Infinite Value

Bulk Tests

Open the window via Tools > Infinite Value > Bulk Tests. It runs randomized tests across casting, parsing, formatting, arithmetic, and MathInfVal methods, comparing InfVal results against their primitive-type equivalents to catch correctness regressions.

Bulk Tests window

A primitive floating point type rounds where InfVal stays exact, so float, double, and decimal results can differ from InfVal ones without either being wrong. In the Arithmetic and Functions modes, Ignore System Type Limits counts such a result as passed and reports how many there were. A result passes when the two agree within the significant digits the type guarantees (6 for float, 15 for double, 28 for decimal), widened by how far the result can move while each primitive input stays within its own guaranteed digits. Any failure left points at InfVal.

Unit tests

A NUnit test suite is included and runs via Unity's built-in Test Runner (Window > General > Test Runner, Edit Mode tab). The suite covers:

InfValPropertiesTests: precision, isZero, isOne, isInteger, isEven, isPowerOfTwo, sign, digits, exponent.
InfValComparisonTests: All six comparison operators and CompareTo.
InfValArithmeticTests: Every arithmetic, bitwise, and shift operator.
InfValMethodsTests: ToPrecision, ToExponent, RemoveTrailingZeros, MovePointLeft/Right, Deconstruct.
InfValCastTests: Round-trip casts for all integer types, BigInteger, float, and double.
InfValParseTests: InfValFormatHelper.TryParse with known strings.
FormatterTests: Formatting output for all built-in formatters.
MathInfValTests: Most MathInfVal methods.

Run both tools after modifying any scripts to catch regressions.