Class to convert a LEL node from one numerical type to another. More...

#include <LELConvert.h>

Inheritance diagram for casa::LELConvert< T, F >:

Public Member Functions
	LELConvert (const CountedPtr< LELInterface< F > > &expr)
	Constructor.
	~LELConvert ()
	Destructor does nothing.
virtual void	eval (LELArray< T > &result, const Slicer &section) const
	Recursively evaluate the expression.
virtual LELScalar< T >	getScalar () const
	Recursively evaluate the scalar.
virtual Bool	prepareScalarExpr ()
	Do further preparations (e.g.
virtual String	className () const
	Get class name.
virtual Bool	lock (FileLocker::LockType, uInt nattempts)
	Handle locking/syncing of a lattice in a lattice expression.
virtual void	unlock ()
virtual Bool	hasLock (FileLocker::LockType) const
virtual void	resync ()
Private Attributes
CountedPtr< LELInterface< F > >	pExpr_p

Detailed Description

template<class T, class F>
class casa::LELConvert< T, F >

Class to convert a LEL node from one numerical type to another.

Intended use:

Internal

Review Status

Date Reviewed:: yyyy/mm/dd

Prerequisite

Etymology

This derived LEL letter class handles numerical type conversions

Synopsis

This LEL letter class is derived from LELInterface. It is used to construct LEL objects that know how to convert between numerical types, such as Double to Float. They operate on numerical Lattices and return a numerical Lattice. The LELConvert object is embedded in the tree, and the conversion actually happens at tree evaluation time.

A description of the implementation details of the LEL classes can be found in Note 216

Example

Examples are not very useful as the user would never use these classes directly. Look in LatticeExprNode.cc to see how it invokes these classes. An example of how the user would indirectly use this class (through the envelope) is:

    IPosition shape(2,5,10);
    ArrayLattice<Float> x(shape); x.set(1.0);
    ArrayLattice<Double> y(shape); 
    y.copyData(x);                 // y = x;

The LELConvert class is embedded in the tree at construction time so as to handle the conversion from Float to Double at evaluation time