open-standards/wai/wai-rs/tests/require_vectors_rule.rs
Transaction Science 648f407f09 wai: the headline decode tests fail on a missing input, and CI supplies their inputs
The tests behind WAI's whole-decode claims read inputs from outside the
repository and returned early — reporting a pass — when those inputs were
absent. No workflow supplied them, so the claims were not checked
continuously.

- WAI_REQUIRE_SOTA_VECTORS=1 turns a missing input into a failure. A path
  variable naming a missing or incomplete input fails with or without the
  switch. Local runs without either still skip, and every skip prints a
  banner saying the claim was not verified.
- wai-sota-vectors.yml regenerates the int_hyper vectors from the pretrained
  mbt2018_mean checkpoint (pinned by SHA-256, with a pinned Python stack) and
  runs int_hyper_synth and int_video_sota_keyframe with the switch set. The
  regenerated vectors identify one run, not one machine: the encoder side is
  float.
- wai-jpegai-e2e.yml fetches the pinned JPEG AI inputs from the public models
  root, checks each against a SHA-256 manifest, and runs the whole-decode and
  reconstruction-binding tests with the switch set, requiring each test by
  name and the pinned pixel hash.
- Both jobs fail on any skip banner, and trigger on the kernels and float
  helpers the decode imports.
- The int_hyper claims are stated as the checks CI runs, with its quality
  measured on the pinned image; JPEG AI cross-machine identity is stated as a
  measurement, not a registry-class claim. determinism-tiers §5–§6 record the
  jobs.

Neither workflow has yet run through the forge.

