On Mon, Dec 8, 2008 at 5:27 AM, Andrew Patterson <andrewpatto at gmail.com> wrote: > The grammar (pasted in the spec) for the use_node > slot refers to object_path which is not actually > defined in the path grammar. Furthermore, I can't > actually tell from the spec whether use_node > is restricted to absolute paths, or whether it can > also accept relative paths. > > archetype_internal_ref: > SYM_USE_NODE type_identifier c_occurrences object_path > > So obviously the grammar needs to be updated to > use the right path rule, but I am also thinking that > perhaps the use_node section needs some > explanation of what types of paths are allowed > to be used and what the ramifications are > if relative paths are also acceptable.
It's implemented as absolute_path in the adl-parser on the trunk of the Java project. The grammar of absolute_path is defined in the ADL spec. All the use_node slots from the openEHR archetypes seem to use absolute path so far. I doubt the usefulness of relative_path in use_node slots. But the example ADL for use_node uses a relative_path. I agree the spec needs to be clearer. /Rong > > Andrew > _______________________________________________ > openEHR-technical mailing list > openEHR-technical at openehr.org > http://lists.chime.ucl.ac.uk/mailman/listinfo/openehr-technical >
