Develop Reference

File download from the browser via DB has extra bytes added causing corruption messages from microsoft applications

<%
Document downloadFile = null;
String mimeType = null;
try{
downloadFile = new DocumentsDao().loadById(Long.parseLong(request.getParameter("id")));
// gets MIME type of the file
mimeType = downloadFile.getFileType();
if (mimeType == null) {
// set to binary type if MIME mapping not found
mimeType = "application/octet-stream";
}
System.out.println("MIME type: " + mimeType);
}catch (Exception e){
return;
}
// modifies response
response.reset();
response.resetBuffer();
response.setContentType(mimeType);
response.setContentLength((int) downloadFile.getDocumentData().length);
// forces download
String headerKey = "Content-Disposition";
String headerValue = String.format("attachment; filename=\"%s\"", downloadFile.getFileName());
response.setHeader(headerKey, headerValue);
// obtains response's output stream
OutputStream outStream = response.getOutputStream();
byte[] buffer = new byte[8192];
int bytesRead = -1;
System.out.println("### Length from db = "+downloadFile.getDocumentData().length);
ByteArrayInputStream inStream = new ByteArrayInputStream(downloadFile.getDocumentData());
while ((bytesRead = inStream.read(buffer)) != -1) {
outStream.write(buffer, 0, bytesRead);
}
inStream.close();
outStream.close();
response.flushBuffer();
return;
%>
The above code in a JSP produces a file to download which has an additional sequence of 0d0a x 4 at the end which causes the microsoft applications word and excel to complain and have to repair the file which has been downloaded.
I thought it might be the upload of the file but it was not, and retrieving from the database is fine. So the input stream is fine the problem occurs after the output stream is closed.
Errors you get are 'Word found unreadable content' 'Excel found unreadable content'
Has anyone seen this?
cheers
Charlie

The conversion of the JSP to servlet introduced the 0d0a X 4 characters. I used fiddle to find that the Apache Tomcat web server was altering the content length and sending the bytes. I looked at working apps at work and they all used servlets to do the job so I converted the above code to a servlet and it worked perfectly.
So don't use JSPs for this purpose.

Java: Binary File Upload using Restlet + Apache Commons FileUpload

I have a REST API with Restlet 2.3 and need to implement a file-uploading functionality to it.
The problem is that, when someone uploads a file using a POST (with a multipart/form-data Content-Type), the file reaches the server with another encoding.
To test this, I printed the contents of the original file in a Unix Terminal and then printed it again before parsing the requet with Apache Commons FileUpload (with almost the same code of this example http://restlet.com/technical-resources/restlet-framework/guide/2.2/extensions/fileupload). Both printed contents are very similar, but the original file has less characters, so i assume that my Java server is using the wrong encoding to interpret the file.
The file I sent is a PNG image. With text files the server works perfectly, but when I send photos or any binary file, the problem appears.

I don't know how you exactly did to check the received content. First you should check the content type that is used for your file part within the content of your multipart request. You should have something like that for a JPG image:
-----------------------------75956101888331271337088331
Content-Disposition: form-data; name="fileToUpload"; filename="myimage.jpg"
Content-Type: image/jpeg
Secondly, I don't know how you actually write the content you received. Apache Commons IO brings an utility method IOUtils.copy that provides a simple solution to write in an OutputStream the content received from an InputStream. See how ti can be used in your context:
while (fileIterator.hasNext()) {
FileItemStream fi = fileIterator.next();
if (fi.getFieldName().equals("fileToUpload")) {
FileOutputStream fos = new FileOutputStream(
"output"+File.separator+fi.getFieldName());
IOUtils.copy(fi.openStream(), fos);
fos.close();
}
}
IMO, the encoding aspect only applies for text not for binary content.
Hope it helps,
Thierry

I actually solved it by using Google's ByteStreams class:
while (fileIterator.hasNext()) {
FileItemStream fi = fileIterator.next();
if (fi.getFieldName().equals(FILE_TO_UPLOAD)) {
byte[] byteArray = ByteStreams.toByteArray(fi.openStream());
result = new String(byteArray,Charset.forName("ISO-8859-1"));
}
}

