The ant task and OpenOffice.org conversion filter which build my blog now read directly from OpenOffice.org documents – I no longer need to export into an intermediate file format. This means that the source for my articles is simply a directory containing regular word processor documents. When I run the ant task, it automatically rebuilds the web site based on them.
As I add to my blog, I have also needed to extend the document conversion filter. I have now placed an updated version of the code on the site. Although the new conversion remains very limited, it now supports the following features: hyperlinks, embedded images, and unordered lists. I have also enhanced the supporting ant buildfile; it produces better permalinks in subdirectories named by year and month, and is capable of copying supporting images.