I am troubleshooting a problem with a Word 2007 file I created from an original one (created in Word 2007). . . . I found that even if I simply open (unzip, get word/document.xml) and save the file, it messes up images in the document.  The images, which are screen shots, are mostly covered by a large grey block (treated as part of the image).

    Any idea on what could cause this?    

    I am doing this in Java.  The relevant source code may be found below.      Thanks, Alan

    public static void removeHyperlinks(String DocxFileName) throws
Exception
    {
        File file = new File(DocxFileName);
        ZipFile DocxFile = new ZipFile(file);
        if (file.exists())
        {
            OpenXmlFile DocxPack = new
OpenXmlFile(DocxFile.getName());
            Document doc =
DocxPack.getOpenXMLdoc("word","document.xml");

            OpenXmlFile.saveDocXfile(DocxFile, doc);
            DocxFile.close();
        }
    }

public class OpenXmlFile extends ZipFile
{
    public OpenXmlFile(String filename) throws Exception
    {
        super(filename);
    }
    public Enumeration getOpenXMLEntries()
    {
        return this.entries();
    }
    public static void saveDocXfile(ZipFile docxFile, Document doc) throws Exception
    {
        Transformer t = TransformerFactory.newInstance().newTransformer(); 
        ByteArrayOutputStream baos = new ByteArrayOutputStream(); 
        t.transform(new DOMSource(doc), new StreamResult(baos)); 
        ZipOutputStream docxOutFile = new ZipOutputStream(new FileOutputStream(
                                                    "XMLoutput.docx")); 
        Enumeration entriesIter = docxFile.entries(); 
        while (entriesIter.hasMoreElements()) 
        {   
            ZipEntry entry = (ZipEntry)entriesIter.nextElement();   
            if (entry.getName().equals("word/document.xml"))   
            {     
                byte[] data = baos.toByteArray();     
                docxOutFile.putNextEntry(new ZipEntry(entry.getName()));     
                docxOutFile.write(data, 0, data.length);     
                docxOutFile.closeEntry();   
            }   
            else   
            {     
                InputStream incoming = docxFile.getInputStream(entry);
                int blocks = getByteArraySize(entry.getSize());
                byte[] data = new byte[1024 * blocks];
                int readCount = incoming.read(data, 0, data.length);
                docxOutFile.putNextEntry(new ZipEntry(entry.getName()));     
                docxOutFile.write(data, 0, readCount);     
                docxOutFile.closeEntry();   
            } 
        } 
        docxOutFile.close();     
    }
    private static int getByteArraySize (long EntrySize)
    {
        long size = 0;
        size = (long)((double)EntrySize / (double)1024) + 1;
        return (int)size;     
    }
. . .

 

You'll probably have more luck with docx4j (the package handling part of which was originally based on OpenXML4J); there are a few posts in our forums recently relating to image manipulation.

   Thank you, but this is the last problem to resolve for my application, then I`m done.  The image is not being saved correctly, it turns out, so I think I just need to take an InputStream from a ZipFile and turn it into a BufferedImage.

 

I am also having a similiar problem my image is deleting when I insert  an Altchunk, does anyone know a reason why or have any type of solution for that problem?