On Tue, 31 Jul 2012 17:35:38 +0300, Oxygen XML Editor Support <support@oxygenxml.com> wrote:
We know about this bug and we plan to address it in a future version. You will be notified when this happens. Basically right now the keys are collected in the DITA Map after the conrefs (or topicrefs) are resolved so when the conkeyref is first encountered there are not yet any keys collected.
As you've realized, that's not the best order of processing. FYI, DITA2Go does it this way: 1. Parse original maps, including maps referenced from topicrefs and maprefs. Filter conditions during parse. 2. Kesolve keydefs in the parsed maps. 3. Resolve keyrefs to maps, including conkeyrefs. 4. Parse maps that used keyrefs and conkeyrefs. 5. Process conrefs between maps and from topics to maps (the OP's case). 6. Parse topics referenced in the maps. 7. Process conrefs between topics and from maps to topics. 8. Process autonumbers, generated lists, links, and xrefs, in that order. 9. Set page sequence. 10. Process related links. HTH! -- Jeremy H. Griffith <jeremy@omsys.com> DITA2Go site: http://www.dita2go.com/