Currently, I'm trying to move a docx to a mediawiki file and preserve the proper filenames in the [[Image:]] tags. For some reason, the proper image file gets swallowed (ie, normally it'd be media/image4.jpg, but instead it's just empty).
I've tried extracting the docx and looking at docx/word/_rels/document.xml.rels but I have no idea how to figure out what images are duplicated. I made a simple script to do some find/replace, but in one file I have 130 [[Image:]] tags and only 105 images.
As such, I would like to have the MediaWiki filter output the proper image name when doing this:
soffice --headless --convert-to txt:MediaWiki myfile.docx
I'm on ubuntu 14.10.
Is this possible?
This doesn't appear to be possible, but I have written a workaround found here that solves it. The long and short of it is that I convert the file and manage uploading / linking of images manually.