Files
msd-core/tests/complexity-trigger.test.cjs
Tom Boucher 107eb8c1d9 feat(#3753): run docs guards on the PR that changes the docs they read (#3787)
A PR whose diff is entirely under docs/ runs zero tests, so a guard whose INPUT
is shipped prose cannot protect the PR lane of the diffs it exists to check. Its
only firing opportunity is after merge, on the shared branch -- which is how next
went red on dacae9273 while the PR that caused it (#3746) was green on every
check.

The docs-lint job in .github/workflows/docs-required.yml -- an ALREADY-REQUIRED
context -- now selects and runs the docs guards that read the specific docs files
the PR changed.

  scripts/docs-guard-registry.cjs    test file -> the docs paths it reads (63)
  scripts/select-docs-guards.cjs     pure (changedPaths, registry) -> test files
  scripts/lint-docs-guard-registration.cjs   drift guard, wired into lint:ci

scripts/ci-test-scope.cjs is NOT touched -- `git diff origin/next --` on it is
empty -- so #764's saving stands and its 21 pinning tests are untouched.

Selection: exact path; trailing-slash directory prefix (boundary-checked --
docs/adrenaline.md does NOT match docs/adr/, which a naive startsWith gets
wrong); and '*' for the 6 entries that walk docs/ generally or read a computed
path. Unknown maps to '*' -- guessing narrow is how a guard silently stops
running. Measured: a typo fix selects 6 of 63; docs/AGENTS.md selects 12;
docs/COMMANDS.md selects 18.

Four things this got wrong first, each found by an independent reviewer or by
probe, and each having been asserted safe in a comment:

1. The registry started as a RULE in ci-test-scope.cjs's RULES, on the theory
   that classify()'s !codeChanged normalization made it inert. True for
   docs-ONLY diffs; false for MIXED docs+code diffs, where codeChanged is true
   and the normalization never runs:

     node scripts/ci-test-scope.cjs --files "docs/a.md src/semver.cts"
       with the RULE:  25 targeted_tests
       origin/next:     3 targeted_tests

   Category error: RULES is the scoped lane's input; a docs-guard registry is a
   lane manifest for a consumer that never calls classify(). Extracted; pinned
   by value.

2. The second attempt was a dedicated workflow with paths: [docs/**]. Such a
   workflow never reports on a non-docs PR, so it can never be a required
   context without hanging every non-docs PR -- and a non-required check does not
   block a merge, so the guard would have been advisory and #3753 unfixed.
   docs-required.yml already has no paths: filter, already supplies the required
   docs-lint context, already computes docs_changed, and already ran one docs
   guard gated on it. Generalizing that step needs no ruleset edit at all.

3. The registry and the drift lint were built from ONE path-segment heuristic, so
   both were blind identically -- and blind at the guard that motivated the issue.
   The reader-call regex required a character BEFORE its keyword, so a callee
   named exactly read( / load( / parse( / doc( / file( / content( could never
   match; and only an INLINE path.join(ROOT,'docs','X.md') argument was caught,
   missing the two-step-via-variable form -- the MAJORITY spelling -- plus
   template literals and concatenation. Detector 1 fired on 14 of ~450 files, so
   35 genuine guards sat unregistered while the lint reported 0 violations,
   including cursor-reviewer (reads docs/COMMANDS.md, asserts
   .includes('--cursor')) and inventory-headings-countfree. The "accepted blind
   spot" this shipped with was the common case, not a fringe.

4. With detection fixed the true population is 115 files: 63 genuine guards, 52
   incidental. Running all 63 in a REQUIRED check on a one-line typo fix is the
   cost #764 exists to avoid -- install.test.cjs is 7840 lines and reads exactly
   one docs file, docs/AGENTS.md, for its frontmatter. Dropping it reproduces the
   bug; running it for a typo elsewhere is waste. Hence the map.

Then a second review round found six more, all fixed here:

- fragment-single-edit-propagation.install.test.cjs was EXEMPTED as
  "overlay fixture only". False: it reads the real docs/registries/eos.json and
  asserts on a registry entry name, and reads the real ADR-0001 and asserts its
  H1. A docs-only PR touching either would have gone green and red next -- #3753
  shipping again, from inside the fix for it. Now registered against both paths,
  and all 52 remaining exemptions were re-audited one by one.
- The SUITES-collision guard compared RAW registry keys, but run-tests.cjs strips
  a leading `tests/` BEFORE its suite check. So it caught 'all' and missed
  'tests/all' -- the only spelling that can actually occur, since every key
  carries the prefix. One typo would have run all 824 test files inside the
  required job. Now normalized the same way run-tests.cjs normalizes.
- The lint failed OPEN on an unreadable tests dir or candidate file: 0 violations,
  ok:true. A guard that cannot read its input must never report success.
- The exemption ratchet gated identity only, so a baselined file that later
  STARTED asserting on shipped docs stayed exempt silently -- 52 permanently blind
  files. The baseline now fingerprints the docs paths each exempted file
  references and fails when that set changes, naming what changed.
- The exemption marker was still honored inside a multi-line template literal in
  the header window. The scanner now tracks template-literal and block-comment
  state.
- `git diff --name-only | grep '^docs/'` silently dropped C-quoted non-ASCII docs
  paths, making docs_changed=false a green zero-guard check. Both call sites now
  pass -c core.quotepath=false.
- The run step was gated on hashFiles(), which a force-committed
  .docs-guard-tests.txt would satisfy. The step now rm -f's both scratch files
  first and gates on an output it sets itself.

Three empty states, deliberately distinct, because conflating them rebuilds
#3753: an empty or malformed registry HARD-FAILS; docs changed with no guard
covering them logs and skips; no docs change is already gated. The middle state
must never be expressed as an empty --files-from, which prints `no tests in suite
"all"` and exits 0 -- a green check that guarded nothing. With the current
registry that state is unreachable, because the six '*' entries always match;
the branch is kept as defensive handling for a future registry and says so.

timeout-minutes: 15 bounds the required job against a hanging fork-supplied test;
it had none. npm ci was added because the job never installed dependencies -- the
previous single-file step got away without it, the registry does not.

docs/contributing/docs-guard-registration.md documents the rule, following its
sibling cross-platform-portability-rules.md, and CONTRIBUTING.md's CI Test
Quality Checks table links to it. It is also load-bearing: without a docs/ file
in the diff this PR would not have triggered its own lane, shipping an
unexercised change to a required check.

