在使用PDFBox和Java Filter Servlet下载PDF文件之前，将其添加到PDF文件中

Question

I am trying to add a Header to a PDF file just before displaying to the client. I have a filter servlet which will be intercepting the request/response before the app asks for Save. I have tested my PDFBox class on a standalone PDF and that works. I am having a difficulty with the Filter Servlet. I have tried the HttpServletRequestWrapper but not able to get eh inputstream. It seems to be empty. Here is my filter code:

public class AddFOUOFilterServlet implements Filter {
/** The log. */
private static Log log = LogFactory.getLog(AddFOUOFilterServlet.class);

private String loginUrl = null;

/**
 * @desc - This function initializes the global variables.
 * @param filterConfig - FilterConfig settings that come from the web.xml
 */
public void init(FilterConfig filterConfig) throws ServletException {
    log.debug("Inside AddFOUOFilterServlet init ");
    ServletContext context = filterConfig.getServletContext();
} 

/**
 * @desc - the method called by the server before each server request
 * @param req - the servlet request
 * @param res - the servlet response
 * @param chain - a FilterChain object that lists all filters to be called by the server
 * @throws IOException, ServletException
 */
public void doFilter(ServletRequest req, ServletResponse res, FilterChain chain) throws IOException, ServletException { 
    //log.debug("Inside doFilter");

    // cast the HTTP request and response objects and get the url path info
    HttpServletRequest request = (HttpServletRequest)req; 
    HttpServletResponse response = (HttpServletResponse)res;  
    HttpServletRequestWrapper requestWrapper = new HttpServletRequestWrapper(request);
    InputStream is = request.getInputStream();
        log.debug("is.available() "+is.available());
        byte[] buf = new byte[1024];
        while((is.read(buf))>0) {
            log.debug("Reading the InputStream from the wrapper ");
    }
    if (request.getParameter("DocumentName") != null && request.getParameter("BlobID") != null)
    {
        log.debug("Processing the PDF URIs... ");
    try 
    {   
            AddFOUOToReport addfouo = new AddFOUOToReport();
            OutputStream os = null;
            os = addfouo.doIt(is, "For Official Use Only");
            chain.doFilter(requestWrapper, res);
    }
    catch (IOException ex)
    {
        log.error("AddFOUOToReport Filter error: " + ex.getMessage());
        response.sendRedirect(loginUrl); 
    } catch (COSVisitorException e) {
        // TODO Auto-generated catch block
        e.printStackTrace();
    }
    } else {
        chain.doFilter(req, res);
    }

}

public void destroy() {log.info("Inside destroy");  }

}

Web.xml entry:

    <!-- BEGIN AddFOUOFilter Changes  -->
<filter>
    <filter-name>AddFOUO Filter</filter-name>
    <filter-class>com.jmar.bo.controller.AddFOUOFilterServlet</filter-class>
</filter>
<filter-mapping>
    <filter-name>AddFOUO Filter</filter-name>
    <url-pattern>/cdz/*</url-pattern>
</filter-mapping>
<filter-mapping>
    <filter-name>AddFOUO Filter</filter-name>
    <url-pattern>/cdzServlet?getBlob*</url-pattern>
</filter-mapping>
<filter-mapping>
    <filter-name>AddFOUO Filter</filter-name>
    <url-pattern>/*</url-pattern>
</filter-mapping>
<!-- END AddFOUOFilter Changes -->

There is no unique pattern to catch this event that is the reason I have as /* as pattern.

Answer 1

If the InputStream is empty that simply means the browser is not sending the PDF.

I'm guessing the PDF does not come from the browser, but from the server...