Add toFloat, fix toInt truncation, and add MIN/MAX constants to Int64/Int32/UInt64

[Copilot]

Addresses several gaps in the Int64/Int32/UInt64 API: missing toFloat, inconsistent toInt overflow behavior, and lack of MIN/MAX constants.

Int64.toFloat()

New instance method across all targets. Cross-platform and eval/C++ use high * 2^32 + unsigned_low; JVM and HL use native signed-to-float casts.

var v = Int64.parseString("9007199254740991");
trace(v.toFloat()); // 9007199254740991.0 (exact)

trace(Int64.MAX.toFloat()); // ~9.22e18

Int64.toInt() — silent truncation

Previously threw "Overflow" when the value exceeded 32-bit range. Now silently returns the low 32 bits, consistent with UInt64.toInt() and all other numeric types. Applied to all implementations (cross-platform, JVM, HL, eval, C++).

MIN/MAX constants

Added static final MIN and static final MAX to haxe.Int64, haxe.Int32, and haxe.UInt64:

Type	MIN	MAX
Int64	-2^63	2^63 - 1
Int32	-2^31	2^31 - 1
UInt64	0	2^64 - 1

Tests

• TestInt64: removed overflow-throw expectations, added testToFloat and testMinMax
• TestInt32: both test files updated to reference Int32.MAX/Int32.MIN; added testMinMax
• TestUInt64: added testMinMax

---

📍 Connect Copilot coding agent with Jira, Azure Boards or Linear to delegate work to Copilot in one click without leaving your project management tool.