關於GZIPInputStream的bug

http://cin-ie.iteye.com/blog/859822

關於GZIPInputStream的bug,在jdk的最新版本上竟然還沒解決這個問題。用到gzip的需要注意了:

 

http://bugs.sun.com/bugdatabase/view_bug.do?bug_id=4691425

 

問題描述:

 

在使用GZIPInputStream對gizp文件進行讀取的時候,使用read方法,方法返回-1,表示該文件讀取真正的結束,但是實際卻並不是這樣。

 

 

這些文件在windowsxp用win-rar解壓或者linux上zcat aaa.gz命令進行解壓後,是可以獲得文件的全部內容的。但是在使用GZIPInputStream卻不行。

 

這個bug是java在讀取gzip文件的時候,還沒有讀取完畢,過早的返回-1,造成有部分文件沒有讀取完畢。

 

不過好在已經有人解決這個問題了,下面是解決的方法:

package x.y.z;



import java.io.InputStream;

import java.io.PushbackInputStream;

import java.io.IOException;



public

class MultiMemberGZIPInputStream extends GZIPInputStream {



    public MultiMemberGZIPInputStream(InputStream in, int size) throws IOException

    {

        // Wrap the stream in a PushbackInputStream…

        super(new PushbackInputStream(in, size), size);

        this.size=size;

    }



    public MultiMemberGZIPInputStream(InputStream in) throws IOException

    {

        // Wrap the stream in a PushbackInputStream…

        super(new PushbackInputStream(in, 1024));

        this.size=-1;

    }



    private MultiMemberGZIPInputStream(MultiMemberGZIPInputStream parent) throws IOException

    {

        super(parent.in);

        this.size=-1;

        this.parent=parent.parent==null ? parent : parent.parent;

        this.parent.child=this;

    }



    private MultiMemberGZIPInputStream(MultiMemberGZIPInputStream parent, int size) throws IOException

    {

        super(parent.in, size);

        this.size=size;

        this.parent=parent.parent==null ? parent : parent.parent;

        this.parent.child=this;

    }



    private MultiMemberGZIPInputStream parent;

    private MultiMemberGZIPInputStream child;

    private int size;

    private boolean eos;



    public int read(byte[] inputBuffer, int inputBufferOffset, int inputBufferLen) throws IOException {



        if (eos) { return -1;}

        if (this.child!=null)

            return this.child.read(inputBuffer, inputBufferOffset, inputBufferLen);



        int charsRead=super.read(inputBuffer, inputBufferOffset, inputBufferLen);

        if (charsRead==-1)

        {

            // Push any remaining buffered data back onto the stream

            // If the stream is then not empty, use it to construct

            // a new instance of this class and delegate this and any

            // future calls to it…

            int n = inf.getRemaining() – 8;

            if (n > 0)

            {

                // More than 8 bytes remaining in deflater

                // First 8 are gzip trailer. Add the rest to

                // any un-read data…

                ((PushbackInputStream)this.in).unread(buf, len-n, n);

            }

            else

            {

                // Nothing in the buffer. We need to know whether or not

                // there is unread data available in the underlying stream

                // since the base class will not handle an empty file.

                // Read a byte to see if there is data and if so,

                // push it back onto the stream…

                byte[] b=new byte[1];

                int ret=in.read(b,0,1);

                if (ret==-1)

                {

                    eos=true;

                    return -1;

                }

                else

                    ((PushbackInputStream)this.in).unread(b, 0, 1);

            }



            MultiMemberGZIPInputStream child;

            if (this.size==-1)

                child=new MultiMemberGZIPInputStream(this);

            else

                child=new MultiMemberGZIPInputStream(this, this.size);

            return child.read(inputBuffer, inputBufferOffset, inputBufferLen);

        }

        else

            return charsRead;

    }



}

重寫了GZIPInputStream這個類。然後你在使用的時候直接引入MultiMemberGZIPInputStream這個類,並調用即可。

例如: 