Co-Authored-By: Claude Opus 5.5 <[email protected]>
2026-09-28 18:40:59 -04:00

300 lines
10 KiB
Rust

//! The missing-input rule of `support/require_vectors.rs`, checked in the
//! environment it is about.
//!
//! The vector tests' CI jobs always set `WAI_REQUIRE_SOTA_VECTORS=1`, so those runs
//! never exercise what the helpers do WITHOUT the switch: that a variable naming a
//! test's inputs still makes them required, and that a test skips only when no
//! variable names what is missing (an unset variable, or an absent or incomplete
//! default location). These tests check that rule directly. Each case starts this
//! test binary again as a child process running the ignored `probe` test, with the
//! switch and the input variables set exactly as the case needs (and every other
//! one removed, so the parent's environment cannot leak in), then checks whether
//! the child passed or failed and what it printed.
//!
//! Their inputs are scratch directories and small files that each case creates
//! under the system temporary directory, so these tests never skip.
#[path = "support/require_vectors.rs"]
mod require_vectors;
use std::path::PathBuf;
use std::process::Command;
/// Which helper the probe runs: `dir` (`env_dir` + `load_all`, as the int_hyper
/// tests use them) or `paths` (`env_paths`, as the JPEG-AI tests use it).
const MODE: &str = "WAI_RV_PROBE";
/// The default directory the probe passes to `env_dir`.
const DEFAULT: &str = "WAI_RV_PROBE_DEFAULT";
/// The directory variable the probe reads, as `int_hyper_synth` does.
const DIR_VAR: &str = "WAI_SOTA_INT_DIR";
/// The two path variables the probe reads, as the JPEG-AI tests read three.
const PATH_VAR: &str = "WAI_JPEGAI_CODESTREAM";
const PATH_VAR2: &str = "WAI_JPEGAI_NORM";
/// The one input file the `dir` probe needs.
const INPUT: &str = "cond.json";
/// Runs one helper the way a vector test does and prints what it returned. It is
/// only meaningful in a child started by [`run`], with the environment [`run`]
/// gives it; run any other way, it returns without doing anything.
#[test]
#[ignore = "a probe that the other tests in this file run in a child process"]
fn probe() {
let Ok(mode) = std::env::var(MODE) else {
return;
};
let loaded = match mode.as_str() {
"dir" => {
let default = std::env::var(DEFAULT).expect("the default directory");
let d = require_vectors::env_dir("probe", DIR_VAR, &default, "the probe");
require_vectors::load_all("probe", &d, [INPUT], "the probe").is_some()
}
"paths" => require_vectors::env_paths("probe", [PATH_VAR, PATH_VAR2], "the probe").is_some(),
other => panic!("unknown probe mode {other}"),
};
println!(
"probe: {}",
if loaded {
"inputs loaded"
} else {
"returned without inputs"
}
);
}
/// What a probe run did.
struct Outcome {
passed: bool,
/// The child's stdout followed by its stderr.
out: String,
}
impl Outcome {
/// The probe passed because it found nothing to check: the loud skip.
fn assert_skipped_loudly(&self) {
assert!(self.passed, "expected a skip, got a failure:\n{}", self.out);
assert!(
self.out.contains("probe: returned without inputs"),
"{}",
self.out
);
// The banner CI's backstop looks for, so a skip is never read as a pass.
assert!(
self.out.contains("SKIPPED") && self.out.contains("NOT VERIFIED"),
"{}",
self.out
);
}
fn assert_failed_with(&self, why: &str) {
assert!(
!self.passed,
"expected a failure naming {why:?}, got a pass:\n{}",
self.out
);
assert!(
self.out.contains(why),
"the failure does not say {why:?}:\n{}",
self.out
);
}
}
/// Runs the probe in a child process in `mode`, with `env` set and the switch and
/// every input variable otherwise removed.
fn run(mode: &str, env: &[(&str, &str)]) -> Outcome {
let exe = std::env::current_exe().expect("this test binary");
let mut cmd = Command::new(exe);
cmd.args([
"probe",
"--exact",
"--ignored",
"--nocapture",
"--test-threads=1",
]);
for v in [require_vectors::SWITCH, MODE, DEFAULT, DIR_VAR, PATH_VAR, PATH_VAR2] {
cmd.env_remove(v);
}
cmd.env(MODE, mode);
for (k, v) in env {
cmd.env(k, v);
}
let o = cmd.output().expect("start the probe");
let out = format!(
"{}{}",
String::from_utf8_lossy(&o.stdout),
String::from_utf8_lossy(&o.stderr)
);
// A filter that matched nothing would also exit 0.
assert!(
out.contains("running 1 test"),
"the probe did not run:\n{out}"
);
Outcome {
passed: o.status.success(),
out,
}
}
/// A scratch directory unique to one test case, removed when dropped.
struct Scratch(PathBuf);
impl Scratch {
fn new(case: &str) -> Scratch {
let p = std::env::temp_dir().join(format!(
"wai-require-vectors-rule-{}-{case}",
std::process::id()
));
let _ = std::fs::remove_dir_all(&p);
std::fs::create_dir_all(&p).expect("create the scratch directory");
Scratch(p)
}
/// A path inside the scratch directory, not created.
fn at(&self, name: &str) -> String {
