org.docx4j.model.datastorage.XHTMLAttrInjector Maven / Gradle / Ivy

Show more of this group Show more artifacts with this name
Show all versions of docx4j-core Show documentation

docx4j is a library which helps you to work with the Office Open XML file format as used in docx documents, pptx presentations, and xlsx spreadsheets.

There is a newer version: 11.4.11

Show newest version

package org.docx4j.model.datastorage;

import java.io.IOException;
import java.io.StringReader;

import javax.xml.parsers.ParserConfigurationException;

import org.docx4j.XmlUtils;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import org.w3c.dom.Document;
import org.w3c.dom.Element;
import org.w3c.dom.Node;
import org.w3c.dom.NodeList;
import org.xml.sax.InputSource;
import org.xml.sax.SAXException;

public class XHTMLAttrInjector {
	
	public static Logger log = LoggerFactory.getLogger(XHTMLAttrInjector.class);		
	
	
	protected static String injectAttrs(String content, String classVal, String styleVal) throws ParserConfigurationException, SAXException, IOException {
		
		// Convert xhtml string to DOM
    	int firstChar = content.codePointAt(0);
    	if (firstChar==0xFEFF) {
    		log.info("Removing BOM..");
    		content = content.substring(1);
    	}	
    	
    	InputSource is = new InputSource();
    	is.setCharacterStream(new StringReader(content));

    	Document doc = XmlUtils.getNewDocumentBuilder().parse(is);    	
		
		// traverse + inject (need traverse, to get p in table)
    	injectAttrs(doc, classVal, styleVal);
		
		// convert back to string again
    	return XmlUtils.w3CDomNodeToString(doc);
	}
	
	public static void injectAttrs( Node sourceNode, String classVal, String styleVal ) {
			    	
    	log.debug("node type" + sourceNode.getNodeType());
    	
        switch (sourceNode.getNodeType() ) {

	    	case Node.DOCUMENT_NODE: // type 9
        	case Node.DOCUMENT_FRAGMENT_NODE: // type 11
        
//        		log.debug("DOCUMENT:" + w3CDomNodeToString(sourceNode) );
//        		if (sourceNode.getChildNodes().getLength()==0) {
//        			log.debug("..no children!");
//        		}
        		
                // recurse on each child
                NodeList nodes = sourceNode.getChildNodes();
                if (nodes != null) {
                    for (int i=0; i clonedChildren = new ArrayList();
//                		List deletions = new ArrayList();
//                        NodeList children = sourceNode.getChildNodes();
//                        if (children != null) {
//                            for (int i=0; i");
//                    break;
//
//                case Node.COMMENT_NODE:
//                    writer.write(indentLevel + "");
//                    writer.write(lineSeparator);
//                    break;
//
//                case Node.PROCESSING_INSTRUCTION_NODE:
//                    writer.write("");
//                    writer.write(lineSeparator);
//                    break;
//
//                case Node.ENTITY_REFERENCE_NODE:
//                    writer.write("&" + node.getNodeName() + ";");
//                    break;
//
//                case Node.DOCUMENT_TYPE_NODE:
//                    DocumentType docType = (DocumentType)node;
//                    writer.write("");
//                    writer.write(lineSeparator);
//                    break;
        }
    }
	

	/**
	 * @param args
	 * @throws IOException 
	 * @throws SAXException 
	 * @throws ParserConfigurationException 
	 */
	public static void main(String[] args) throws ParserConfigurationException, SAXException, IOException {
		
//		String content = "";
		String content = "
p1
p2
p1";
		
		
		
		System.out.println( injectAttrs(content, "foo", "bar") );

	}

}