We Keep Coding

Read excel file byte array

byte[] test = getByteArry(excelfikepath)
I have one method where it returns the bytearray of the excel .xlsx file. To read this file i need to write these byte array using FileOutputStream on one server and from there i am calling another method which will read and process that excel from the server.
There is some limitation because of which i cant read excel file directly i have to put it onto another server and process.
Just wanted to know is there any way by which i can make use of this byte array and read excel file IN MEMORY instead of writing it on server.

This will help to get byte array out of an excel file.
public static byte[] getFileByteArr(String fileName) throws InvalidFormatException, IOException {
try (OPCPackage opcPackage = OPCPackage.open(new File(fileName))) {
try (XSSFWorkbook workbook = (XSSFWorkbook) WorkbookFactory.create(opcPackage)) {
try (ByteArrayOutputStream bos = new ByteArrayOutputStream()) {
workbook.write(bos);
return bos.toByteArray();
}
}
}
}

how to get the inputStream from Workbook object in java

When I upload the file i am passing the inputstream to the workbook. Now I want to use this InputStream from workbook in another method like save where I save the file InputStream in to DB. Here is my code.
public void FileUpload(FileUploadEvent event) throws ParseException {
UploadedFile item = event.getUploadedFile();
Workbook workbook = org.apache.poi.ss.usermodel.WorkbookFactory.create(item.getInputStream());
}
Now I want to make Workbook object as instance variable and pass to another method like below.
public String save() throws SQLException, IOException{
fileId = dao.savefile(workbook,fileName);
}
In my savefile method
InputStream inptest= **workbook.getStream**
ps.setBinaryStream(2,fin,fin.available());
So inptest variable accepts InputStream which I wanted to get it from Workbook.

It sounds like what you are asking for is a way to use the InputStream for multiple purposes:
To create a Workbook object (which you're already doing)
To save the content of that InputStream somewhere else
Since reading from an InputStream is usually a one-time-only operation that cannot be repeated, then you can do the following:
Save the full content of the InputStream to a buffer.
Open two new InputStreams from the buffer.
Pass your InputStreams to your two methods.
Code might look like this:
public void FileUpload(FileUploadEvent event) throws ParseException {
UploadedFile item = event.getUploadedFile();
InputStream originalInputStream = item.getInputStream();
byte[] buffer = IOUtils.toByteArray(originalInputStream);
InputStream is1 = new ByteArrayInputStream(buffer);
InputStream is2 = new ByteArrayInputStream(buffer);
Workbook workbook = org.apache.poi.ss.usermodel.WorkbookFactory.create(is1);
}
InputStream inptest = is2;
ps.setBinaryStream(2,fin,fin.available());
Note: this uses Apache Commons IO library for IOUtils.

If you are trying to save the Workbook object to a file, there is a method write() which takes in an OutputStream. Saving to a file can then be accomplished by
FileOutputStream fos = new FileOutputStream("path/to/file/[filename]");
workbook.write(fos);
fos.close();

Appending Byte[] to end of a binary file

I'm parsing a file. I'm creating a new output file and will have to add the 'byte[] data' to it. From there I will need to append many many other 'byte[] data's to the end of the file. I'm thinking I'll get the user to add a command line parameter for the output file name as I already have them providing the file name which we are parsing. That being said if the file name is not yet created in the system I feel I should generate one.
Now, I have no idea how to do this. My program is currently using DataInputStream to get and parse the file. Can I use DataOutputStream to append? If so I'm wondering how I would append to the file and not overwrite.

If so I'm wondering how I would append to the file and not overwrite.
That's easy - and you don't even need DataOutputStream. Just FileOutputStream is fine, using the constructor with an append parameter:
FileOutputStream output = new FileOutputStream("filename", true);
try {
output.write(data);
} finally {
output.close();
}
Or using Java 7's try-with-resources:
try (FileOutputStream output = new FileOutputStream("filename", true)) {
output.write(data);
}
If you do need DataOutputStream for some reason, you can just wrap a FileOutputStream opened in the same way.

