Brainy Upload Bean.

[Al Hollis]

Was wondering if any one has similar experiances with this bean...

I have the following code based on....

OnJava Upload Bean

 
package beans;
 
import java.io.*;
import javax.servlet.*;
import javax.servlet.http.*;
 
public class FileUploadBean 
{
 
String filepath;
String filename;
 
private void setFilename(String s) 
{
    if (s==null)
    return;
 
    int pos = s.indexOf("filename=\"");
   
    if (pos != -1) 
{
         filepath = s.substring(pos+10, s.length()-1);
 
         /* Windows browsers include the full path on the client
         But Linux/Unix and Mac browsers only send the filename
         test if this is from a Windows browser
    */
         pos = filepath.lastIndexOf("\\");
         if (pos != -1)
{
        filename = filepath.substring(pos + 1);
      }
   else
{
        filename = filepath;
 }
}
}
 
public void doUpload(HttpServletRequest request, HttpServletResponse response) throws IOException 
 
{
 
response.setContentType("text/html");
PrintWriter out = response.getWriter();
 
ServletInputStream in = request.getInputStream();
 

 
byte[] line = new byte[128];
 
int readLine = in.readLine(line, 0, 128);
 
if (readLine < 3)
{
  return;
}
 

int boundaryLength = readLine - 2;
 
String boundary = new String(line, 0, boundaryLength);
 
try{
 
/* THe following while loop reads from the entire input steam, the if statement inside
checks to see when the line starts with Content-Disposition and specifically file name,
by determining this the data part of the input steam can be found and stored */
 
while (readLine != -1) 
{
String newLine = new String(line, 0, readLine);

out.println(newLine+"<BR>"); // Used to view loop through input stream,


if(newLine.startsWith("Content-Disposition: form-data; name=\"filename\""))
{

if (newLine.indexOf("filename=\"") != -1) 
{
          setFilename(new String(line, 0, readLine-2));
 

          if (filename==null)
{
            return;
}

readLine = in.readLine(line, 0, 128); //content type here
readLine = in.readLine(line, 0, 128); //blank line
readLine = in.readLine(line, 0, 128); //this is the data
 
/* THe following code gets the data part of the information from the input stream
It is then stored in the /tmp/ folder in the application directory with the file 
name that was retrieved from the input stream. */
 
newLine = new String(line, 0, readLine);
PrintWriter pw = new PrintWriter(new BufferedWriter(new FileWriter("/tmp/"+filename)));

/* This while loop now reads just from the data to the end, the if statement checks to 
   see if the current line is the last line of data, if so we take off the last to bytes of data
  as these relate to the return character witch is at the end of the data section before the final
  boundary */
 
while (readLine != -1 && !newLine.startsWith(boundary))
{
readLine = in.readLine(line,0,128);
            if ((readLine==boundaryLength+2 || readLine==boundaryLength+4) && (new String(line, 0, readLine).startsWith(boundary)))
{
pw.print(newLine.substring(0, newLine.length()-2));
newLine = new String(line, 0, readLine);
            }
else
{
             pw.print(newLine);
                    newLine = new String(line, 0, readLine);
}
//readLine = in.readLine(line,0,128);
}
out.println("File transfered");
pw.close();
}
}
else
{
   
}
readLine = in.readLine(line, 0, 128);
}
 
}
catch(Exception e)
{
 out.println(e);
}
 
}
}
 

Now this code works perfeclty with basic text files, but for images and pictures it seems to get more information than actually there, e.g a 76kb jpeg image becomes 120kb....

Any one know why this is doing this, or can point me in a better direction. I dont really want to use one of these jsp upload beans, and orielly wont work with the java version installed on the machine when program is to be run, and again is allready all there something i dont want to have Thanks in advance
[ December 12, 2005: Message edited by: Al Hollis ]

[Paul Clapham]

You're using Readers and Writers, which are specifically meant for text only, to operate on binary data such as image files. Which are not text, so they get damaged when you do that. Change your code to use InputStreams and OutputStreams; they will work for text files as well.

[Al Hollis]

Thank you very much was doing my head in lol.... Should have read the api's

Now i have a problem...

How big can i set my byte array...

byte[] line = new byte[5242880];

Does that equal 5 megs?? if so why when i upload a 1.2 meg pdf do i only recieve 78k??

Thanks in advance

[Paul Clapham]

Don't try to get the file in one bite. Assign a reasonable-sized buffer and copy the file in a loop.

[Al Hollis]

What do i copy the file to?

Just a bigger array of bytes?

[Al Hollis]

Thanks for your help, now have program working sucessfully quite proud actually