We’re getting there — 0.8 is now out. Big changes in this release: 1. The Item class now carries a binary version of the document, which can be useful for transmitting and saving it in a pipeline stage. It’s also useful for the second big change: 2. DocFilters are now a stage in the pipeline. The DocFilter interface has been refactored to look like a stage. This makes it much easier to handle documents that should generate multiple items. (Imagine an XML file with multiple subitems, or a large document that should have one item per chapter). It’s also a much cleaner design, because now connectors don’t need to know anything about DocFilters.

Plenty of bug fixes and small niceties added. Check it out.