Hi! Just "stumbled" across this "bug"/"missing feature" in the codegenerator.
If the user has defined its own UserType, e.g. DoubleStringType then the codegenerator actually naively assumes the property is of type DoubleStringType and not String[] (which is the case for exactly this situation). My suggestion for solving this is as follows: Have the codegenerator try to instantiate each type and check if they implements UserType. If it does, then call getReturnedClass() and use that as an typename instead. - Problems: The UserType then needs to be on the classpath when the codegenerator runs., it will "slow-down" the codegenerator and there has to be some additional logic to figure out how the returned type should be defined in the java file. - Good stuff: This will be more or less transparent for the user. I had thought of just use an <meta> tag (as Gavin has suggested in another context) that said it is actually this type, and not the type listed in the hbm.xml file that should be used - Problems: The hbm.xml file then contains unneccesary redundant information that has to be repeated every time the usertype is used. - Good stuff: It it is the simplest thing that could possibly work :) Thus I suggest: First add support for an extra metatag that overrules the actual type when the codegenerator renderes the code. (This can also be used in other situations) Secondly add the "fancy" UserType handling :) hmm - maybe I should just shut up and go implement it :) /max ------------------------------------------------------- This SF.NET email is sponsored by: SourceForge Enterprise Edition + IBM + LinuxWorld = Something 2 See! http://www.vasoftware.com _______________________________________________ hibernate-devel mailing list [EMAIL PROTECTED] https://lists.sourceforge.net/lists/listinfo/hibernate-devel
