HTML to PDF converter for Java and .NET

Mass-converting of HTML documents to a PDF with PD4ML

If an HTML document structure and layout allow that, try to represent the HTML as a set of smaller independent parts, convert them separately to PDFs and merge the resulting PDFs to a single document.

A typical use case of the approach is a conversion of bank or telecom account statements to PDF. That type of document can be extremely volumous, but its structure as a rule is repeated (a sequence of content portions correspond to single pages) and easy to be split.

The below example requests for a new HTML by getNextDocument() call. It collects a chunk of such docs, convert them to a PDF and keeps in a list. After all HTMLs are converted it merges the kept PDFs to a single doc.

getNextDocument() in the sample is to be overridden with a more practical code. For example, it could request a new XML data portion from a database, transform it to HTML and return as a StringReader object.

As in the approach we deal with relatively independent HTML documents, a parallelizing of conversion and merging procedures is possible and promises performance benefits on multi-CPU workstations/servers.

import java.io.ByteArrayInputStream;
import java.io.ByteArrayOutputStream;
import java.io.File;
import java.io.FileOutputStream;
import java.io.InputStream;
import java.net.MalformedURLException;
import java.net.URL;
import java.util.ArrayList;
import java.util.Iterator;

import org.zefer.pd4ml.PD4Document;
import org.zefer.pd4ml.PD4ML;

public class MassConvert {
	
	private final static int DOCS_PER_CHUNK = 20;
	
	private static ArrayList chunks = new ArrayList();
	
	public URL getNextDocument() throws MalformedURLException {
		// to be overridden
		//
		// if source HTML documents are generated on-a-fly, probably it is a good idea 
		// to change the method to
		// public StringReader getNextDocument()
		// and to adjust the rest of the class code correspondingly
		return new URL("file:/O:/work/testarea/AccountStatement.htm");
	}

	public int getDocumentNumber() {
		// to be overridden
		return 1000;
	}

	public static void main(String[] args) {
		MassConvert mc = new MassConvert();
		mc.convert();
	}
	
	public void convert() {
		
		long start = System.currentTimeMillis();
		int perc = 0;
		int oldperc = 0;
		
		byte[] pdf = null;

		int docNumber = getDocumentNumber();
		URL[] urls = new URL[DOCS_PER_CHUNK];
		
        try {

        	System.out.println("Converting chunks");
        	
        	for ( int i = 0; docNumber > i; i++ ) {
        		urls[i % DOCS_PER_CHUNK] = getNextDocument();
        		
                perc = i * 100 / docNumber;
                	    
            	if ( (i+1) % DOCS_PER_CHUNK == 0 || i == docNumber - 1 ) {
                	PD4ML pd4ml = new PD4ML();
                	pd4ml.setHtmlWidth(800);
                	ByteArrayOutputStream baos = new ByteArrayOutputStream();
                	pd4ml.render(urls, baos);
                	
                	pdf = baos.toByteArray();

                    if (perc / 10 > oldperc / 10) {
                        oldperc = perc;
                    	System.out.print((perc/10)*10 + "%");
                    	System.out.println(" " + 
                    			(System.currentTimeMillis() - start)/1000 + "sec");
                    }
                	
                	chunks.add(pdf);
                	pdf = null;
                	urls = new URL[DOCS_PER_CHUNK > docNumber - i ? 
                			docNumber - 1 - i : DOCS_PER_CHUNK];
            	}
        	}

        	System.out.println("Merging " + chunks.size() + " chunks");
        	
        	int i = 0;
    		ArrayList buf = new ArrayList();
        	while ( chunks.size() > 1 ) {
            	Iterator ii = chunks.iterator();
            	while ( ii.hasNext() ) {
            		pdf = (byte[])ii.next();
            		ii.remove();
                	if ( ii.hasNext() ) {
                		byte[] pdf2 = (byte[])ii.next();
                		ii.remove();
            			InputStream is1 = new ByteArrayInputStream(pdf);
            			InputStream is2 = new ByteArrayInputStream(pdf2);
            			ByteArrayOutputStream osMerge = new ByteArrayOutputStream();
            			PD4Document.mergePDFs(is1, is2, osMerge);
            			buf.add(osMerge.toByteArray());
                	} else {
                		buf.add(pdf);
                		break;
                	}

                	System.out.print('.');
                	if ( (i+1) % DOCS_PER_CHUNK == 0 ) {
                    	System.out.println(" " + 
                    			(System.currentTimeMillis() - start)/1000 + "sec");
                	}
            		i++;
            	}
            	ii = null;
            	chunks.clear();
            	chunks = buf;
            	buf = new ArrayList();
        	}

        	if ( chunks.size() != 1 ) {
        		// not likely
        		System.out.println("\nERROR?");
        	} else {
            	pdf = (byte[])chunks.get(0);
            	
                System.out.println("\ndone in " + 
                		(System.currentTimeMillis() - start)/1000 + "sec");
            	System.out.println("Resulting PDF size: " + pdf.length + "bytes");
            	
            	File pdfFile = File.createTempFile("merge", ".pdf");
            	
            	FileOutputStream fos = new FileOutputStream(pdfFile);
            	fos.write(pdf);
            	fos.close();
            	
                String params = "C:\\Program Files (x86)\\Adobe\\Reader 11.0\\Reader\\AcroRD32.exe " + 
                		pdfFile.getAbsolutePath(); 
                Runtime.getRuntime().exec(params);
        	}
        	
		} catch (Exception ex) {
			ex.printStackTrace();
		}
	}
}