nx_photogram_front_gate.nx
buildroot/runtime/nx_photogram_front_gate.nx
about
nx_photogram_front_gate.nx -- proof gate for the photogrammetry front end.
Every test here has an ORACLE (a ground truth computed independently of the code under test) and the
two headline claims each carry a NEGATIVE CONTROL, because a rotation-invariance test that only shows
"the numbers came out small" proves nothing unless you also show they come out LARGE when the mechanism
is disabled. T4 disables the steering; T5b feeds pure noise. If either control fails to fire, the gate
goes RED even when every positive test passed -- an instrument that cannot fail is not measuring.
license_tier: ORIGINAL expect_exit: 0
dependencies 2 imports · 0 importers
imports: nx_photogram_front.nxnx_epose.nx
imported by: nobody (leaf or entry point)
call flow from main pre-order; caps 40 nodes / depth 6 declared; ↻ = already shown
structs
| none |
consts
| 13 | const G_S: i64 = 96 // synthetic image side |
| 19 | const G_NPTS: i64 = 60 // synthetic 3D points |
| 20 | const G_NOUT: i64 = 18 // deliberate outlier correspondences (30%) |
| 26 | const G_TOL: i64 = 20 |
functions
| 31 | func gw(s: *u8) -> i64 |
| 38 | func gn(v: i64) -> i64 called by 1: main |
| 55 | func gchk(name: *u8, ok: i64) -> i64 |
| 64 | func g_tex(x: i64, y: i64) -> i64 called by 1: g_fill_tex |
| 70 | func g_fill_tex(img: *u8, s: i64) -> i64 |
| 85 | func g_rot90(src: *u8, dst: *u8, s: i64) -> i64 called by 1: main |
| 99 | func g_refpop(v: i64) -> i64 called by 1: main |
| 112 | func main() -> i64 |