One unrelated fix, included because this PR surfaced it and CLAUDE.md forbids
deferring a defect found while working. On this branch's first CI run,
`full test (windows-latest, 24, shard 3/3)` was CANCELLED at exactly 30 minutes;
tests were still passing 0.8s before the cancel, so it is a wall-clock timeout,
not a hang, and a cancelled job reddens `Required tests`.

The cause is not this PR's test file, which costs ~60ms. Shard composition is
unstable: adding ONE file to the unit suite reshuffled 115 of 268 files between
shards, and shard 3 drew a heavier mix. Underneath that is a real pre-existing
defect. tests/ci-test-job-timeout-budget.test.cjs requires every lane's budget to
be >= 1.5x its MEASURED cost -- "a lane that got slower must be re-budgeted, not
excused" -- and its test-full entry recorded 19m from a windows-22 shard. That is
stale. Measured on `next` with none of this PR's changes present: 26m18s (run
32614439702, windows-latest/24 shard 3/3), 23m36s and 23m17s on shard 2/3. So the
lane costs ~26m and the 30-minute cap carried 1.14x headroom, not 1.5x. The gate
had been out of compliance with its own rule; this PR was merely the file
addition that reshuffled shard 3 past the cliff.

Fixed as that file prescribes: measuredMinutes 19 -> 27 with fresh evidence, and
test-full timeout-minutes 30 -> 45. The rule's minimum for 27m is 41; 45 is
deliberately above it because the reshuffle means per-shard worst case moves run
to run, and a budget pinned to the exact minimum would be re-breached by the next
test file anyone adds. Only that one job's timeout changed; test.yml's scope,
matrix and steps are untouched, so #764's saving is unaffected.

Raising that cap let the Windows shard finish (28m45s, inside 45) and uncovered
a real failure the 30-minute cancel had been masking:
`new quick-task branch branches off origin/main (#2916)` died with
`outcome=timed_out exitCode=null`, SIGTERM, at the 15000ms bound.

tests/quick-branching.test.cjs:149 `runStep` runs a `#!/usr/bin/env bash` script
executing MULTIPLE git commands, but was bound to GIT_TIMEOUT_MS (15000) -- the
norm for a SINGLE git plumbing call. tests/helpers/timeouts.cjs already documents
this exact failure and exists to fix it: HOOK_FANOUT_TIMEOUT_MS was created after
PR #3285 recorded "outcome=timed_out exitCode=null at exactly the 15000ms probe
bound while every other lane passed the same commit", and calls that "a bound
sized for the wrong class, not a slow machine". Our failure is that case
verbatim, so both sites move to the class norm rather than to a bigger number.

The same class also failed on `next` itself 21 hours earlier -- run 32608945654,
windows-latest/24 shard 1/3, `plan touching only src/ in a submodule project
keeps worktree isolation ENABLED` -- where tests/worktree-safety.test.cjs:5845
`runGate` fans out to `git config --file .gitmodules` under a hardcoded 30000.
Fixed too, since it is a defect in the tree regardless of which branch surfaced
it.

A survey of the whole tests/ tree found the same class-mismatch at further
bash fan-out sites bound under 60000ms, and the maintainer approved sweeping
them rather than leaving them latent to surface the same way one at a time. 16
fan-out sites across 16 files now use the class norm.

The sweep is class-correctness, not raising numbers until things pass. Sites
were moved ONLY where the bash body demonstrably spawns something (git, node,
npm, a CLI); self-contained shell snippets were left where they are, and are
listed as deliberately unchanged: pure if/printf bodies (copilot-install), pure
array/case builtins (code-review-pipeline-regression:638), a documented
pure-shell gsd_run stub (host-integration), single-process hook calls
(workflow-guard:222/271/302), and a deliberately tight 5000ms fast-check hook
(gsd-write-guard.property). Nothing was lowered. process-seam.test.cjs:513
(literal 300) is untouched on purpose -- it tests timeout BEHAVIOR, so raising
it would destroy what it asserts.

Shared file-level constants were the trap here, and were handled per file rather
than by redefinition: GIT_TIMEOUT_MS has ~15 users in git-base-branch and only 1
is a fan-out; WORKTREE_TIMEOUT_MS has 16 users in worktree.test.cjs and 3 are;
PROBE_TIMEOUT_MS has several in three more files. In each the CALL SITE was
changed and the constant left alone, so no single-plumbing-call site silently
inherited a 60s bound. The one exception is hooks-opt-in.test.cjs, where
HOOK_TIMEOUT_MS has exactly one consumer -- spawnHook, the fan-out itself -- so
redefining it is identical in effect and reads better.

Only two of these sites have actually been observed failing. The rest cite that
shared class and those two run ids rather than inventing evidence of their own.

