topic Word to PDF: NullPointerException in JODConverter in Alfresco Archive

Word to PDF: NullPointerException in JODConverter

nicolasraoul — Wed, 10 Mar 2010 10:44:36 GMT

In Java, I transform a Word file to a PDF file, but get a NullPointerException in jodconverter. It seems that officeManager is null. Here is the exceptionrg.alfresco.service.cmr.repository.ContentIOException: 02100003 Content conversion failed: reader: ContentAccessor[ contentUrl=store://2010/3

Re: Word to PDF: NullPointerException in JODConverter

rascio — Sat, 17 Apr 2010 12:16:19 GMT

i've the same problem…have you solved it?

Re: Word to PDF: NullPointerException in JODConverter

nicolasraoul — Mon, 19 Apr 2010 03:52:51 GMT

i've the same problem…have you solved it?

Not yet. I am busy on other aspects of the project, but this problem is a huge threat on the whole project 😞

Re: Word to PDF: NullPointerException in JODConverter

nicolasraoul — Mon, 26 Apr 2010 09:40:18 GMT

I had disabled the PDF feature so I hadn't looked at this for ages.
Today I gave it another chance, and surprise, it works! 🙂
I have no idea what was the problem 😞

Re: Word to PDF: NullPointerException in JODConverter

rascio — Mon, 26 Apr 2010 10:37:32 GMT

yes for me is the same…
now it works correctly…bah it's random

Re: Word to PDF: NullPointerException in JODConverter

kimmoilppola — Tue, 08 Jun 2010 13:28:29 GMT

There, I fixed it!
It's not random, I looked through Alfresco sources and the actual transformation needs to be done in it's own transaction. I don't know why it needs it and the exception we get is not very hintful -
but here's a working example:


        ContentService contentService = serviceRegistry.getContentService();
        NodeService nodeService = serviceRegistry.getNodeService();
        MimetypeService mimetypeService = serviceRegistry.getMimetypeService();

        String newName = transformName(mimetypeService, document.getName(), targetMimeType, true);
        NodeRef parentNodeRef = (nodeService.getPrimaryParent(sourceNodeRef)).getParentRef();

        NodeRef targetNodeRef = serviceRegistry.getCopyService().copy(
                sourceNodeRef,
                parentNodeRef,
                ContentModel.ASSOC_CONTAINS,
                QName.createQName(NamespaceService.CONTENT_MODEL_1_0_URI, "copy"),
                false);

        nodeService.setProperty(targetNodeRef, ContentModel.PROP_NAME, newName);

        ContentReader reader = contentService.getReader(sourceNodeRef, ContentModel.PROP_CONTENT);
        String sourceMimeType = reader.getMimetype();

        if (null == reader || !reader.exists()) {
            throw new ContentException("Could not aqcuire reader");
        }

        ContentWriter writer = contentService.getWriter(targetNodeRef, ContentModel.PROP_CONTENT, true);
        writer.setMimetype(targetMimeType);
        writer.setEncoding(reader.getEncoding());

        if (null != contentService.getTransformer(sourceMimeType, targetMimeType)) {
            serviceRegistry.getTransactionService().getRetryingTransactionHelper().doInTransaction(doTransformationCallback(sourceNodeRef, targetNodeRef, reader, writer), false, true);          
        }
    }

    protected RetryingTransactionHelper.RetryingTransactionCallback doTransformationCallback(final NodeRef sourceNodeRef, final NodeRef targetNodeRef, final ContentReader reader, final ContentWriter writer) {
        return new RetryingTransactionHelper.RetryingTransactionCallback() {
            @Override
            public Object execute() throws Throwable {
                TransformationOptions options = new TransformationOptions(
                        sourceNodeRef, ContentModel.PROP_NAME, targetNodeRef, ContentModel.PROP_NAME);
                serviceRegistry.getContentService().transform(reader,writer,options);
                return null;
            }
        };
    }
‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍‍

Re: Word to PDF: NullPointerException in JODConverter

kimmoilppola — Tue, 08 Jun 2010 13:46:29 GMT

Sorry I cheered too early, it still fails randomly but seems to work more often.

Re: Word to PDF: NullPointerException in JODConverter

neilm — Mon, 25 Oct 2010 10:56:24 GMT

If anyone does manage to get a test case that reproduces this failure, please JIRA it.
Thank you!