Jim Alateras wrote: > Tried sending this email to the 'implementers forum' but it bounced. > > I have spent a day or so looking over the java-kernel code and have some > questions and I will apologise in advance if the questions do not make > sense. > > 1. I notice that there is some talk (on the technical forum) about an > XML Schema for ADL bur couldn't quite figure out the resolution. Can I > define my archetypes using XML Schema and then load them into memory at > runtime. Does such a feature exists?
No it doesn't. But there are existing java ADL parsers, which is related to your second question. > > 2. In the openehr-java-kernel is there a class that will read an ADL > file and construct the appropriate archetypes in memory. Not in the kernel, but ADL parser just does that. Currently there are two open source java ADL parsers that I am aware of. There is one java wrapper of the reference eiffel parser done by Thomas himself. The eiffel parser is always in sync with the ADL specification, but there is no link between the eiffel parser and the java kernel currently. It can be found at http://my.openehr.org/wsvn/ref_impl_eiffel/. We at Acode made a java ADL parser from scratch and released it under GPL at http://www.acode.se/frameset.jsp?cat=4&subcat=2&lang=en. The parser is implemented purely in java (javaCC actually) so the footprint is small. More importantly, the output is the Archetype class in the java kernel. So it can be used to feed archetypes directly into the kernel during runtime. However, the ADL supported by this parser is little bit behind the specification at the moment. > > 3. How are people persisting the archetype definitions (in file, in > rdbms etc)? ADL text with versioning information in rdbms is our current choice. File format is useful during authoring process on the client side I suppose. > > 4. How are people persisting the reference model data in an RDBMS? What > patterns are you using to map the adaptive object model into a > relational database? Hibernate + Spring + DAO > > 5. Is there a document outside > (http://www.openehr.org/advice/implementers-priv/msg00014.html) that > describes how to use the API for using the RM/AM Java API in an > application (the test cases are a goo source but was looking for > something that is more descriptive)? Good question. My feeling is (please correct me if I am wrong) it is right now what's lacking in the specification. We roughly know what is a kernel and what does it suppose to do. But it will be good to have defined API for the kernel as well, e.g. for data validation and creation. > > > Once again thanks for all the good work on the java-kernel. We sincerely hope this would give your project a good start and we would in the future join forces on the development of openehr components. Regards, Rong > > cheers > </jima> > > > > > > - > If you have any questions about using this list, > please send a message to d.lloyd at openehr.org > - If you have any questions about using this list, please send a message to d.lloyd at openehr.org