Co-authored-by: sim <sim@local>
2026-08-23 21:21:21 -04:00

997 lines
38 KiB
JavaScript

// docs-guard-exempt: 'docs/readme.md' is a synthetic fixture path fed to isAnalyzablePath(), never read as content.
'use strict';
/**
* Complexity-triggered refactor extension point — analyzer, evaluator, and
* baseline-persistence behavioral + property tests.
*
* Module: gsd-core/bin/lib/complexity-trigger.cjs (compiled from
* src/complexity-trigger.cts — deliberately absent; this file is
* written failing-first per issue #1953's TDD directive).
*
* Spec sources (authoritative, do not drift from these without a design
* change):
* - .gsd/phase/feat-1953-complexity-triggered-refactor/41-api-contract.md
* - .gsd/phase/feat-1953-complexity-triggered-refactor/50-test-matrix.md
* - .gsd/phase/feat-1953-complexity-triggered-refactor/40-design.md
*
* Coverage map — test matrix rows 1-54 (analyzer counting, the literal-strip
* leak surface, CRLF/empty/size bounds, properties, file selection,
* threshold/jump evaluation, baseline persistence) plus the typed-surface
* enum/constant lock tests. Rows 55+ (git adapter, CLI, ship gate, loop
* wiring) live in tests/refactor-trigger-cli.test.cjs — out of scope here.
*
* Hermetic: every temp dir via createTempDir + t.after(cleanup); every fs
* mock via mock.method(fs, ...) restored via t.after(...mock.restore()).
* No shared module-level fixture state (matrix row 91). Ambient GSD_* env
* is not touched by this file — every fixture here is either pure in-memory
* input or an isolated tmpdir (matrix row 92 is structural, not exercised
* by this leaf-module suite).
*/
const { describe, test, mock } = require('node:test');
const assert = require('node:assert/strict');
const fs = require('node:fs');
const path = require('node:path');
const { createTempDir, cleanup } = require('./helpers.cjs');
const fc = require('./helpers/fast-check-setup.cjs');
const {
BASELINE_FILE_NAME,
PROPOSAL_SUFFIX,
SCHEMA_VERSION,
DEFAULTS,
ANALYZABLE_EXTENSIONS,
VERDICT,
REASON,
stripLiterals,
analyzeSource,
isAnalyzablePath,
evaluateCandidates,
nextBaseline,
reanchorBaseline,
readBaseline,
writeBaseline,
} = require('../gsd-core/bin/lib/complexity-trigger.cjs');
// ---------------------------------------------------------------------------
// Analyzer — counting (matrix rows 1-10)
// ---------------------------------------------------------------------------
describe('complexity-trigger: analyzer — counting decision points', () => {
test('scoresFlatFunctionAsOne', () => {
const source = ['function f() {', ' return 1;', '}'].join('\n');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.equal(r.functions.length, 1);
assert.equal(r.functions[0].name, 'f');
assert.equal(r.functions[0].startLine, 1);
assert.equal(r.functions[0].score, 1);
});
test('scoresSingleIfAsTwo', () => {
const source = [
'function f(x) {',
' if (x) {',
' return 1;',
' }',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.equal(r.functions[0].score, 2);
});
test('countsEachDecisionConstructOnce', () => {
// Each of the twelve listed constructs exactly once: if, else if, for,
// for..of, for..in, do, while (via one do-while so "while" is not
// double-counted by a separate stand-alone while statement), case,
// catch, &&, ||, ?:. Base 1 + 12 = 13.
const source = [
'function f(x) {',
' if (x === 1) { return 1; }',
' else if (x === 2) { return 2; }',
' for (let i = 0; i < 1; i++) { g(i); }',
' for (const y of x) { g(y); }',
' for (const k in x) { g(k); }',
' let n = 0;',
' do { n += 1; } while (n < 1);',
' switch (x) {',
' case 1: g(1); break;',
' default: g(0);',
' }',
' try { g(x); } catch (e) { g(e); }',
' const a = x && 1;',
' const b = x || 1;',
' const c = x ? 1 : 2;',
' return a + b + c;',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.equal(r.functions.length, 1);
assert.equal(r.functions[0].score, 13);
});
test('doesNotCountBareElse', () => {
const source = [
'function f(x) {',
' if (x) { return 1; }',
' else { return 2; }',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.functions[0].score, 2, 'bare else must not add a point beyond the if');
});
test('doesNotCountSwitchDefault', () => {
const source = [
'function f(x) {',
' switch (x) {',
' case 1: return 1;',
' default: return 0;',
' }',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.functions[0].score, 2, 'base 1 + one case; default: must not add a point');
});
test('doesNotConflateLengthWithComplexity', () => {
const bodyLines = Array.from({ length: 200 }, (_, i) => ` const v${i} = ${i};`);
const source = ['function f() {', ...bodyLines, ' return 0;', '}'].join('\n');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.equal(r.functions[0].score, 1, 'a long but flat function must still score 1');
});
test('scoresFlatSwitchByCaseCountPinningKnownBias', () => {
const caseLines = Array.from({ length: 12 }, (_, i) => ` case ${i + 1}: return ${i + 1};`);
const source = [
'function f(x) {',
' switch (x) {',
...caseLines,
' }',
' return 0;',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.functions[0].score, 13, 'flat 12-case switch scores 1 + 12 by design — the known bias is pinned, not accidental');
});
test('attributesScoresToEachFunction', () => {
const source = [
'function a() {',
' return 1;',
'}',
'',
'function b(x) {',
' if (x) {',
' return 1;',
' }',
' return 0;',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.functions.length, 2);
assert.equal(r.functions[0].name, 'a');
assert.equal(r.functions[0].startLine, 1);
assert.equal(r.functions[0].score, 1);
assert.equal(r.functions[1].name, 'b');
assert.equal(r.functions[1].startLine, 5);
assert.equal(r.functions[1].score, 2);
});
test('doesNotDoubleCountNestedFunction', () => {
const source = [
'function outer() {',
' function inner(x) {',
