Files
sx/examples/1129-diagnostics-array-dim-not-const.sx
agra d8076b9333 lang: rename signed integer types sN -> iN
Surface rename of the signed integer family: s1..s64 become i1..i64
(u1..u64, usize, isize unchanged). 'string' keeps the s-prefix arm in
name classification; width parsing moves to the i-prefix arm next to
isize.

Internal TypeId tags follow the surface (.s8/.s16/.s32/.s64 ->
.i8/.i16/.i32/.i64), as do mono-key mangle fragments (ptr_i64,
tu_i64_bool) and all display/diagnostic formatting (i{d}).

Migrated in the same sweep: stdlib + examples + issue repros + FFI C
companions (shared symbol names like ffi_id_i64), expected
stdout/stderr/ir snapshots, specs.md, readme.md, CLAUDE.md/AGENTS.md,
implementation_plan.md, docs/, issue writeups. Vendored stb_image and
historical flow state left untouched.

zig build test: 426/426; examples suite: 595/595.
2026-06-12 09:31:53 +03:00

25 lines
1018 B
Plaintext

// An array dimension that is not a compile-time integer constant is a hard
// error, not a silently-fabricated 0-length array. Here a type alias's
// dimension is a runtime function call (`get()`), which is genuinely not
// compile-time-known — the registration-time resolver cannot evaluate it.
//
// (A const-FOLDABLE expression dimension such as `[M + 1]` is NOT an error — it
// folds; see examples/0144-types-const-expr-array-dim.sx. Only a dimension with
// a genuinely runtime operand halts here.)
//
// Regression (issue 0083): the stateless resolver printed a non-fatal warning
// and fabricated length 0, then let compilation continue — producing a 0-byte
// alloca and corrupt element access. It now yields the `.unresolved` sentinel,
// which the alias registration surfaces as this diagnostic, aborting the build
// with a non-zero exit.
#import "modules/std.sx";
get :: () -> i64 { return 5; }
BadArr :: [get()]i64;
main :: () {
a : BadArr = ---;
a[0] = 7;
print("a0={}\n", a[0]);
}