[kepler-dev] Export Archive Context Menu

[Edward A. Lee]

OK with me to check it in, but it will probably break tests...
In particular, the tests are written in Tcl which can't distinguish
constructors except by arity.  The tests should be easy enough to fix.
Christopher: Can you oversee this?

Edward

At 12:18 PM 10/6/2006, Chad Berkley wrote:
>I think I've finally figured out why this is happening.  The class
>ptolemy.actor.TypeAttribute has two constructors:
>
>public TypeAttribute(Workspace workspace)
>public TypeAttribute(TypedIOPort container, String name)
>
>The second one is what is causing problems for kepler.  TypeAttribute
>assumes that only a TypedIOPort can be its container, which makes sense
>for PTII but not for kepler.  Kepler has a PortAttribute that keeps
>track of ports within the kar files.  The parser was dying because
>Kepler was trying to assign a PortAttribute (which is not a TypedIOPort)
>as the container.  A simple fix is just to add another constructor to
>TypeAttribute that allows an Attribute to be a container like so:
>
>public TypeAttribute(Attribute container, String name)
>
>I tried this and it works fine.  Is it ok if I check this change into
>the PTII tree, or should I put it in the kepler/src/exp tree (I'd rather
>put it in PTII).
>
>chad
>
>
>Edward A. Lee wrote:
> >
> > The MoML parser will only use a two-argument constructor that
> > takes a container and a name.  The right way to set a port type
> > in a MoML file is with a type attribute.
> >
> > I'm happy to look at a concrete case, if you send me one...
> >
> > Edward
> >
> > At 01:38 PM 10/5/2006, Chad Berkley wrote:
> >
> >> Hey Dan,
> >>
> >> I fixed the ontology problem.  It now loads correctly unless you change
> >> the type of the port.  This seems to be a ptolemy bug or something.  I
> >> remember this happening before and I could never figure it out.  The
> >> error is thrown from the MoMLParser and I have no idea why it can't find
> >> the correct constructor for a port with a type.  the constructor exists.
> >>   I'll keep digging into it.
> >>
> >> chad
> >>
> >> Dan Higgins wrote:
> >> > Hi Chad,
> >> >    Yes, the 'Export Archive (KAR)...' actor context menu is just what I
> >> > was asking for!
> >> >
> >> >    I tested it by creating a customized RExpression actor where I added
> >> > 2 input ports and one output port. It writes the kar file with no
> >> > problem, but I did have several problems when I tried loading the
> >> > exported KAR file back into Kepler (using the 'File/Import
> >> Archive(KAR)'
> >> > menu.
> >> >    First, if I set the 'Type' of the output port to something like
> >> > '{int}' I see an error (Cannot find suitable constructor...')
> >> >    If I do not set  a type, I get a different error - "There are no
> >> > semantic types defined in this KAR file..." even though there were
> >> > semantic types defined in the orginal RExpression actor. Shouldn't the
> >> > saved KAR include sematic types from the original actor? (Or ask for
> >> new
> >> > ones?)
> >> >
> >> > Dan
> >> >
> >> _______________________________________________
> >> Kepler-dev mailing list
> >> Kepler-dev at ecoinformatics.org
> >> http://mercury.nceas.ucsb.edu/ecoinformatics/mailman/listinfo/kepler-dev
> >
> >
> > ------------
> > Edward A. Lee
> > Professor, Chair of EECS
> > 231 Cory Hall, UC Berkeley, Berkeley, CA 94720-1770
> > phone: 510-642-0253 or 510-642-0455, fax: 510-642-2845
> > eal at eecs.Berkeley.EDU, http://ptolemy.eecs.berkeley.edu/~eal
> >
>_______________________________________________
>Kepler-dev mailing list
>Kepler-dev at ecoinformatics.org
>http://mercury.nceas.ucsb.edu/ecoinformatics/mailman/listinfo/kepler-dev

------------
Edward A. Lee
Professor, Chair of EECS
231 Cory Hall, UC Berkeley, Berkeley, CA 94720-1770
phone: 510-642-0253 or 510-642-0455, fax: 510-642-2845
eal at eecs.Berkeley.EDU, http://ptolemy.eecs.berkeley.edu/~eal