' if (x) {',
' return 1;',
' }',
' return 0;',
' }',
' return inner;',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.functions.length, 2);
const outer = r.functions.find((fn) => fn.name === 'outer');
const inner = r.functions.find((fn) => fn.name === 'inner');
assert.equal(outer.score, 1, 'the nested if must not double-count into outer');
assert.equal(inner.score, 2, 'the nested if is attributed to the innermost function');
});
test('detectsAllFunctionForms', () => {
const source = [
'const arrow = (x) => x + 1;',
'const obj = {',
' method(x) {',
' return x;',
' },',
'};',
'async function asyncFn() {',
' return 1;',
'}',
'function* genFn() {',
' yield 1;',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.equal(r.functions.length, 4, 'arrow, method shorthand, async function, and generator must all be detected');
assert.ok(r.functions.every((fn) => fn.score === 1), 'none of these forms has a branch of its own');
});
});
// ---------------------------------------------------------------------------
// Analyzer — the leak surface (matrix rows 11-21)
// ---------------------------------------------------------------------------
describe('complexity-trigger: analyzer — the leak surface', () => {
test('ignoresDecisionKeywordInLineComment', () => {
const source = [
'function f() {',
' // if (x) { return 1; }',
' return 0;',
'}',
].join('\n');
assert.equal(analyzeSource(source).functions[0].score, 1);
});
test('ignoresDecisionKeywordInBlockComment', () => {
const source = [
'function f() {',
' /* if (x) { return 1 } && y */',
' return 0;',
'}',
].join('\n');
assert.equal(analyzeSource(source).functions[0].score, 1);
});
test('ignoresDecisionKeywordInString', () => {
const source = [
'function f() {',
' const s = "if (x) { return 1; }";',
' return s;',
'}',
].join('\n');
assert.equal(analyzeSource(source).functions[0].score, 1);
});
test('ignoresDecisionKeywordInTemplateLiteral', () => {
const source = [
'function f() {',
' const s = `if (x) { return 1; }`;',
' return s;',
'}',
].join('\n');
assert.equal(analyzeSource(source).functions[0].score, 1);
});
test('ignoresAlternationInsideRegexLiteral', () => {
const source = [
'function f(s) {',
' return /a|b/.test(s);',
'}',
].join('\n');
assert.equal(analyzeSource(source).functions[0].score, 1, 'the | inside /a|b/ must not count as ||');
});
test('doesNotMistakeDivisionForRegexLiteral', () => {
const source = [
'function f(a, b, c) {',
' return a / b / c;',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.equal(r.functions[0].score, 1, 'a / b / c must be read as division, never as a regex literal');
});
test('doesNotCountOptionalChainingAsTernary', () => {
const source = [
'function f(a) {',
' return a?.b;',
'}',
].join('\n');
assert.equal(analyzeSource(source).functions[0].score, 1);
});
test('doesNotCountNullishCoalescingAsBranch', () => {
const source = [
'function f(x, y) {',
' return x ?? y;',
'}',
].join('\n');
assert.equal(analyzeSource(source).functions[0].score, 1);
});
test('handlesEscapedQuoteWithoutSwallowingCode', () => {
const source = [
'function f(x) {',
" const s = 'it\\'s fine';",
' if (x) { return s; }',
' return null;',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.equal(r.functions[0].score, 2, 'the escaped quote must not swallow the following if');
});
test('countsBranchInsideTemplateInterpolation', () => {
const source = [
'function f(x) {',
' const s = `value: ${x ? 1 : 2}`;',
' return s;',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.equal(r.functions[0].score, 2, 'the ternary inside ${} is real code and must be counted');
});
test('refusesToScoreUnterminatedLiteral', () => {
const unterminatedString = [
'function f() {',
' const s = "never closed;',
'}',
].join('\n');
const unterminatedTemplate = [
'function f() {',
' const s = `never closed;',
'}',
].join('\n');
const unterminatedComment = [
'function f() {',
' /* never closed',
' return 1;',
'}',
].join('\n');
for (const src of [unterminatedString, unterminatedTemplate, unterminatedComment]) {
const r = analyzeSource(src);
assert.equal(r.ok, false);
assert.equal(r.reason, REASON.REFACTOR_ANALYZER_UNPARSEABLE);
assert.equal(r.functions, undefined, 'an unparseable source must never emit a number');
}
});
test('producesIdenticalScoresForCrlfAndLf', () => {
const lfSource = [
'function f(x) {',
' if (x) {',
' return 1;',
' } else if (x === 2) {',
' return 2;',
' }',
' return 0;',
'}',
'',
'function g(y) {',
' return y && y || y ? 1 : 0;',
'}',
].join('\n');
const crlfSource = lfSource.replace(/\n/g, '\r\n');
const lfResult = analyzeSource(lfSource);
const crlfResult = analyzeSource(crlfSource);
assert.equal(lfResult.ok, true);
assert.equal(crlfResult.ok, true);
assert.deepEqual(
crlfResult.functions,
lfResult.functions,
'CRLF and LF forms of the same source must yield byte-identical scores AND line numbers',
);
});
test('returnsEmptyResultForEmptyFile', () => {
const r = analyzeSource('');
assert.equal(r.ok, true);
assert.deepEqual(r.functions, []);
});
test('returnsEmptyResultForWhitespaceOnlyFile', () => {
const r = analyzeSource(' \n\t\n \n');
assert.equal(r.ok, true);
assert.deepEqual(r.functions, []);
});
test('returnsEmptyResultWhenNoFunctionsPresent', () => {
const source = [
'const x = 1;',
'if (x) {',
' console.log(x);',
'}',
].join('\n');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.deepEqual(r.functions, [], 'top-level code belongs to no function, and must never be a synthetic entry');
});
test('analyzesLargeFileWithinBounds', () => {
const N = 20000;
const chunks = Array.from({ length: N }, (_, i) => [
`function f${i}(x) {`,
' if (x) { return 1; }',
' return 0;',
'}',
].join('\n'));
const source = chunks.join('\n\n');
assert.ok(source.length > 1_000_000, 'fixture must actually be large enough to exercise the bound');