BufferedReader     bufferedReader = new BufferedReader(new InputStreamReader(new MultiMemberGZIPInputStream(new FileInputStream(gzFile)),

 

http://blog.csdn.net/hwq1987/article/details/6279130

在解壓gz文件時,如果直接用java.util.zip.GZIPInputStream來處理問題只能解壓很少一部分內容,通過類MultiMemberGZIPInputStream 可以完全解壓一個gz文件。
 

import java.io.IOException;

import java.io.InputStream;

import java.io.PushbackInputStream;

import java.util.zip.GZIPInputStream;



public class MultiMemberGZIPInputStream extends GZIPInputStream {



public MultiMemberGZIPInputStream(InputStream in, int size)

    throws IOException {

   // Wrap the stream in a PushbackInputStream...

   super(new PushbackInputStream(in, size), size);

   this.size = size;

}



public MultiMemberGZIPInputStream(InputStream in) throws IOException {

   // Wrap the stream in a PushbackInputStream...

   super(new PushbackInputStream(in, 1024));

   this.size = -1;

}



private MultiMemberGZIPInputStream(MultiMemberGZIPInputStream parent)

    throws IOException {

   super(parent.in);

   this.size = -1;

   this.parent = parent.parent == null ? parent : parent.parent;

   this.parent.child = this;

}



private MultiMemberGZIPInputStream(MultiMemberGZIPInputStream parent,

    int size) throws IOException {

   super(parent.in, size);

   this.size = size;

   this.parent = parent.parent == null ? parent : parent.parent;

   this.parent.child = this;

}



private MultiMemberGZIPInputStream parent;



private MultiMemberGZIPInputStream child;



private int size;



private boolean eos;



public int read(byte[] inputBuffer, int inputBufferOffset,

    int inputBufferLen) throws IOException {



   if (eos) {

    return -1;

   }

   if (this.child != null)

    return this.child.read(inputBuffer, inputBufferOffset,

      inputBufferLen);



   int charsRead = super.read(inputBuffer, inputBufferOffset,

     inputBufferLen);

   if (charsRead == -1) {

    // Push any remaining buffered data back onto the stream

    // If the stream is then not empty, use it to construct

    // a new instance of this class and delegate this and any

    // future calls to it...

    int n = inf.getRemaining() - 8;

    if (n > 0) {

     // More than 8 bytes remaining in deflater

     // First 8 are gzip trailer. Add the rest to

     // any un-read data...

     ((PushbackInputStream) this.in).unread(buf, len - n, n);

    } else {

     // Nothing in the buffer. We need to know whether or not

     // there is unread data available in the underlying stream

     // since the base class will not handle an empty file.

     // Read a byte to see if there is data and if so,

     // push it back onto the stream...

     byte[] b = new byte[1];

     int ret = in.read(b, 0, 1);

     if (ret == -1) {

      eos = true;

      return -1;

     } else

      ((PushbackInputStream) this.in).unread(b, 0, 1);

    }



    MultiMemberGZIPInputStream child;

    if (this.size == -1)

     child = new MultiMemberGZIPInputStream(this);

    else

     child = new MultiMemberGZIPInputStream(this, this.size);

    return child.read(inputBuffer, inputBufferOffset, inputBufferLen);

   } else

    return charsRead;

}



}

 

應用示例:

try {

    int nnumber;



    FileInputStream fin = new FileInputStream(gzPath);



    MultiMemberGZIPInputStream MmGz = new MultiMemberGZIPInputStream(fin);

    FileOutputStream fout = new FileOutputStream(topath);



    byte[] buf = new byte[1024];



    nnumber = MmGz.read(buf, 0, buf.length);



    while (nnumber != -1) {



         fout.write(buf, 0, nnumber);

         nnumber = MmGz.read(buf, 0, buf.length);



}

    MmGz.close();

    fout.close();

    fin.close();



   } catch (Exception e) {

    e.printStackTrace();

   }

 

發表評論
所有評論
還沒有人評論,想成為第一個評論的人麼? 請在上方評論欄輸入並且點擊發布.
相關文章