Files.write(new Path('/path/to/file'), byteArray, StandardOpenOption.APPEND);
This is for byte append. Don't forget about Exception

File file =new File("your-file");
FileWriter fileWritter = new FileWriter(file.getName(),true);
BufferedWriter bufferWritter = new BufferedWriter(fileWritter);
bufferWritter.write(your-string);
bufferWritter.close();
Of coruse put this in try - catch block.

could not delete file neither via java nor manually

I have the following situation, within a servlet a create a file and then have to delete it.
When executing the file, I figured out that the file is still in the server, so I tried to remove it manually, I can't, I get the following message :
this file is opened by another program : javaw.exe
Here is my code :
public class GenerateFile extends Action {
public ActionForward execute(ActionMapping mapping, ActionForm form,
HttpServletRequest request, HttpServletResponse response) throws IOException {
System.out.println("ok");
String fileName = request.getParameter("fileName");
Integer nbrParam = Integer.parseInt(request.getParameter("nbrParam"));
String[] valueParam = new String[nbrParam+1];
for(int i =1;i<=nbrParam;i++)
{ System.out.println(request.getParameter("param"+i));
valueParam[i]=request.getParameter("param"+i);
}
FileInputStream in = new FileInputStream("C:\\Users\\free\\Desktop\\myworkspace\\gestionRH\\WebRoot\\fiches\\"+fileName+".doc");
POIFSFileSystem fs = new POIFSFileSystem(in);
HWPFDocument doc = new HWPFDocument(fs);
Range r = doc.getRange();
for(int i=1;i<=nbrParam;i++)
{ System.out.println("<param"+i+">");
System.out.println(valueParam[i]);
r.replaceText("<param"+i+">", valueParam[i]);
}
File file = new File("C:\\Users\\free\\Desktop\\myworkspace\\gestionRH\\WebRoot\\fiches\\temp");
File temp = File.createTempFile("monfile",".doc",file);
String tempName =temp.getName();
doc.write( new FileOutputStream(temp));
OutputStream out = response.getOutputStream();
response.setContentType("application/rtf");
response.setHeader("Content-Disposition","attachment; filename=Decision");
FileInputStream in1 = new FileInputStream(temp);
byte[] buffer = new byte[4096];
int length;
while ((length = in1.read(buffer)) > 0){
out.write(buffer, 0, length);
}
in1.close();
out.flush();
System.out.println("C:\\Users\\free\\Desktop\\myworkspace\\gestionRH\\WebRoot\\fiches\\temp\\"+tempName);
File f = new File("C:\\Users\\free\\Desktop\\myworkspace\\gestionRH\\WebRoot\\fiches\\temp\\"+tempName);
f.delete();
return null;
}
}

You should close all the file-reading object instances. Besides, if you can delete the file manually, you should close java and then delete it, javaw is the process that launches java outside the console.

The problem is you are creating a new FileOutputStream(tempName) to write on that file, but never closing that outputstream (or another outputstream linked to it).
Do this:
FileOutputStream fos = newFileOutputStream(tempName);
// use it
fos.close(); // CLOSE IT!!
// then you can delete the file
Simplify
Maybe you could do the work another way, without temp files...
by example: doc.write(new FileOutputStream(tempName)) could be replaced by:
doc.write(response.getOutputStream());
This way doc sends its bytes directly to where you need them, not to a temp file eliminating the need for it.
The idea behind input/output streams is composing them. Input/OutputStream are the abstract base classes. And there are a lot of implementations:
based on memory: ByteArrayInput/OutputStream
based on files: FileInputOutputStream
compressing/decompressing to another outputstream: GZipInputOutputStream
and so on
The beauty of it is applying decorator pattern to add functionality. By example:
new GZipOutputStream(new ByteArrayOutputStream());
// creates an outputstreams that compress data received and send it to the other stream
// the BAOS then writes the received bytes to memory
new GZipOutputStream(new FileOutputStream());
// it's the same but sending compressed bytes to a file.