const r = analyzeSource(source);
assert.equal(r.ok, true);
assert.equal(r.functions.length, N);
assert.ok(r.functions.every((fn) => fn.score === 2));
});
});
// ---------------------------------------------------------------------------
// Analyzer — properties (matrix rows 27-28)
// ---------------------------------------------------------------------------
describe('complexity-trigger: analyzer — properties', () => {
test('propertyStripNeverManufacturesDecisionPoints', () => {
// #1953 defect 3: the invariant is asserted through the TYPED surface
// (analyzeSource's numeric score), never by pattern-matching the text
// stripLiterals produced (CONTRIBUTING.md:800 bans grepping the SUT's
// own output, with no carve-out for a text transform's own return
// value). A fuzzed string embedded as INERT content (a block comment,
// or a string literal) can never manufacture a decision point — proven
// by the reported score staying byte-for-byte the same number.
const BASE = ['function f(x) {', ' if (x) { return 1; }', ' return 0;', '}', ''].join('\n');
const baseAnalyzed = analyzeSource(BASE);
assert.equal(baseAnalyzed.ok, true);
const baselineScores = baseAnalyzed.functions.map((fn) => fn.score);
fc.assert(
fc.property(fc.string({ maxLength: 300 }), (source) => {
const stripped = stripLiterals(source);
if (stripped.ok) {
assert.ok(
stripped.stripped.length <= source.length,
`stripLiterals must never grow the source (in=${source.length}, out=${stripped.stripped.length})`,
);
}
const analyzed = analyzeSource(source);
if (analyzed.ok) {
for (const fn of analyzed.functions) {
assert.ok(fn.score >= 1, 'every detected function must score at least 1');
}
}
// Sanitize just enough to keep the wrapper well-formed (never
// "*/" inside the comment; never an unescaped quote, backslash, or
// raw newline inside the string) — the content is otherwise
// untouched, including any decision-point-shaped substrings.
const commentSafe = source.replace(/\*/g, ' ');
const withComment = analyzeSource(`${BASE}\n/* ${commentSafe} */\n`);
assert.equal(withComment.ok, true, 'a sanitized block comment must never make the source unparseable');
assert.deepEqual(
withComment.functions.map((fn) => fn.score),
baselineScores,
'embedding fuzzed text inside a block comment must never change the reported score',
);
const stringSafe = source.replace(/["\\\r\n]/g, ' ');
const withString = analyzeSource(`${BASE}\nconst __s = "${stringSafe}";\n`);
assert.equal(withString.ok, true, 'a sanitized string literal must never make the source unparseable');
assert.deepEqual(
withString.functions.map((fn) => fn.score),
baselineScores,
'embedding fuzzed text inside a string literal must never change the reported score',
);
}),
{ numRuns: 50, seed: 42 },
);
});
test('propertyCommentsAndStringsAreScoreNeutral', () => {
const BASE_SOURCES = [
['function f(x) {', ' if (x) { return 1; }', ' return 0;', '}'].join('\n'),
['function g(x, y) {', ' return x && y;', '}'].join('\n'),
['function h(x) {', ' for (let i = 0; i < x; i++) { g(i); }', '}'].join('\n'),
];
const APPENDERS = [
(src) => [src, '// if (x) { return 1; } && y || z ? 1 : 2', ''].join('\n'),
(src) => [src, '/* if (x) { return 1 } && y */', ''].join('\n'),
(src) => [src, 'const __s = "if (x) { return 1; } && y || z ? 1 : 2";', ''].join('\n'),
];
fc.assert(
fc.property(
fc.constantFrom(...BASE_SOURCES),
fc.constantFrom(...APPENDERS),
(base, appendFn) => {
const before = analyzeSource(base);
const after = analyzeSource(appendFn(base));
assert.equal(before.ok, true);
assert.equal(after.ok, true);
assert.deepEqual(
before.functions.map((fn) => fn.score),
after.functions.map((fn) => fn.score),
'appending a comment/string containing decision keywords must never change a score',
);
},
),
{ numRuns: 20, seed: 42 },
);
});
});
// ---------------------------------------------------------------------------
// File selection (matrix rows 29-32)
// ---------------------------------------------------------------------------
describe('complexity-trigger: file selection', () => {
test('analyzesSupportedExtensions', () => {
for (const ext of ANALYZABLE_EXTENSIONS) {
assert.equal(isAnalyzablePath(`src/foo${ext}`), true, `expected ${ext} to be analyzable`);
}
});
test('skipsUnsupportedExtensionsWithoutScoringZero', () => {
for (const p of ['docs/readme.md', 'package.json', 'ci.yml', 'notes.txt', 'Makefile']) {
assert.equal(isAnalyzablePath(p), false, `expected ${p} to be unsupported`);
}
});
test('excludesTestFilesByDefault', () => {
assert.equal(isAnalyzablePath('tests/foo.test.cjs'), false);
assert.equal(isAnalyzablePath('tests\\foo.test.cjs'), false, 'separators are normalized unconditionally');
});
test('excludesGeneratedLibPathsExplicitly', () => {
assert.equal(isAnalyzablePath('gsd-core/bin/lib/foo.cjs'), false);
assert.equal(isAnalyzablePath('gsd-core\\bin\\lib\\foo.cjs'), false);
});
});
// ---------------------------------------------------------------------------
// Threshold / jump evaluation (matrix rows 33-44)
// ---------------------------------------------------------------------------
describe('complexity-trigger: threshold / jump evaluation', () => {
test('doesNotTriggerBelowThreshold', () => {
const T = DEFAULTS.threshold;
const analyzed = [{ file: 'a.js', ok: true, method: 'decision-points', functions: [{ name: 'f', startLine: 1, endLine: 3, score: T - 1 }] }];
const r = evaluateCandidates({ analyzed, baseline: {} });
assert.equal(r.verdict, VERDICT.BELOW_THRESHOLD);
assert.deepEqual(r.candidates, []);
assert.equal(r.target, null);
});
test('doesNotTriggerAtExactThreshold', () => {
const T = DEFAULTS.threshold;
const analyzed = [{ file: 'a.js', ok: true, method: 'decision-points', functions: [{ name: 'f', startLine: 1, endLine: 3, score: T }] }];
