opining pdf file using servlets

[Ahmed M. Omar]

i am trying to opining pdf files in servlet application this the code

protected void doGet(HttpServletRequest request, HttpServletResponse response) throws ServletException, IOException {
    	String contractId = request.getParameter("contractId");
		String year = request.getParameter("year");
		String letter = request.getParameter("letter");
		String contType = request.getParameter("contType");
		String orgUnitId = request.getParameter("orgUnitId");
		dbConn dbConn = new dbConn();
		dbConn.connect();
		ServletOutputStream out = null;
		try {
			TContractBean tContractBean = new TContractBean();
			tContractBean.setContId(contractId);
			tContractBean.setFkTBookTBYear(year);
			tContractBean.setFkTBookTBLetter(letter);
			tContractBean.setFkTccContType(contType);
			tContractBean.setOrganizationUnitId(orgUnitId);
			Vector imgPathes = new TContractDao().getImages(dbConn.getConn(),
					tContractBean);
			
			out = response.getOutputStream();
			response.setContentType("application/pdf");
                        //FileConversions.convertAll(imgPathes) get byt[] of pdf file and iam sure that it works correctly every time
			out.write(FileConversions.convertAll(imgPathes));
			out.flush();			   
		} catch (Exception e) {
			e.printStackTrace();
		} finally {
			dbConn.disconnect();
			out.close();
		}
	}

when i runing this code the pdf file works randomly (ie:works fine for 3 times and appears corrupted 2 times)


how can i write more professional code to make the pdf file works correctly every time

[Pete Nelson]

When you get a "corrupted" pdf, what do you get? A zero-byte file, or garbage?

One possibility I see, if something goes wrong with the database, your servlet may produce no output. Any exceptions that occur prior to your out.write(...) call in the "try" block will cause that out.write(...) call to never happen.

A better solution may be to create an empty buffer before the "try" block, write to that empty buffer within the "try", and then in the "finally", write that buffer to "out". The other advantage to this is that you should also be able to create & send the "Content-Length" header, which can help the Acrobat Reader plugin know how much data it should expect.

Good luck!