I had the similar problem when uploading the image file. This is how I fixed. The problem was in my case the data read from the inputstream. As it is reading from a socket no guarantee that you will have the full buffer of your array filled. Therefore you should check your data size before writing it to the outputbuffer/file. Here is my code hope it helps. Also available in repository https://github.com/esabilbulbul/java-servlet-fileupload/blob/master/README.md
// Create a new file upload handler
ServletFileUpload upload = new ServletFileUpload();
upload.setHeaderEncoding("UTF-8");
// Parse the request
FileItemIterator iter = upload.getItemIterator(request);
while (iter.hasNext())
{
FileItemStream item = iter.next();
String name = item.getFieldName();
//InputStream attachmentStream = item.openStream();
//byte[] attachmentBytes = ByteStreams.toByteArray(attachmentStream);
//InputStream stream = item.getInputStream();
InputStream stream = item.openStream();
if (item.isFormField())
{
//System.out.println("Form field " + name + " with value " + Streams.asString(stream) + " detected.");
}
else
{
System.out.println("File field " + name + " with file name "+ item.getName() + " detected.");
// Process the input stream
FileOutputStream fout= new FileOutputStream ("c:\\" + item.getName());
BufferedOutputStream bout= new BufferedOutputStream (fout);
BufferedInputStream bin= new BufferedInputStream(stream);
byte buf[] = new byte[2048];
int len=0;
while ((len = bin.read(buf)) > 0)//((bin.read(buf)) != -1)
{
bout.write(buf, 0, len);
if (len<2048)
len = len;
}
bout.close();
bin.close();
}
}

Downloading a file created in server using Java

I am trying to download a file from a given URL. The URL is not a direct file URL. When this URL is provided in browser manually, we get a prompt for download/save.
For example, http://www.my-domain.com/download/type/salary/format/excel
I have no issues in the URL which has the file name directly in the URL. In the above URL, based on the format and type, server generates the file.
In Java I am trying to download the file using the below code. The file is created, but the content is just the domain content and not the actual excel data.
URL url = new URL("http://www.my-domain.com/download/type/salary/format/excel");
HttpURLConnection connection = (HttpURLConnection) url.openConnection();
float totalDataRead = 0;
BufferedInputStream in = new BufferedInputStream(connection.getInputStream());
FileOutputStream fos = new FileOutputStream("c:\\test.xls");
BufferedOutputStream bout = new BufferedOutputStream(fos, 1024);
byte[] data = new byte[1024];
int i = 0;
while ((i = in.read(data, 0, 1024)) >= 0) {
totalDataRead = totalDataRead + i;
bout.write(data, 0, i);
}
bout.close();
in.close();

The content is whatever the server sent for that URL. You can't do anything about that from the client end. If it contained Javascript for example it won't get executed.

When you want to solve a problem you have to use the adequate tools to get the thing done. The adequate tools can be found at poi.apache.org. Have a look at Apache POI.

Java servlet and IO: Create a file without saving to disk and sending it to the user

I`m hoping can help me out with a file creation/response question.
I know how to create and save a file. I know how to send that file back to the user via a ServletOutputStream.
But what I need is to create a file, without saving it on the disk, and then send that file via the ServletOutputStream.
The code above explains the parts that I have. Any help appreciated. Thanks in Advance.
// This Creates a file
//
String text = "These days run away like horses over the hill";
File file = new File("MyFile.txt");
Writer writer = new BufferedWriter(new FileWriter(file));
writer.write(text);
writer.close();
// Missing link goes here
//
// This sends file to browser
//
InputStream inputStream = null;
inputStream = new FileInputStream("C:\\MyFile.txt");
byte[] buffer = new byte[8192];
ByteArrayOutputStream baos = new ByteArrayOutputStream();
int bytesRead;
while ( (bytesRead = inputStream.read(buffer)) != -1)
baos.write(buffer, 0, bytesRead);
response.setContentType("text/html");
response.addHeader("Content-Disposition", "attachment; filename=Invoice.txt");
byte[] outBuf = baos.toByteArray();
stream = response.getOutputStream();
stream.write(outBuf);

You don't need to save off a file, just use a ByteArray stream, try something like this:
inputStream = new ByteArrayInputStream(text.getBytes());
Or, even simpler, just do:
stream.write(text.getBytes());
As cHao suggests, use text.getBytes("UTF-8") or something similar to specify a charset other than the system default. The list of available charsets is available in the API docs for Charset.

Need help optimising bufferedReader output

I am sending a file to the browser in a servlet. The highest JDK I can use is 1.4.2, and I also have to retrieve the file via a URL. I am also trying to use "guessContentTypeFromStream", but I keep getting null which raises an exception when used in the code sample below. I currently have to hard code or work out the content-type myself.
What I would like to know is, how I can re-factor this code so the file transmission is as fast as possible and also use guessContentTypeFromStream ? (Note "res" is HttpServletResponse).
URL servletUrl = new URL(sFileURL);
URLConnection conn = servletUrl.openConnection();
int read;
BufferedInputStream bis = new BufferedInputStream(conn.getInputStream());
String sContentType =conn.guessContentTypeFromStream(conn.getInputStream());
res.setContentType(sContentType);
//res.setContentType("image/jpeg");
PrintWriter os = res.getWriter();
while((read = bis.read()) != -1){
os.write(read);
}
//Clean resources
os.flush();

This is how you normally read/writes data.
in = new BufferedInputStream(socket.getInputStream(), BUFFER_SIZE);
byte[] dataBuffer = new byte[1024 * 16];
int size = 0;
while ((size = in.read(dataBuffer)) != -1) {
out.write(dataBuffer, 0, size);
}