const r = evaluateCandidates({ analyzed, baseline: {} });
assert.equal(r.verdict, VERDICT.BELOW_THRESHOLD, 'strictly-greater semantics: score === T must not trigger');
assert.deepEqual(r.candidates, []);
});
test('triggersAboveThreshold', () => {
const T = DEFAULTS.threshold;
const analyzed = [{ file: 'a.js', ok: true, method: 'decision-points', functions: [{ name: 'f', startLine: 1, endLine: 3, score: T + 1 }] }];
const r = evaluateCandidates({ analyzed, baseline: {} });
assert.equal(r.verdict, VERDICT.TRIGGERED);
assert.equal(r.candidates.length, 1);
assert.deepEqual(r.candidates[0].reasons, ['threshold']);
assert.equal(r.candidates[0].baseline, null);
assert.equal(r.candidates[0].delta, null);
});
test('doesNotTriggerBelowJumpDelta', () => {
const T = DEFAULTS.threshold;
const D = DEFAULTS.jumpDelta;
const score = T - 5;
const baseline = { 'a.js::f': { score: score - (D - 1) } };
const analyzed = [{ file: 'a.js', ok: true, method: 'decision-points', functions: [{ name: 'f', startLine: 1, endLine: 3, score }] }];
const r = evaluateCandidates({ analyzed, baseline });
assert.equal(r.verdict, VERDICT.BELOW_THRESHOLD);
assert.deepEqual(r.candidates, []);
});
test('doesNotTriggerAtExactJumpDelta', () => {
const T = DEFAULTS.threshold;
const D = DEFAULTS.jumpDelta;
const score = T - 5;
const baseline = { 'a.js::f': { score: score - D } };
const analyzed = [{ file: 'a.js', ok: true, method: 'decision-points', functions: [{ name: 'f', startLine: 1, endLine: 3, score }] }];
const r = evaluateCandidates({ analyzed, baseline });
assert.equal(r.verdict, VERDICT.BELOW_THRESHOLD, 'strictly-greater semantics: delta === D must not trigger');
});
test('triggersAboveJumpDelta', () => {
const T = DEFAULTS.threshold;
const D = DEFAULTS.jumpDelta;
const score = T - 5;
const baseline = { 'a.js::f': { score: score - (D + 1) } };
const analyzed = [{ file: 'a.js', ok: true, method: 'decision-points', functions: [{ name: 'f', startLine: 1, endLine: 3, score }] }];
const r = evaluateCandidates({ analyzed, baseline });
assert.equal(r.verdict, VERDICT.TRIGGERED);
assert.deepEqual(r.candidates[0].reasons, ['jump']);
});
test('emitsSingleCandidateCarryingBothReasons', () => {
const T = DEFAULTS.threshold;
const D = DEFAULTS.jumpDelta;
const score = T + 1;
const baseline = { 'a.js::f': { score: score - (D + 1) } };
const analyzed = [{ file: 'a.js', ok: true, method: 'decision-points', functions: [{ name: 'f', startLine: 1, endLine: 3, score }] }];
const r = evaluateCandidates({ analyzed, baseline });
assert.equal(r.candidates.length, 1, 'one candidate, never two entries');
assert.deepEqual(r.candidates[0].reasons, ['threshold', 'jump'], 'both reasons, ordered');
});
test('doesNotTriggerWhenComplexityDecreased', () => {
const analyzed = [{ file: 'a.js', ok: true, method: 'decision-points', functions: [{ name: 'f', startLine: 1, endLine: 3, score: 10 }] }];
const baseline = { 'a.js::f': { score: 20 } };
const r = evaluateCandidates({ analyzed, baseline });
assert.equal(r.verdict, VERDICT.BELOW_THRESHOLD);
assert.deepEqual(r.candidates, []);
});
test('doesNotTreatMissingBaselineAsZeroBaseline', () => {
const T = DEFAULTS.threshold;
const analyzed = [{ file: 'a.js', ok: true, method: 'decision-points', functions: [{ name: 'f', startLine: 1, endLine: 3, score: T + 1 }] }];
const r = evaluateCandidates({ analyzed, baseline: {} });
assert.equal(r.candidates.length, 1);
assert.equal(r.candidates[0].baseline, null);
assert.equal(r.candidates[0].delta, null, 'no baseline means delta is null — never equal to score');
assert.notEqual(r.candidates[0].delta, r.candidates[0].score);
assert.deepEqual(r.candidates[0].reasons, ['threshold'], 'jump must not be evaluated with no baseline');
});
test('ordersCandidatesDeterministically', () => {
const analyzed = [
{ file: 'b.js', ok: true, method: 'decision-points', functions: [{ name: 'y', startLine: 1, endLine: 3, score: 20 }] },
{
file: 'a.js',
ok: true,
method: 'decision-points',
functions: [
{ name: 'x', startLine: 1, endLine: 3, score: 20 },
{ name: 'z', startLine: 10, endLine: 13, score: 18 },
],
},
];
const baseline = { 'a.js::x': { score: 10 } }; // y and z have no baseline entry
const r = evaluateCandidates({ analyzed, baseline, threshold: 15, jumpDelta: 100 });
const ids = r.candidates.map((c) => `${c.file}::${c.name}`);
assert.deepEqual(ids, ['a.js::x', 'b.js::y', 'a.js::z'], 'score desc, delta desc (null last), file asc, name asc');
assert.equal(r.target, r.candidates[0]);
});
test('breaksCandidateTiesStably', () => {
const analyzed = [
{
file: 'a.js',
ok: true,
method: 'decision-points',
functions: [
{ name: 'bravo', startLine: 20, endLine: 22, score: 20 },
{ name: 'alpha', startLine: 5, endLine: 7, score: 20 },
],
},
];
const r = evaluateCandidates({ analyzed, baseline: {}, threshold: 15, jumpDelta: 100 });
assert.deepEqual(r.candidates.map((c) => c.name), ['alpha', 'bravo'], 'tie on every other key breaks by name ascending');
});
test('rejectsNonNumericThresholdInsteadOfNaNComparing', () => {
const scoreAboveDefaultThreshold = DEFAULTS.threshold + 1;
const analyzed = [{
file: 'a.js',
ok: true,
method: 'decision-points',
functions: [{ name: 'f', startLine: 1, endLine: 3, score: scoreAboveDefaultThreshold }],
}];
for (const bad of [0, -5, 'not-a-number', NaN, undefined, null, {}]) {
const r = evaluateCandidates({ analyzed, baseline: {}, threshold: bad, jumpDelta: bad });
assert.equal(r.thresholdUsed, DEFAULTS.threshold, `threshold ${String(bad)} must fall back to the default`);
assert.equal(r.jumpDeltaUsed, DEFAULTS.jumpDelta, `jumpDelta ${String(bad)} must fall back to the default`);
assert.equal(r.verdict, VERDICT.TRIGGERED, 'fallback must still evaluate — never a silent NaN-comparison never-trigger');
