[kepler-dev] [Bug 2240] - add support for null values to data passing among ports

bugzilla-daemon@ecoinformatics.org bugzilla-daemon at ecoinformatics.org
Tue Feb 7 14:23:21 PST 2006


http://bugzilla.ecoinformatics.org/show_bug.cgi?id=2240





------- Additional Comments From cxh at eecs.berkeley.edu  2006-02-07 14:23 -------
Jing will modify the EML source.
Dan will modify the R actor.
I'll will look at the XYPlot actor and possibly other actors such as
SequencePlotter.  
I'll create Array2Array and remove nils, changing the length.
SequenceToArray, preserves nils in the array.
ArrayToSequence, preserves nils in the output.

What happens when we hook up an actor that produces nils and we call getValue(),
what happens.  Right now do we throw an exception?

Also, we need to handle IntToken.intValue() for nil tokens.
Also and perhaps the constructors that take a string need to properly handle
"nil"

% set d [java::new ptolemy.data.DoubleToken "nil"]
ptolemy.kernel.util.IllegalActionException: For input string: "nil"
% set d [java::new ptolemy.data.DoubleToken "1.0"]
java0x6
% 

Perhaps we need some sort of constructor that constructs a nil token.
It makes no sense to set a Double to a value and then set it nil.
Should Double(boolean isNil) be able to create a nil token?


More information about the Kepler-dev mailing list