedList() } ;
  this ._threadCount = 0 ;
  }
  /** */ /**
  * 同步方法,完成任务后将资源放回线程池中
  * @param repool
  */

  protected synchronized void repool(Pooled repool) ... {
  if ( this ._shutdown) ... {
  if ( this ._debug) ... {
  System.out.println( " ThreadPool.repool():重设中…… " );
  }
  // 优先级别判定
  switch (repool.getPriority()) ... {
  case Thread.MIN_PRIORITY:
  this ._idxThreads[PRIORITY_LOW].add(repool);
  break ;
  case Thread.NORM_PRIORITY:
  this ._idxThreads[PRIORITY_NORMAL].add(repool);
  break ;
  case Thread.MAX_PRIORITY:
  this ._idxThreads[PRIORITY_HIGH].add(repool);
  break ;
  default :
  throw new IllegalStateException( " 没有此种级别 " );
  }
  // 通知所有线程
  notifyAll();
  } else ... {
  if ( this ._debug) ... {
  System.out.println( " ThreadPool.repool():注销中…… " );
  }
  repool.shutDown();
  }
  if ( this ._debug) ... {
  System.out.println( " ThreadPool.repool():完成 " );
  }
  }
  public void setDebug( boolean debug) ... {
  this ._debug = debug;
  }
  public synchronized void shutDown() ... {
  this ._shutdown = true ;
  if ( this ._debug) ... {
  System.out.println( " ThreadPool.shutDown():关闭中…… " );
  }
  for ( int index = 0 ;index <= PRIORITY_NORMAL;index ++ ) ... {
  List threads = this ._idxThreads[index];
  for ( int threadIndex = 0 ;threadIndex < threads.size();threadIndex ++ ) ... {
  Pooled idleThread = (Pooled)threads.get(threadIndex);
  idleThread.shutDown();
  }
  }
  notifyAll();
  }
  /** */ /**
  * 以指定的优先级启动线程
  * @param target
  * @param priority
  */

  public synchronized void start(Runnable target, int priority) ... {
  Pooled thread = null ;
  List idleList = this ._idxThreads[priority];
  int idleSize = idleList.size();
  if (idleSize > 0 ) ... {
  int lastIndex = idleSize - 1 ;
  thread = (Pooled)idleList.get(lastIndex);
  idleList.remove(idleList);
  thread.setTarget(target);
  } else ... {
  this ._threadCount ++ ;
  thread = new Pooled(target, " Pooled-> " + this ._threadCount, this );
  switch (priority) ... {
  case PRIORITY_LOW:
  thread.setPriority(Thread.MIN_PRIORITY);
  break ;
  case PRIORITY_NORMAL:
  thread.setPriority(Thread.NORM_PRIORITY);
  break ;
  case PRIORITY_HIGH:
  thread.setPriority(Thread.MAX_PRIORITY);
  break ;
  default :
  thread.setPriority(Thread.NORM_PRIORITY);
  }
  // 启动
  thread.start();
  }
  }
  /** */ /**
  * 返回线程数量
  *
  * @return
  */
  public int getThreadsCount() ... {
  return this ._threadCount;
  }
  }

[1] [2] 下一页

责任编辑:小草

文章搜索:
 相关文章
热点资讯
热门课程培训