Windows本身的复制、粘贴操作，在二个盘间进行大文件复制或转移时，如果目标盘的写入速度太慢，会异致操作占用大量的内存，及至虚拟内存。于是考虑自己写个可以自己控制内存占用的小程序用用。
本来是用同步的，但同步的太占CPU，并且很奇怪的不能自己控制内存的占用。于是想用异步，但在进行异步不知道怎么控制异步读写的“继续”，在我写的代码里，CallBack只被执行一次，就再不被激发了。很奇怪。
struct Block { internal long offset; internal int count;}
struct DataBlock { internal Block block; internal byte[] data;}
IAsyncResult iarRead, iarWrite;
int iBufferSize = GlobalData.ConstantVaries.iMega * 4;//默认每次读取4M的数据
int iBufferTimes = GlobalData.ConstantVaries.iEight;//默认读四次
private void Conductor()
{
if (queueBlock.Count > GlobalData.ConstantVaries.iZero && queueDataBlock.Count < iBufferTimes)
{
bool blCanRead = (iarRead == null);
if (!blCanRead) blCanRead = iarRead.IsCompleted;
if (blCanRead && queueBlock.Count > GlobalData.ConstantVaries.iZero && queueDataBlock.Count < iBufferTimes)
{
DataBlock dbT = new DataBlock();
dbT.data = new byte[iBufferSize];
dbT.block = queueBlock.Dequeue();
fsSource.Seek(dbT.block.offset, System.IO.SeekOrigin.Begin);
iarRead = fsSource.BeginRead(dbT.data, GlobalData.ConstantVaries.iZero, dbT.block.count, new AsyncCallback(ReadCallback), dbT);
//Console.WriteLine(queueBlock.Count);
}
}
if (queueDataBlock.Count > GlobalData.ConstantVaries.iZero)
{
}
else
{
}
}
private void ReadCallback(IAsyncResult result)
{
DataBlock dbT = (DataBlock)result.AsyncState;
queueDataBlock.Enqueue(dbT);
Console.WriteLine(queueDataBlock.Count);
Conductor();
}
/// <summary>
/// 开始异步复制文件
/// </summary>
internal void BeginCopyFile()
{
Conductor();
}
