Folks, I am exploring the idea of using OpenJPA internal APIs to obtain meta-data from a persistence unit (e.g. JAR) for use in custom code generation.
Since I'm a bit of a newbie to OpenJPA, it would be helpful if someone could give me some starting pointers to save me a little time. Suppose I have the path to an O/S file (e.g. /mydevice/myentity.jar). What would be the appropriate entry point into the OpenJPA APIs to have it parse the deployment descriptor (if present) and combine that meta data with the meta data collected from annotations? What would be the appropriate entry point (I I were say to use a visitor pattern) to 'visit' all the meta data thus obtained? Finally, in regard to named queries, if the meta data indicates a named query, what would be the appropriate API entry point to obtain generated SQL for that query, without actually being connected to a database (just from the meta data)? Thanks in advance.
