Shape script - conditionally apply?

[Dave_Bullet]

Is it possible to have shape script conditionally apply - depending on the package / view / model containing a diagram on which it appears?

My problem is.... one set of architects would like an application (a package with an "application" stereotype) to appear one way, the other group want just a plain UML rendering (ie. a standard package).

Problem is - it is the one "element" in the model and I need to be able to render it 2 ways.

Any help really appreciated.

Cheers,
David.

[mark.myers]

Dave,

did you get anywhere with this?  I want to do exactly the same thing - seems like it would be a very useful feature!

Regards,
Mark.

[KP]

There is a function you can use in shape scripts to revert to default rendering: DrawNativeShape(). You can use it like this:

shape main
{
 if (HasTag("UseShapeScript","true"))
 {
   // put shape script commands here...
 }
 else
 {
   DrawNativeShape();
 }
}

You would need to define the boolean tagged value "UseShapeScript".

[mark.myers]

There is a function you can use in shape scripts to revert to default rendering: DrawNativeShape(). You can use it like this:

shape main
{
  if (HasTag("UseShapeScript","true"))
  {
    // put shape script commands here...
  }
  else
    DrawNativeShape();
}

You would need to define the boolean tagged value "UseShapeScript".

Yes I know about that but this is not what I am after.  I want to conditionally change the shape based on the diagram the shape is in NOT on any property of the element itself (because a property of the element will be the same in every diagram it is in).

[AshK]

Sorry, currently there are no means to access diagram properties from a shape script.

Would really be nice if diagrams had tagged values for this purpose.

To achieve a per diagram symbol rendering, we could expose "diagram.name" and "diagram.stereotype"?

Thoughts?

Ash

[AshK]

Oh, and diagram.type (thanks kp)

[«Midnight»]

Ash,

Sounds like a very powerful idea, and it wouldn't bend the scripts all out of shape [sorry, I just had to].  

I've got a few half-baked ideas for consideration, but will wait until I can state them clearly before sending them in via the usual channels.

David

[mark.myers]

Sorry, currently there are no means to access diagram properties from a shape script.

Would really be nice if diagrams had tagged values for this purpose.

To achieve a per diagram symbol rendering, we could expose "diagram.name" and "diagram.stereotype"?

Thoughts?

Ash

Thanks, Ash.  That sounds good.
For my immeadiate requirement I have re-thought my approach and no longer need to do this - for the time being. I still think it would be a very useful feature.

Mark.

[«Midnight»]

Does anyone (speak up Sparxians!) know whether this has yet been done? Perhaps for 804 (he says, hopefully).

[Dave_Bullet]

In terms of diagram properties... testing within a shape script is getting too fancy and doesn't actually achieve my requirement.

I'd just like a static "use shape script" property on every diagram that EA internally would use to determine whether to apply shape scripts or not.

It doesn't have to be script based.

The reason to do this - is you don't have to use a tagged value (which would require some extension to EA I believe) and you don't have to obscure the setting via a composite of stereotype, name, type etc... ie. a simple property makes it easier for a user making diagrams.

Besides - being able to test stereotype, name, type for a diagram within the shapescript doesn't actually help.  I just want to be able to say "yes" or "no" to render shapesscripts on a per diagram basis - regardless of its type, stereotype, name etc.. (there's no logic here!)

Sparx - also don't forget... whatever you apply, make sure it works for the HTML and RTF rendering too.

Cheers,
David.

[Dave_Bullet]

Hi there,

Any new features in 7.0 to allow me to test what diagram the shapescript is being executed on then conditionally apply (or drawnativeshape())

I retract my last post.  Actually being able to test diagram properties in shapescript is the way to go (as it would be the most flexible for more EA users).

AshK / KP - any more thoughts re features for 7.0?  I am happy to submit a formal request if nothing has been included in the current release.

Thanks,
DAvid.

[Dave_Bullet]

PS: The main reason I am asking for this is so ex-Visio users and UML purists can use the same EA model.

Visio people - like symbols on their deployment diagrams
UML architects - want to render nodes and the like as plain UML boxes.

I need the same node (which I have stereotyped "server") to be plain rendered or use an image depending on the diagram it is used on.

In my case, I could use diagram properties in shapescript - but would test the diagram type (ie. deployment = render image, otherwise = drawnativeshape())

Cheers,
DAvid.,

[Dave_Bullet]

Solved.  EA7.0 supports this (may have been supported earlier?) via the "diagram.type" property.

THANKYOU SPARX!    

Here's some sample script to demonstrate:

shape main
{
layouttype="border";
noshadow="true";
if(hasproperty("diagram.type", "Deployment"))
{
println("i am in a deployment diagram");
}
else
{
println("i am in something else");
}
}

[KP]

Solved.  EA7.0 supports this (may have been supported earlier?) via the "diagram.type" property.

"diagram.name" and "diagram.stereotype" have also been exposed. The full list of properties are in
the SDK on the page "Shape Scripts > Writing Scripts > Displaying Element Properties".

[Dave_Bullet]

Thanks KP - that's how I found it.  Downloaded the SDK file.  Thanks guys for putting that together.  Will come in quite handy and will save me from asking you all the annoying RTFM questions  

Cheers,
David.