}
});
});
// ---------------------------------------------------------------------------
// Baseline persistence (matrix rows 45-54)
// ---------------------------------------------------------------------------
describe('complexity-trigger: baseline persistence', () => {
test('treatsMissingBaselineAsEmpty', (t) => {
const tmp = createTempDir('complexity-trigger-');
t.after(() => cleanup(tmp));
const r = readBaseline(tmp);
assert.equal(r.ok, true);
assert.deepEqual(r.baseline, {});
assert.equal(r.reason, undefined);
});
test('doesNotClobberMalformedBaselineOnReadFailure', (t) => {
const tmp = createTempDir('complexity-trigger-');
t.after(() => cleanup(tmp));
const baselinePath = path.join(tmp, BASELINE_FILE_NAME);
fs.writeFileSync(baselinePath, '{not valid json', 'utf8');
const before = fs.readFileSync(baselinePath, 'utf8');
const r = readBaseline(tmp);
assert.equal(r.ok, false);
assert.equal(r.reason, REASON.REFACTOR_BASELINE_MALFORMED);
assert.deepEqual(r.baseline, {});
const after = fs.readFileSync(baselinePath, 'utf8');
assert.equal(after, before, 'a failing read must never rewrite the malformed file');
});
test('rejectsWrongShapedBaseline', (t) => {
const tmp = createTempDir('complexity-trigger-');
t.after(() => cleanup(tmp));
const baselinePath = path.join(tmp, BASELINE_FILE_NAME);
for (const bad of ['[]', '"a string"', 'null', '42']) {
fs.writeFileSync(baselinePath, bad, 'utf8');
const r = readBaseline(tmp);
assert.equal(r.ok, false, `shape ${bad} must be rejected`);
assert.equal(r.reason, REASON.REFACTOR_BASELINE_MALFORMED);
}
});
test('anchorsBaselineOnFirstObservation', () => {
const prev = {};
const analyzed = [{
file: 'a.js', ok: true, method: 'decision-points',
functions: [{ name: 'f', startLine: 1, endLine: 3, score: 6 }],
}];
const next = nextBaseline(prev, analyzed, { analyzedFiles: ['a.js'] });
assert.equal(next['a.js::f'].score, 6, 'first observation inserts the anchor at the current score');
});
test('doesNotAdvanceAnchorOnPlainEvaluate', () => {
const prev = { 'a.js::f': { score: 3 } };
const analyzed = [{
file: 'a.js', ok: true, method: 'decision-points',
functions: [{ name: 'f', startLine: 1, endLine: 3, score: 6 }],
}];
const next = nextBaseline(prev, analyzed, { analyzedFiles: ['a.js'] });
assert.equal(next['a.js::f'].score, 3, 'a stable anchor must not advance on a plain evaluate, even for a non-triggering function');
});
test('freezesBaselineWhileProposalUntriaged', () => {
// The anchor is unconditionally stable across evaluate — not merely
// "frozen because it triggered". This proves it holds even when the
// function is (still) triggering on a second evaluate.
const prev = { 'a.js::f': { score: 10 } };
const analyzed = [{
file: 'a.js', ok: true, method: 'decision-points',
functions: [{ name: 'f', startLine: 1, endLine: 5, score: DEFAULTS.threshold + 5 }],
}];
const evaluation = evaluateCandidates({ analyzed, baseline: prev });
assert.equal(evaluation.verdict, VERDICT.TRIGGERED);
const next = nextBaseline(prev, analyzed, { analyzedFiles: ['a.js'] });
assert.equal(next['a.js::f'], prev['a.js::f'], 'the anchor carries forward unchanged, not merely equal');
assert.equal(next['a.js::f'].score, 10);
});
test('reanchorsBaselineOnDisposition', () => {
// Proves reanchorBaseline is the ONLY thing that moves an anchor:
// nextBaseline leaves it untouched regardless of candidates; only an
// explicit disposition call moves it.
const key = 'src/a.cts::target';
const prev = { [key]: { score: 3 } };
const analyzed = [{
file: 'src/a.cts', ok: true, method: 'decision-points',
functions: [{ name: 'target', startLine: 1, endLine: 20, score: 11 }],
}];
const evaluation = evaluateCandidates({ analyzed, baseline: prev });
assert.equal(evaluation.verdict, VERDICT.TRIGGERED, 'setup sanity: the function does trigger, proving the freeze below is not merely "nothing to move"');
const afterEvaluate = nextBaseline(prev, analyzed, { analyzedFiles: ['src/a.cts'] });
assert.equal(afterEvaluate[key].score, 3, 'a plain evaluate must never move the anchor');
// Accept: re-anchor to a LOWER post-refactor score.
const afterAccept = reanchorBaseline(prev, key, 4);
assert.equal(afterAccept[key].score, 4, 'accept re-anchors to the post-refactor score');
// Decline: re-anchor to the current HIGHER score (debt consciously accepted).
const afterDecline = reanchorBaseline(prev, key, 11);
assert.equal(afterDecline[key].score, 11, 'decline re-anchors to the current score');
});
test('detectsIncrementalCreepAcrossRuns', () => {
const T = 15;
const D = 5;
const anchor = { 'a.js::f': { score: 8 } };
let baseline = anchor;
let firstTriggeringRound = null;
let firstTriggeringCandidate = null;
for (let round = 1; round <= 5; round += 1) {
const score = 6 + round * 2; // round1=8, round2=10, round3=12, round4=14, round5=16
const analyzed = [{
file: 'a.js', ok: true, method: 'decision-points',
functions: [{ name: 'f', startLine: 1, endLine: 5, score }],
}];
const evaluation = evaluateCandidates({ analyzed, baseline, threshold: T, jumpDelta: D });
if (firstTriggeringRound === null && evaluation.verdict === VERDICT.TRIGGERED) {
firstTriggeringRound = round;
firstTriggeringCandidate = evaluation.candidates[0];
}
baseline = nextBaseline(baseline, analyzed, { analyzedFiles: ['a.js'] });
}
// Anchor is stable at 8 throughout (a plain evaluate never advances it),
// so the delta is CUMULATIVE since the anchor: round1=0, round2=2,
// round3=4, round4=6, round5=8. Delta first exceeds D=5 at round 4
// (6 > 5) — one full run before the absolute score would cross T=15
// (score 16 at round 5). This is the point of the test: the jump-delta
// catches the creep earlier than, and independently of, the absolute
// threshold — it is not redundant with it.
