使用 Flying Saucer PDF Rendering 将格式错误的 HTML 转换为 PDF

Question

In a project ^GitHub I'm trying to convert any arbitrary HTML string into a PDF version. By convert I mean parse the HTML, and render it into a PDF file.

To achieve that I'm using Flying Saucer PDF Rendering like this:

Main.java

public class Main {

    public static void main(String [] args) {
        final String ok = "<valid html here>: see github rep for real html markup here";
        final String html = "<invalid html here>: see github rep for real html markup here";
        try {
            // final byte[] bytes = generatePDFFrom(ok); // works!
            final byte[] bytes = generatePDFFrom(html); // does NOT work :(
            try(FileOutputStream fos = new FileOutputStream("sample-file.pdf")) {
                fos.write(bytes);
            }

        } catch (IOException | DocumentException e) {
            e.printStackTrace();
        }
    }

    private static byte[] generatePDFFrom(String html) throws IOException, DocumentException {
        final ITextRenderer renderer = new ITextRenderer();
        renderer.setDocumentFromString(html);
        renderer.layout();
        try (ByteArrayOutputStream fos = new ByteArrayOutputStream(html.length())) {
            renderer.createPDF(fos);
            return fos.toByteArray();
        }
    }
}

In the above code if I use the html string stored in ok variable (this is a "valid" html), it creates the PDF correctly (if you run the GitHub project by using the ok variable it will create a file sample-file.pdf inside the project folder with some rendered html).

Now, if I use the value in html variable (html with invalid tags, tags maybe not closed properly, etc) it throws the following error (the error can vary depending on the incorrect value):

ERROR:  'The markup in the document following the root element must be well-formed.'
Exception in thread "main" org.xhtmlrenderer.util.XRRuntimeException: Can't load the XML resource (using TrAX transformer). org.xml.sax.SAXParseException; lineNumber: 22; columnNumber: 9; The markup in the document following the root element must be well-formed.
    at org.xhtmlrenderer.resource.XMLResource$XMLResourceBuilder.transform(XMLResource.java:222)
    at org.xhtmlrenderer.resource.XMLResource$XMLResourceBuilder.createXMLResource(XMLResource.java:181)
    at org.xhtmlrenderer.resource.XMLResource.load(XMLResource.java:84)
    at org.xhtmlrenderer.pdf.ITextRenderer.setDocumentFromString(ITextRenderer.java:171)
    at org.xhtmlrenderer.pdf.ITextRenderer.setDocumentFromString(ITextRenderer.java:166)
    at Main.generatePDFFrom(Main.java:84)
    at Main.main(Main.java:72)
Caused by: javax.xml.transform.TransformerException: org.xml.sax.SAXParseException; lineNumber: 22; columnNumber: 9; The markup in the document following the root element must be well-formed.
    at com.sun.org.apache.xalan.internal.xsltc.trax.TransformerImpl.transform(TransformerImpl.java:740)
    at com.sun.org.apache.xalan.internal.xsltc.trax.TransformerImpl.transform(TransformerImpl.java:343)
    at org.xhtmlrenderer.resource.XMLResource$XMLResourceBuilder.transform(XMLResource.java:220)
    ... 6 more
Caused by: org.xml.sax.SAXParseException; lineNumber: 22; columnNumber: 9; The markup in the document following the root element must be well-formed.
    at com.sun.org.apache.xerces.internal.parsers.AbstractSAXParser.parse(AbstractSAXParser.java:1239)
    at com.sun.org.apache.xalan.internal.xsltc.trax.TransformerImpl.transformIdentity(TransformerImpl.java:659)
    at com.sun.org.apache.xalan.internal.xsltc.trax.TransformerImpl.transform(TransformerImpl.java:728)
    ... 8 more

Now, as far as I understood this is because of the "invalid" parts of the html string.

Important notes:

• The values assigned to variables ok and html here are just a placeholder for the question. Real ones are here .
• In the real project, the html string is an input that comes from the user. Yes, he/she must know what to put there, but, of course, he/she can do some mistakes in the html conformation, so I have to handle this.

Question(s)

• Is there any way I can "tell" to Flying Saucer PDF Rendering to ignore / autocomplete / clean itself / or any other , those "invalid" parts and move on with the creation of the PDF file ^(preferred) .
• Is there a better approach I can use in order to overcome this.

Answer 1

Since I had the same issue while using Flying Saucer to generate a PDF from an HTML, I used the HtmlCleaner library (see maven link ) to clean the HTML code before parsing into Flying Saucer library.

// Clean the html to use in the flying saucer converting tool
// get the element you want to serialize
HtmlCleaner cleaner = new HtmlCleaner();
TagNode rootTagNode = cleaner.clean(html);
// set up properties for the serializer (optional, see online docs)
CleanerProperties cleanerProperties = cleaner.getProperties();
// use the getAsString method on an XmlSerializer class
XmlSerializer xmlSerializer = new PrettyXmlSerializer(cleanerProperties);
String cleanedHtml = xmlSerializer.getAsString(rootTagNode);

// use the https://github.com/flyingsaucerproject/flyingsaucer to convert cleaned HTML to PDF
ITextRenderer renderer = new ITextRenderer();
renderer.setDocumentFromString(cleanedHtml);
// ....

Answer 2

An initial thought would be to parse your input through another library that would be able to handle html better and then toString() that library's results into the PDF Renderer.

https://jsoup.org/

Five minutes of Googling found this as a pretty reasonable library to use. There's even a test utility you can try throwing your malformed input into:

https://try.jsoup.org/