Release notes: CalM version 4.0.0

Overview

The .sh extension was removed from the GNU/Linux launcher script.

Support for arbitrary-precision arithmetic has been added to CalM 4. Two new symbols pertaining to floating-point number types have been added:

• big: creates arbitrary-precision floating points
• small: creates standard floating points

The logical operators and and or have been made sequence points. That means they only evaluate the first argument if that's sufficient to determine the result.

API changes and new features

Significant structural changes have been made to implement arbitrary-precision arithmetic in particular. Some API changes:

• The classes Numbers and Bools in the math.mp.type package have been replaced with an interface named TypeContext. Rather than calling static methods in Numbers and Bools, an instance implementing TypeContext is passed to various constructors. The type context is created by the math parser configuration.
• A new math parser config has been added: ArbitraryPrecisionConfig. It's based on StandardConfig. ExtendedConfig changed name to ScriptConfig and based on ArbitraryPrecisionConfig.
• A class named Real has been added to the math package. It represents an arbitrary-precision real number.
• The class Fraction got the method pow(int exponent).
• The math.util package was removed. MathUtil was moved to the math package and MatrixUtil to matrix. ArgUtil, EvalUtil and TypeUtil was moved to math.mp.util.
• The interface math.rpn.override.OperationOverride was deleted. Overridable.addOverride now accepts Operation instead. A special operand instance (Overridable.PASS) can be returned by the overriding operation to pass the calculation.