0

I am trying to convert this Word document with a header showing an image on the right

http://www.filesnack.com/files/cduiejc7

to PDF using this sample code:

https://github.com/plutext/docx4j/blob/master/src/samples/docx4j/org/docx4j/samples/ConvertOutPDF.java

Here's the result:

http://www.filesnack.com/files/ctjs659h

While the Word document has the header image on the right, the converted PDF shows it on the left.

How can I make docx4j to reproduce the original document as PDF?

Matthias
  • 1,759
  • 16
  • 34

1 Answers1

1

Your image is positioned relative to a paragraph:

      <w:drawing>
    <wp:anchor distT="0" distB="0" distL="114300" distR="114300" simplePos="0" relativeHeight="251658240" behindDoc="0" locked="0" layoutInCell="1" allowOverlap="1" wp14:anchorId="791936E3" wp14:editId="575B92C8">
      <wp:simplePos x="0" y="0"/>
      <wp:positionH relativeFrom="column">
        <wp:posOffset>5317388</wp:posOffset>
      </wp:positionH>
      <wp:positionV relativeFrom="paragraph">
        <wp:posOffset>-325755</wp:posOffset>
      </wp:positionV>

docx4j potential to support stuff like that in PDF output is limited by what XSL FO supports. See docx4j's TextBoxTest class for what we can do with text boxes.

Currently, although we can position some textBoxes; we don't do the same for floating images: https://github.com/plutext/docx4j/issues/127

In the meantime, a possible workaround for some cases (eg float right) is to use a table.

Or possibly, you could try putting the image inside a text box!

JasonPlutext
  • 15,352
  • 4
  • 44
  • 84
  • Thank you! IMHO, in a perfect world docx4j should hide this hard-to-read-and-understand XSL FO stuff completely from the user. But alas... I've tried the "image inside textbox" approach. Although it looked promising I did not manage to get reliable results. However, I got what I want by using a table. Good old layout tables... – Matthias Jul 28 '14 at 22:26