So, the time has come for features and releases.
There was discussion (around January) of adding a form of language auto-detection to BSF, a la JAXP.
What ideas or suggestions do people have on the matter? I'm up for doing implementation, but I'm looking for some thoughts on how people would expect it to work and on a preferred direction for implementation.
One this is complete, error logging is somewhat cleaned up, and I add some more docs to the tree, I think we could have a real 2.3 release candidate on our hands for a final 2.3.
Thanks in advance, Victor
I was wondering if the Commons Discovery project being at 0.2 is ready enough to use as the implementation:
http://jakarta.apache.org/commons/discovery.html
Since it not only promotes reuse of Jakarta projects and certainly seems to implement the discovery the way everyone seemed to think it should work, a la JAXP.
Some may think this impractical but one implementation would be to use org.apache.commons.discovery.ServiceDiscovery
and have SPI files META-INF/services/org.apache.bsf.BSFEngine
which contained one or multiple entries such as org.apache.bsf.BSFEngine=com.wolfram.bsf.engines.MathematicaBSFEngine
and either add a public String[] getLanguages() method on the BSFEngine interface for runtime registering discovered engines or alternatively have another entry syntax in META-INF/services/org.apache.bsf.BSFEngine that somehow specifies the language string for a given engine implementations
Jeff
--------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
