Re: [Geotools-gt2-users] Problems with .prj while saving a shapefile

[Jody G. <jod...@gm...>]

Hey Rodrigo:

Rodrigo Paes wrote:
> Hi Jody,
> I tried your approach and here is what happened:
>
> - When I call the createSchema, it creates the schema but the crs is null;
> - Then I call the forceSchemaCRS and it creates the .prj properly but 
> the schema is set to null;
> - Then when I call the command: featureStore = 
> (FeatureStore<SimpleFeatureType, SimpleFeature>) shapeFileDataStore
>                     .getFeatureSource(featureType.getTypeName());
>    it is not able to get the right type, throwing an exception:
>
We are quickly heading into bug report territory; and should watch this 
turn over in a debugger. As Michael pointed out we are having a problem 
transforming the bounds; something we should not let get in the way of 
changing the crs.

I was startled that forcing the CRS would set the schema to null - we 
should explore how that went wrong next.

The shapefile module maintainer is very busy; I will see if he can help 
on this one. In the meantime can we arrange a test case to show the 
problem? The part we are missing (appart from the code you provided) is 
making that initial feature collection (even with a single point in it). 
The CRS you have chosen (that produces a transform exception) is the 
critical part we will of not tested before.

>         shapeFileDataStore.createSchema(collection.getSchema());
>         SimpleFeatureType schema = shapeFileDataStore.getSchema(); // 
> gets the created schema
>         shapeFileDataStore.forceSchemaCRS(crs); // this call is 
> setting the schema to null
>         shapeFileDataStore.schema = schema; //  it sets the schema again

If we cannot get a hold of the module maintainer; I have opened up a 
couple slots for hands on geotools work and I could help you debug and 
patch this problem.

Jody