self.0
.join(name)
.to_str()
.expect("a UTF-8 temp path")
.to_owned()
}
}
impl Drop for Scratch {
fn drop(&mut self) {
let _ = std::fs::remove_dir_all(&self.0);
}
}
// ---- without the switch ----
#[test]
fn without_the_switch_a_named_directory_that_does_not_exist_fails() {
let s = Scratch::new("dir-missing");
let (named, default) = (s.at("not-there"), s.at("default"));
run("dir", &[(DIR_VAR, &named), (DEFAULT, &default)]).assert_failed_with("is not a directory");
}
#[test]
fn without_the_switch_a_named_path_that_is_a_file_fails_as_a_directory() {
let s = Scratch::new("dir-is-file");
let (named, default) = (s.at("a-file"), s.at("default"));
std::fs::write(&named, b"{}").unwrap();
run("dir", &[(DIR_VAR, &named), (DEFAULT, &default)]).assert_failed_with("is not a directory");
}
#[test]
fn without_the_switch_a_named_empty_directory_fails() {
// The case this file was written for: the directory exists, so only the
// missing file can fail the run.
let s = Scratch::new("dir-empty");
let (named, default) = (s.at("vectors"), s.at("default"));
std::fs::create_dir(&named).unwrap();
run("dir", &[(DIR_VAR, &named), (DEFAULT, &default)])
.assert_failed_with(&format!("{DIR_VAR} names this directory"));
}
#[test]
fn without_the_switch_an_unset_or_empty_directory_variable_skips_on_an_absent_default() {
let s = Scratch::new("dir-default");
let default = s.at("default");
run("dir", &[(DEFAULT, &default)]).assert_skipped_loudly();
run("dir", &[(DIR_VAR, ""), (DEFAULT, &default)]).assert_skipped_loudly();
}
#[test]
fn without_the_switch_an_incomplete_default_directory_skips() {
let s = Scratch::new("dir-default-empty");
let default = s.at("default");
std::fs::create_dir(&default).unwrap();
run("dir", &[(DEFAULT, &default)]).assert_skipped_loudly();
}
#[test]
fn a_named_directory_holding_the_input_loads_it() {
// The positive control: without it, a probe that always failed would make the
// failure cases above pass for the wrong reason.
let s = Scratch::new("dir-full");
let (named, default) = (s.at("vectors"), s.at("default"));
std::fs::create_dir(&named).unwrap();
std::fs::write(PathBuf::from(&named).join(INPUT), b"{}").unwrap();
let o = run("dir", &[(DIR_VAR, &named), (DEFAULT, &default)]);
assert!(
o.passed && o.out.contains("probe: inputs loaded"),
"{}",
o.out
);
assert!(!o.out.contains("NOT VERIFIED"), "{}", o.out);
}
#[test]
fn without_the_switch_a_path_variable_naming_nothing_fails() {
let s = Scratch::new("path-missing");
let other = s.at("norm.json");
std::fs::write(&other, b"{}").unwrap();
run("paths", &[(PATH_VAR, &s.at("not-there")), (PATH_VAR2, &other)])
.assert_failed_with("does not exist");
// env_paths has no default: an empty value names no existing path.
run("paths", &[(PATH_VAR, ""), (PATH_VAR2, &other)]).assert_failed_with("does not exist");
}
#[test]
fn without_the_switch_a_path_variable_naming_nothing_fails_when_another_is_unset() {
// The other variable being unset would end the test as a skip; the one that
// names a missing path must fail it first.
let s = Scratch::new("path-missing-partial");
run("paths", &[(PATH_VAR, &s.at("not-there"))]).assert_failed_with("does not exist");
run("paths", &[(PATH_VAR2, &s.at("not-there"))]).assert_failed_with("does not exist");
}
#[test]
fn without_the_switch_an_unset_path_variable_skips() {
run("paths", &[]).assert_skipped_loudly();
// One set to a file that exists, the other unset: nothing named is missing.
let s = Scratch::new("path-partial");
let file = s.at("codestream.bin");
std::fs::write(&file, b"x").unwrap();
run("paths", &[(PATH_VAR, &file)]).assert_skipped_loudly();
}
#[test]
fn path_variables_naming_files_that_exist_load_them() {
// The positive control for `env_paths`, as the directory one is for `env_dir`.
let s = Scratch::new("path-full");
let (a, b) = (s.at("codestream.bin"), s.at("norm.json"));
std::fs::write(&a, b"x").unwrap();
std::fs::write(&b, b"{}").unwrap();
let o = run("paths", &[(PATH_VAR, &a), (PATH_VAR2, &b)]);
assert!(
o.passed && o.out.contains("probe: inputs loaded"),
"{}",
o.out
);
assert!(!o.out.contains("NOT VERIFIED"), "{}", o.out);
}
#[test]
fn switch_values_that_leave_it_off_behave_as_unset() {
let s = Scratch::new("switch-off");
let default = s.at("default");
for off in ["", "0", "false", "no", "FALSE", " no "] {
run(
"dir",
&[(require_vectors::SWITCH, off), (DEFAULT, &default)],
)
.assert_skipped_loudly();
}
}
// ---- with the switch ----
#[test]
fn with_the_switch_an_absent_default_fails() {
let s = Scratch::new("switch-on");
let default = s.at("default");
for on in ["1", "true", "yes"] {
run("dir", &[(require_vectors::SWITCH, on), (DEFAULT, &default)])
.assert_failed_with("a missing input is a failure, not a skip");
run("paths", &[(require_vectors::SWITCH, on)])
.assert_failed_with("a missing input is a failure, not a skip");
}
}