assert.equal(firstTriggeringRound, 4);
assert.ok(
firstTriggeringCandidate.reasons.includes('jump'),
'round 4 must trigger via the jump reason, proving the jump-delta adds value over the absolute threshold',
);
});
test('prunesBaselineForDeletedFile', () => {
const analyzed = [{
file: 'a.js', ok: true, method: 'decision-points',
functions: [{ name: 'f', startLine: 1, endLine: 3, score: 5 }],
}];
const prev = { 'a.js::f': { score: 5 }, 'a.js::g': { score: 8 } };
const next = nextBaseline(prev, analyzed, { analyzedFiles: ['a.js'] });
assert.ok(!('a.js::g' in next), 'g no longer exists in the analyzed file; its baseline entry must be pruned');
assert.equal(next['a.js::f'].score, 5);
const prevWithUntouchedFile = { 'a.js::f': { score: 5 }, 'b.js::h': { score: 99 } };
const nextWithUntouchedFile = nextBaseline(prevWithUntouchedFile, analyzed, { analyzedFiles: ['a.js'] });
assert.equal(
nextWithUntouchedFile['b.js::h'].score,
99,
'an entry whose file is absent from analyzedFiles must never be pruned',
);
});
test('leavesNoPartialBaselineWhenWriteFails', (t) => {
const tmp = createTempDir('complexity-trigger-');
t.after(() => cleanup(tmp));
const writeMock = mock.method(fs, 'writeFileSync', () => {
const err = new Error('ENOSPC: no space left on device');
err.code = 'ENOSPC';
throw err;
});
t.after(() => writeMock.mock.restore());
const r = writeBaseline(tmp, { 'a.js::f': { score: 5 } });
assert.equal(r.ok, false);
assert.equal(r.reason, REASON.REFACTOR_BASELINE_WRITE_FAILED);
const leftover = fs.readdirSync(tmp).filter((n) => n.includes('.tmp'));
assert.deepEqual(leftover, [], 'no orphan tmp file must remain after write failure');
});
test('cleansUpTempFileWhenRenameFails', (t) => {
const tmp = createTempDir('complexity-trigger-');
t.after(() => cleanup(tmp));
const renameMock = mock.method(fs, 'renameSync', () => {
const err = new Error('EPERM: operation not permitted');
err.code = 'EPERM';
throw err;
});
t.after(() => renameMock.mock.restore());
const r = writeBaseline(tmp, { 'a.js::f': { score: 5 } });
assert.equal(r.ok, false);
assert.equal(r.reason, REASON.REFACTOR_BASELINE_WRITE_FAILED);
const leftover = fs.readdirSync(tmp).filter((n) => n.includes('.tmp'));
assert.deepEqual(leftover, [], 'the orphan .tmp file must be unlinked when rename fails');
});
test('degradesReadBaselineOnDeniedRead', (t) => {
// An injected EACCES on the baseline read must degrade readBaseline to
// { ok:false, baseline:{}, reason: REFACTOR_BASELINE_MALFORMED } without
// throwing.
const tmp = createTempDir('complexity-trigger-');
t.after(() => cleanup(tmp));
const readMock = mock.method(fs, 'readFileSync', () => {
const err = new Error('EACCES: permission denied');
err.code = 'EACCES';
throw err;
});
t.after(() => readMock.mock.restore());
const r = readBaseline(tmp);
assert.equal(r.ok, false);
assert.deepEqual(r.baseline, {});
assert.equal(r.reason, REASON.REFACTOR_BASELINE_MALFORMED);
});
});
// ---------------------------------------------------------------------------
// Typed surface — frozen enums and constants
// ---------------------------------------------------------------------------
describe('complexity-trigger: typed surface', () => {
test('locksReasonEnumKeys', () => {
const expected = [
'REFACTOR_ALREADY_DISPOSITIONED',
'REFACTOR_ANALYZER_UNPARSEABLE',
'REFACTOR_ANALYZER_UNSUPPORTED',
'REFACTOR_ARTIFACT_NOT_FOUND',
'REFACTOR_BASELINE_MALFORMED',
'REFACTOR_BASELINE_WRITE_FAILED',
'REFACTOR_DECLINE_REASON_EMPTY',
'REFACTOR_DISABLED',
'REFACTOR_FILE_UNREADABLE',
'REFACTOR_GIT_UNAVAILABLE',
'REFACTOR_INVALID_PHASE',
'REFACTOR_NO_TOUCHED_FILES',
'REFACTOR_OK',
'REFACTOR_STRICT_NOT_ENFORCING',
'REFACTOR_USAGE',
].sort();
assert.deepEqual(Object.keys(REASON).sort(), expected);
assert.ok(Object.isFrozen(REASON));
});
test('locksVerdictEnumKeys', () => {
const expected = ['BELOW_THRESHOLD', 'SKIPPED', 'TRIGGERED'].sort();
assert.deepEqual(Object.keys(VERDICT).sort(), expected);
assert.ok(Object.isFrozen(VERDICT));
});
test('exposes frozen DEFAULTS and ANALYZABLE_EXTENSIONS with documented values', () => {
assert.deepEqual(DEFAULTS, { threshold: 15, jumpDelta: 5 });
assert.ok(Object.isFrozen(DEFAULTS));
assert.deepEqual(ANALYZABLE_EXTENSIONS, ['.js', '.cjs', '.mjs', '.ts', '.cts', '.mts']);
assert.ok(Object.isFrozen(ANALYZABLE_EXTENSIONS));
});
test('exposes the documented module-scoped constants', () => {
assert.equal(BASELINE_FILE_NAME, 'complexity-baseline.json');
assert.equal(PROPOSAL_SUFFIX, '-REFACTOR.md');
assert.equal(SCHEMA_VERSION, 1);
});
});
describe('complexity-trigger: proposal fence-width tolerant read (#3657)', () => {
test('parseProposal accepts a formatter-narrowed 3-backtick fence (#3657)', () => {
const { renderProposal, parseProposal } = require('../gsd-core/bin/lib/complexity-trigger.cjs');
const proposal = {
schema_version: 1,
status: 'proposed',
phase: '2',
target_file: 'src/a.ts',
target_function: 'handleThing',
score: 7,
baseline: 5,
delta: 2,
metric: 'decision-points',
recorded_at: '2026-07-19T00:00:00Z',
resolved_at: null,
reason: 'score above threshold',
candidates: [{ name: 'handleThing', score: 7 }],
};
const rendered = renderProposal(proposal);
const parsed4 = parseProposal(rendered);
assert.notEqual(parsed4, null, 'writer form must round-trip');
// Same artifact after a CommonMark formatter narrows the fence to the
// shortest legal width — the identical #3657 defect class.
const narrowed = rendered
.replace(/^````json$/m, '```json')
.replace(/^````$/m, '```');
const parsed3 = parseProposal(narrowed);
assert.notEqual(parsed3, null, 'a formatter-narrowed proposal must parse');
assert.equal(parsed3.status, 'proposed');
assert.equal(parsed3.candidates.length, 1);
assert.equal(parsed3.candidates[0].name, 'handleThing');
});
});