Seems like, you are not closing the file(out), thus it remains with the thread of this action, which is restricting it to get deleted.
Hope it helps.

maybe you should try ProcMon to find out what process exactly holds the file opened

For IO features, I would to suggest to use some kind of jar already provided by community.
For example, common-io.x-x.jar, spring-core.jar
Eg, org.apache.commons.io.FileUtils;
FileUtils.copyDirectory(from, to);
FileUtils.deleteDirectory(childDir);
FileUtils.forceDelete(springConfigDir);
FileUtils.writeByteArrayToFile(file, data);
org.springframework.util.FileSystemUtils;
FileSystemUtils.copyRecursively(from, to);
FileSystemUtils.deleteRecursively(dir);
good luck!

Whenever you open a file handler, you should close it. In a Java application that you want to run for a long period of time, you are strongly recommended to close all unused file handlers soon after you finish working with them.
Examples of common file handlers are FileOutputStream and FileInputstream. Here is a good example of how you open and close the FileOutputStream
FileOutputStream fos = null;
try {
fos = new FileOutputStream(tempName);
// do something
} catch (IOException ex) {
// deal with exceptions
} finally {
// close if fos is not null
if (fos != null) {
fos.close();
}
}
You should never do this:
doc.write( new FileOutputStream(temp));
because you can never close the file handler if it has no refernce to it.

Java FileWriter overwrite

I have a piece of code that generates new data whenever there is new data available as InputStream . The same file is overwritten everytime. Sometimes the file becomes 0 kb before it gets written. A webservice reads these files at regular intervals. I need to avoid the case when the file is 0 bytes.
How do it do this? Will locks help in this case? If the browser comes in to read a file which is locked, will the browser continue to show old data from the cache until the lock is released and file is available to be read again.
try{
String outputFile = "output.html";
FileWriter fWriter = new FileWriter(outputFile);
//write the data ...
fWriter .flush();
outputFile = "anotheroutput.html";
fWriter = new FileWriter(outputFile);
//write the data ...
fWriter .flush();
fWriter.close();
}
catch(Exception e)
{
e.prinStackTrace();
}

Try writing to a temporary file (in the same file system) and once the file write is complete move it into place using File.renameTo(). If you underlying file system supports atomic move operations (most do) then you should get the behaviour that you require. If you are running on windows you will have to make sure you close the file after reading otherwise the file move will fail.
public class Data
{
private final File file;
protected Data(String fileName) {
this.file = new File(filename);
}
/* above is in some class somehwere
* then your code brings new info to the file
*/
//
public synchronized accessFile(String data) {
try {
// Create temporary file
String tempFilename = UUID.randomUUID().toString() + ".tmp";
File tempFile = new File(tempFilename);
//write the data ...
FileWriter fWriter = new FileWriter(tempFile);
fWriter.write(data);
fWriter.flush();
fWriter.close();
// Move the new file in place
if (!tempFile.renameTo(file)) {
// You may want to retry if move fails?
throw new IOException("Move Failed");
}
} catch(Exception e) {
// Do something sensible with the exception.
e.prinStackTrace();
}
}
}

FileWriter fWriter = new FileWriter(fileName,true);
try using above :-)

Your requirement is not very clear. Do you want to write a new name file every time or you want to append to the same file or you want to over write the same file? Anyway all three cases are easy and from the API you can manage it.
If the issue is that a web service is reading the file which is not yet complete i.e. is in writing phase. In your web service you should check if the file is read only, then only you read the file. In writing phase once writing is finished set the file to read only.
The 0Kb file happens because you are overwriting the same file again. Overwriting cleans up all the data and then start writing the new content.

public class Data
{
String fileName;
protected Data(String fileName)
{
this.fileName= fileName;
return; // return from constructor often not needed.
}
/* above is in some class somehwere
* then your code brings new info to the file
*/
//
public synchronized accessFile(String data)
{
try
{
// File name to be class member.
FileWriter fWriter = new FileWriter(fileName);
//write the data ...
fWriter.write(data);
fWriter .flush();
fWriter .close();
return;
}
catch(Exception e)
{
e.prinStackTrace();
}
this is not needed:
outputFile = "anotheroutput.html";
fWriter = new FileWriter(outputFile);
//write the data ...
fWriter .flush();
fWriter.close();
that's because work on the file is a method of class Data