OpenOffice document conversion with ooextract.py and missing fonts

I'm converting Word documents with ooextract.py from OpenOffice. If the document contains a font which is not installed on the Linux system where I run it, the conversion fails. Why is the process not using an alternative font which is available?

Another problem we have seen is with Word documents containing images. Depending on the position of the image, it sometimes 开发者_运维知识库can be converted and sometimes not. If the conversion fails, the sOffice.bin daemon takes 99% CPU usage.