Deprecated API

Contents

• Deprecated Fields
• Deprecated Methods

Deprecated Fields

org.opengis.referencing.cs.AxisDirection.BOTTOM
Use AxisDirection.DISPLAY_DOWN instead.

org.opengis.referencing.cs.AxisDirection.LEFT
Use AxisDirection.DISPLAY_LEFT instead.

org.opengis.referencing.cs.AxisDirection.RIGHT
Use AxisDirection.DISPLAY_RIGHT instead.

org.opengis.referencing.cs.AxisDirection.TOP
Use AxisDirection.DISPLAY_UP instead.

Deprecated Methods

org.opengis.referencing.cs.CSFactory.createCylindricalCS(Map, PolarCS, CoordinateSystemAxis)
Uses the method expecting 3 axis instead.

org.opengis.referencing.crs.CRSFactory.createDerivedCRS(Map, CoordinateReferenceSystem, MathTransform, CoordinateSystem)
Use the method with an OperationMethod argument instead.

org.opengis.spatialschema.geometry.primitive.PrimitiveFactory.createDirectPosition(double[])
This method moved to the GeometryFactory interface.

org.opengis.spatialschema.geometry.primitive.PrimitiveFactory.createPolyhedralSurface(List)
Moved to GeometryFactory since PolyhedralSurface and Polygon are all defined in the geometry package according ISO 19107. Furthermore, PolyhedralSurface extends Geometry.

org.opengis.referencing.crs.CRSFactory.createProjectedCRS(Map, GeographicCRS, MathTransform, CartesianCS)
Use the method with an OperationMethod argument instead.

org.opengis.referencing.crs.CRSFactory.createProjectedCRS(Map, GeographicCRS, String, GeneralParameterValue[], CartesianCS)
This method will be removed for the following reasons:

• It introduces a dependency to MathTransformFactory at the implementation level of this method. This is against the orthogonal aspect of factories (no other methods force this kind of cross-dependency).
• It brings duplication with MathTransformFactory API.
• This is mostly a convenience method; user can do the same with similar efficiency in their own code.
• Experience gained in implementation and usage suggests that it is hard to get a method that meets every need. Some code will ignore this method and build ProjectedCRS in their own way (using CRSFactory.createProjectedCRS(Map,OperationMethod,GeographicCRS,MathTransform,CartesianCS) method) anyway. They may want to use their own MathTransformFactory, use a different OperationMethod, etc.

org.opengis.spatialschema.geometry.geometry.GeometryFactory.createSurfaceBoundary(Ring, List)
Moved to PrimitiveFactory since Ring, Surface and SurfaceBoundary are all primitive objects.

org.opengis.spatialschema.geometry.geometry.PointArray.get(int)
This method may conflict with an implementation of List interface. It also raise a performance and behavior issue: should the direct position be backed by this PointArray, so changes to the position will be reflected in the PointArray and vice-versa? The proposed alternative, PointArray.get(int, DirectPosition), avoid those questions.

org.opengis.go.display.primitive.GraphicArc.ArcClosure.getArray()
Use GraphicArc.ArcClosure.values() instead.

org.opengis.coverage.SampleDimension.getCategoryNames(Locale)
Replaced by SampleDimension.getCategoryNames().

org.opengis.coverage.SampleDimension.getDescription(Locale)
Replaced by SampleDimension.getDescription().

org.opengis.spatialschema.geometry.geometry.PointArray.getDimension()
This method definition conflict with section 6.2.2.10 in ISO 19107: "The operation coordinateDimension shall return the dimension of the coordinates that define this GM_Object, which must be the same as the coordinate dimension of the coordinate reference system for this GM_Object.".

org.opengis.coverage.Coverage.getDimensionNames(Locale)
Replaced by Coverage.getDimensionNames().

org.opengis.referencing.operation.MathTransform.getDimSource()
Renamed MathTransform.getSourceDimensions() for consistency with OperationMethod.

org.opengis.referencing.operation.MathTransform.getDimTarget()
Renamed MathTransform.getTargetDimensions() for consistency with OperationMethod.

org.opengis.referencing.cs.AxisDirection.inverse()
Use AxisDirection.opposite instead.

org.opengis.spatialschema.geometry.geometry.PointArray.toArray()
This method raise a number of implementation issues: what should be the behavior if a single point in this array is modified? Should it be reflected in the PointArray? Furthermore, this method will be inefficient in some implementations, for example an implementations backed by a float[] array. Invoking toArray() in such case may create thousands of DirectPosition objects. PointArray.positions() is safer, since it expose the positions through an iterator.