d8076b9333
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.
23 lines
1.0 KiB
Plaintext
23 lines
1.0 KiB
Plaintext
// A DIRECT array dimension that is genuinely not a compile-time integer (a
|
|
// runtime call) is a hard error — ONE clean diagnostic + non-zero exit. Crucially
|
|
// it must NOT fabricate a length and must NOT crash later in lowering: the bad
|
|
// var is used downstream (element store + read, `.len`), and lowering has to bail
|
|
// gracefully on the `.unresolved` type rather than `@panic` in `sizeOf` or pile
|
|
// on cascade errors.
|
|
//
|
|
// Regression (issue 0083): the stateful `resolveArrayLen` emitted the diagnostic
|
|
// then `return 0` — fabricating a 0-length array (0-byte alloca, OOB access) to
|
|
// dodge the `sizeOf` panic. It now returns null → the `.unresolved` sentinel; the
|
|
// binding's lowering bails on it (a field access on an already-diagnosed
|
|
// `.unresolved` value stays silent), so the single real diagnostic aborts the
|
|
// build with no fabrication and no panic.
|
|
#import "modules/std.sx";
|
|
|
|
get :: () -> i64 { return 5; }
|
|
|
|
main :: () {
|
|
a : [get()]i64 = ---;
|
|
a[0] = 7;
|
|
print("unreachable: {} {}\n", a.len, a[0]);
|
|
}
|