fix: require exact contextual integer to float typing

2026-10-24 01:07:20 +02:00
parent 8ebbf72388
commit 81531208b2
5 changed files with 240 additions and 14 deletions
@@ -244,6 +244,7 @@ Current prototype features:
 - String literals as immutable pointers to static zero-terminated byte arrays
 - Pointer-preserving `.ptr`/`.len`, pointer-to-array indexing and slicing, postfix pointer dereference and optional unwrap, and keyed struct literals
 - Contextual integer constants and typed compile-time evaluation of arithmetic and Zig-style bitwise expressions
+- Contextual integer-to-float constants require exact representation (`f32` accepts `16777218`, not `16777217`; `f64` accepts `9007199254740994`, not `9007199254740993`), including folded and inferred constants; float literals and explicit scalar casts retain IEEE rounding
 - Integer `~`, `&`, `|`, `xor`, guarded `<<` / `>>`, saturating `<<|`, and their compound assignments; postfix `^` remains pointer dereference
 - Scalar-backed nominal `distinct` types with same-identity runtime/comptime operators, explicit backing extraction casts, integer bounds, reflection, and recursive standard formatting
 - Directory packages with merged declarations and file-local relative imports