Page 1 of 1
[Solved] Generating PDF without running OpenOffice engine
Posted: Fri Feb 02, 2018 2:48 am
by riddellg
Hello,
As I understand it from the documentation the way to generate a PDF from an open office text document using pure Java code, using no GUI, ie without the GUI driving Export as PDF, is to use the SDK API to make a call to a separately running OpenOffice engine via the API over HTTP. True?
I'm just getting familiar with the UNO / IDL / c++ codebase - is the PDF Filter in the c++ code, or Java code? Would it be feasible to run the ODF->PDF rendition entirely locally in a single Java JVM?
I've seen other pure Java implementations of ODF->PDF (ie xdocreport with itext underpinnings) but I'm thinking using the native OpenOffice capabilities would be more capable & more accurate. I'm hoping to do it pure Java using the OpenOffice PDF Filter approach. Is this feasible?
thanks,
-Graeme
Re: Generating PDF without running OpenOffice engine
Posted: Mon Feb 05, 2018 1:05 pm
by Zizi64
- You can use a third party virtual PDF-printer software, or
- you can write a code for opening a document with the AOO/LO but without appearing the User Interface and the document itself. use the hidden property in the openPropertyes part of your code:
Code: Select all
Dim OpenProperties(1) As New com.sun.star.beans.PropertyValue
OpenProperties(0).Name = "Hidden"
OpenProperties(0).Value = true
OpenProperties(1).Name = "Password"
OpenProperties(1).Value = "Some secret words"
Re: Generating PDF without running OpenOffice engine
Posted: Mon Feb 05, 2018 5:05 pm
by riddellg
Thankyou Tibor, as you suggested I was able to drive it as follows (Java code).
Note to others doing this in Java to read the "First Steps" doc for guidance on putting the 4 required jars in a UserLibrary and using a 32-bit JVM for compatibility with the OO DLLs. Specifically on my original question it's clear this method exits the JVM and leverages/requires an installed OpenOffice instance.
-gr
Code: Select all
/*
* Reliant on xContext = Bootstrap.bootstrap() call made in constructor..
*/
public boolean export(String odfPath, String pdfPath) throws BootstrapException, Exception {
XMultiComponentFactory xServiceManager = xContext.getServiceManager();
String available = (xServiceManager != null ? "available" : "not available");
LOG.info("remote ServiceManager is " + available);
if (xServiceManager == null)
throw new RuntimeException("Unable to get XMultiComponentFactory from OpenOffice engine");
pdfPath = pdfPath.replace('\\', '/');
// Get access to the component loader..
XComponentLoader xLoader = (XComponentLoader)UnoRuntime.queryInterface(
XComponentLoader.class,
xServiceManager.createInstanceWithContext("com.sun.star.frame.Desktop", xContext));
// Load the document..
PropertyValue[] loadProps = new PropertyValue[2];
loadProps[0] = new PropertyValue();
loadProps[0].Name = "ReadOnly";
loadProps[0].Value = new Boolean(true);
loadProps[1] = new PropertyValue();
loadProps[1].Name = "Hidden";
loadProps[1].Value = new Boolean(true);
XComponent xDocument = (XComponent) xLoader.loadComponentFromURL("file:///" + odfPath,
"_blank1", 0, loadProps);
LOG.info("doc loaded - {}", odfPath);
// Grab the storable interface..
XStorable xStorable = (XStorable) UnoRuntime.queryInterface(XStorable.class, xDocument);
// Export the PDF using the pdf writer filter..
// Note, the PDF writer supports many properties, all the same properties you see in the popup
// dialog such as lossless compression, image resolutions etc. For now taking defaults...
PropertyValue[] aMediaDescriptor = new PropertyValue[1];
aMediaDescriptor[0] = new PropertyValue();
aMediaDescriptor[0].Name = "FilterName";
aMediaDescriptor[0].Value = "writer_pdf_Export";
xStorable.storeToURL("file:///" + pdfPath, aMediaDescriptor);
LOG.info("pdf written - {}", pdfPath);
return true;
}