Completed this (except for one remaining outstanding bug, documented elsewhere on the blog).

The ScanCan project has the following advantages as a testbed for Cocoon/eXist functionality:

1. It uses both XSLT 1.0 and 2.0.
2. It uses XEP to generate PDFs.
3. It has some quite sophisticated AJAX-based searching.
4. It has P4 no-namespace XML as its base, with a DTD, but also has P5 transformed output available which references a namespaced schema.

Therefore it should be able to test all the functionality we care about on Lettuce. Martin will have to port this over from the old XQuery code to what the new eXist requires.