PDF library in Java

Basically, we need to create PDF file in our desktop application (Swing and SWT) and web application(render the PDF report in the view layer).Three Java PDF libraries come into my pocket: iText,PDFBox and ICEpdf. PDFBox is from apache project, but it is volunteer project so it is not powerful as apache other top projects. Even though iText and ICEpdf support open source usage, you need to get the license priviledge if you use them in commercial project.

As far as i know, many developers are familiar with iText if they use Spring MVC before. In Spring MVC view layer, org.springframework.web.servlet.view.document.AbstractPdfView is supplied to render PDF in your pages.

package org.springframework.web.servlet.view.document;

import java.io.ByteArrayOutputStream;
import java.io.OutputStream;
import java.util.Map;
import javax.servlet.http.HttpServletRequest;
import javax.servlet.http.HttpServletResponse;

import com.lowagie.text.Document;
import com.lowagie.text.DocumentException;
import com.lowagie.text.PageSize;
import com.lowagie.text.pdf.PdfWriter;

import org.springframework.web.servlet.view.AbstractView;
 
/**
 * Abstract superclass for PDF views, using Bruno Lowagie's
 * <a href="http://www.lowagie.com/iText">iText</a> package.
 * Application-specific view classes will extend this class.
 * The view will be held in the subclass itself, not in a template.
 *
 * <p>Note: Internet Explorer requires a ".pdf" extension, as
 * it doesn't always respect the declared content type.
 *
 * @author Rod Johnson
 * @author Juergen Hoeller
 * @author Jean-Pierre Pawlak
 * @see AbstractPdfStamperView
 */
public abstract class AbstractPdfView extends AbstractView {

    /**
     * This constructor sets the appropriate content type "application/pdf".
     * Note that IE won't take much notice of this, but there's not a lot we
     * can do about this. Generated documents should have a ".pdf" extension.
     */
    public AbstractPdfView() {
        setContentType("application/pdf");
    }


    @Override
    protected boolean generatesDownloadContent() {
        return true;
    }

    @Override
    protected final void renderMergedOutputModel(
            Map<String, Object> model, HttpServletRequest request, HttpServletResponse response) throws Exception {

        // IE workaround: write into byte array first.
        ByteArrayOutputStream baos = createTemporaryOutputStream();

        // Apply preferences and build metadata.
        Document document = newDocument();
        PdfWriter writer = newWriter(document, baos);
        prepareWriter(model, writer, request);
        buildPdfMetadata(model, document, request);

        // Build PDF document.
        document.open();
        buildPdfDocument(model, document, writer, request, response);
        document.close();

        // Flush to HTTP response.
        writeToResponse(response, baos);
    }

    /**
     * Create a new document to hold the PDF contents.
     * <p>By default returns an A4 document, but the subclass can specify any
     * Document, possibly parameterized via bean properties defined on the View.
     * @return the newly created iText Document instance
     * @see com.lowagie.text.Document#Document(com.lowagie.text.Rectangle)
     */
    protected Document newDocument() {
        return new Document(PageSize.A4);
    }

    /**
     * Create a new PdfWriter for the given iText Document.
     * @param document the iText Document to create a writer for
     * @param os the OutputStream to write to
     * @return the PdfWriter instance to use
     * @throws DocumentException if thrown during writer creation
     */
    protected PdfWriter newWriter(Document document, OutputStream os) throws DocumentException {
        return PdfWriter.getInstance(document, os);
    }

    /**
     * Prepare the given PdfWriter. Called before building the PDF document,
     * that is, before the call to <code>Document.open()</code>.
     * <p>Useful for registering a page event listener, for example.
     * The default implementation sets the viewer preferences as returned
     * by this class's <code>getViewerPreferences()</code> method.
     * @param model the model, in case meta information must be populated from it
     * @param writer the PdfWriter to prepare
     * @param request in case we need locale etc. Shouldn't look at attributes.
     * @throws DocumentException if thrown during writer preparation
     * @see com.lowagie.text.Document#open()
     * @see com.lowagie.text.pdf.PdfWriter#setPageEvent
     * @see com.lowagie.text.pdf.PdfWriter#setViewerPreferences
     * @see #getViewerPreferences()
     */
    protected void prepareWriter(Map<String, Object> model, PdfWriter writer, HttpServletRequest request)
            throws DocumentException {

        writer.setViewerPreferences(getViewerPreferences());
    }

    /**
     * Return the viewer preferences for the PDF file.
     * <p>By default returns <code>AllowPrinting</code> and
     * <code>PageLayoutSinglePage</code>, but can be subclassed.
     * The subclass can either have fixed preferences or retrieve
     * them from bean properties defined on the View.
     * @return an int containing the bits information against PdfWriter definitions
     * @see com.lowagie.text.pdf.PdfWriter#AllowPrinting
     * @see com.lowagie.text.pdf.PdfWriter#PageLayoutSinglePage
     */
    protected int getViewerPreferences() {
        return PdfWriter.ALLOW_PRINTING | PdfWriter.PageLayoutSinglePage;
    }

    /**
     * Populate the iText Document's meta fields (author, title, etc.).
     * <br>Default is an empty implementation. Subclasses may override this method
     * to add meta fields such as title, subject, author, creator, keywords, etc.
     * This method is called after assigning a PdfWriter to the Document and
     * before calling <code>document.open()</code>.
     * @param model the model, in case meta information must be populated from it
     * @param document the iText document being populated
     * @param request in case we need locale etc. Shouldn't look at attributes.
     * @see com.lowagie.text.Document#addTitle
     * @see com.lowagie.text.Document#addSubject
     * @see com.lowagie.text.Document#addKeywords
     * @see com.lowagie.text.Document#addAuthor
     * @see com.lowagie.text.Document#addCreator
     * @see com.lowagie.text.Document#addProducer
     * @see com.lowagie.text.Document#addCreationDate
     * @see com.lowagie.text.Document#addHeader
    */
    protected void buildPdfMetadata(Map<String, Object> model, Document document, HttpServletRequest request) {
    }

    /**
     * Subclasses must implement this method to build an iText PDF document,
     * given the model. Called between <code>Document.open()</code> and
     * <code>Document.close()</code> calls.
     * <p>Note that the passed-in HTTP response is just supposed to be used
     * for setting cookies or other HTTP headers. The built PDF document itself
     * will automatically get written to the response after this method returns.
     * @param model the model Map
     * @param document the iText Document to add elements to
     * @param writer the PdfWriter to use
     * @param request in case we need locale etc. Shouldn't look at attributes.
     * @param response in case we need to set cookies. Shouldn't write to it.
     * @throws Exception any exception that occured during document building
     * @see com.lowagie.text.Document#open()
     * @see com.lowagie.text.Document#close()
     */
    protected abstract void buildPdfDocument(Map<String, Object> model, Document document, PdfWriter writer,
            HttpServletRequest request, HttpServletResponse response) throws Exception;

}

In the AbstractPdfView class, there is one abstract method named as buildPdfDocument. In your detail implementation, you have to use your data and configuration to fill up the PDF document.

 AbstractPdfView pdfView = new AbstractPdfView() {
            protected void buildPdfDocument(Map model, Document document, PdfWriter writer,
                    HttpServletRequest request, HttpServletResponse response) throws Exception {
                document.add(new Paragraph(text));
            }
        };

In your actual practise, some unexpected problem like performance may occur due to poor code standard and quality. You can refer to this article  iText Performance to find the root cause and do your code enhancement.

For PDFBox and ICEpdf, please browser apache and icesoft official site to read their wiki and guide.