Hi Armin,
it seems you are asking for two things:
1) How to visualize a CAS?
2) How to handle multiple views?
1) It strongly depends on what you want to visualize. If you just want to look
at the annotations you might use the CAS Visualizer or CAS Editor. Or you can
always write your own annotator that outputs the desired visualization.
2) As a view is just a "CAS in disguise" you may simply write a view-unaware
annotator doing all the visualization on a single CAS and then use SofA mapping
when assembling the pipeline.
An example using uimaFIT:
AnalysisEngineDescription visualizer = createPrimitiveDescription(
Visualizer.class
);
AggregateBuilder builder = new AggregateBuilder();
builder.add(visualizer, "_InitialView", VIEW_1);
builder.add(visualizer, "_InitialView", VIEW_2);
...
builder.add(visualizer, "_InitialView", VIEW_N);
AnalysisEngineDescription aggr = builder.createAggregateDescription();
where VIEW_1 ... VIEW_N are the names of your views.
-Torsten
> -----Original Message-----
> From: [email protected] [mailto:[email protected]]
> Sent: Thursday, December 22, 2011 8:30 AM
> To: [email protected]
> Subject: Extracting a single view from a CAS
>
> Hi,
>
> how to extract a single view from a CAS with text sofas? It's nice to
> have a CAS with multiple views. But how to visualize all the views? I
> like to see them, to check if they are ok. Is there any analysis engine
> for doing that?
>
> I'm thinking of writing a CAS multiplier that puts each view in its own
> CAS as _InitialView. What do you think about that?
>
> Sincerely,
>
> Armin
