ms-ai-architect/tests/kb-update/test-dedup-dato.test.mjs

// tests/kb-update/test-dedup-dato.test.mjs
// TDD for Enhet 3b — the dual-**Dato:** dedup on the 5 mlops-genaiops files that carry BOTH a
// bold **Dato:** (creation) AND a bold **Last updated:** (modification). Ground truth 2026-07-16
// (re-measured against the committed corpus + the git first-commit "Created-map"):
//   4 files carry a DISTINCT creation date (**Dato:** 2026-02-04, authored BEFORE the git
//     first-commit 2026-04-08 → a genuine creation date) → relabel **Dato:** → **Created:**,
//     value byte-preserved (the same norsk→english label-rename class as Enhet 1/2).
//   1 file (mlops-security-access-control) carries IDENTICAL dates (**Dato:**=**Last updated:**=
//     2026-06-19, both AFTER the git first-commit → NOT a genuine creation date, a metadata-pass
//     artifact) → drop the redundant **Dato:**; relabeling it would assert a FALSE creation date.
//
// This is never a derive-a-value backfill and never an unproven delete: the drop primitive throws
// unless a bold proof field with the IDENTICAL value exists. One new pure primitive
// (dropRedundantBoldField) + the reused relabelHeaderDialect (CREATED_MAP) are pinned here, plus
// a corpus-state guard (the gap-discipline lock) over the 5 committed files.

import { test } from 'node:test';
import assert from 'node:assert/strict';
import { readFileSync } from 'node:fs';
import { join, dirname } from 'node:path';
import { fileURLToPath } from 'node:url';
import { dropRedundantBoldField } from '../../scripts/kb-update/lib/transform.mjs';
import { relabelHeaderDialect } from '../../scripts/kb-update/relabel-dialect.mjs';
import { MANIFEST, CREATED_MAP, applyOp } from '../../scripts/kb-update/dedup-dato.mjs';

const __dirname = dirname(fileURLToPath(import.meta.url));
const PLUGIN_ROOT = join(__dirname, '..', '..');
const DIR = 'skills/ms-ai-engineering/references/mlops-genaiops';

// A dual-**Dato:** fixture: bold Category, bold **Dato:** (creation) + bold **Last updated:**
// (modification) with DISTINCT values, then a body that mentions "Dato:" in prose (never touched).
const FIXTURE = [
  '# Feedback Loops and Continuous Improvement',
  '',
  '**Category:** MLOps & GenAIOps',
  '**Dato:** 2026-02-04',
  '**Last updated:** 2026-06-24',
  '**Status:** Established Practice',
  '',
  '## Innhold',
  '',
  'A body line mentioning **Dato:** [YYYY-MM-DD] in a template — must be left alone.',
  '',
].join('\n');

// mlops-security shape: IDENTICAL **Dato:** and **Last updated:** → the drop case.
const FIXTURE_IDENTICAL = [
  '# Security and Access Control in MLOps',
  '',
  '**Category:** MLOps & GenAIOps',
  '**Last updated:** 2026-06-19',
  '**Dato:** 2026-06-19',
  '**Status:** Established Practice',
  '',
  '---',
  '',
].join('\n');

// ── dropRedundantBoldField ───────────────────────────────────────────────────
test('dropRedundantBoldField: drops **Dato:** when a bold **Last updated:** with the same value proves it redundant', () => {
  const out = dropRedundantBoldField(FIXTURE_IDENTICAL, 'Dato', 'Last updated');
  const header = out.split('\n---')[0];
  assert.ok(/\*\*Last updated:\*\* 2026-06-19/.test(header), 'proof field kept');
  assert.ok(!/\*\*Dato:\*\*/.test(header), 'redundant **Dato:** removed');
  assert.equal(out.split('\n').length, FIXTURE_IDENTICAL.split('\n').length - 1, 'exactly one line removed');
});

test('dropRedundantBoldField: throws when the drop and proof values differ (not redundant, human decision)', () => {
  assert.throws(() => dropRedundantBoldField(FIXTURE, 'Dato', 'Last updated'), /not redundant|≠/);
});

test('dropRedundantBoldField: throws when no proof field exists to prove redundancy', () => {
  const noProof = FIXTURE_IDENTICAL.replace('**Last updated:** 2026-06-19\n', '');
  assert.throws(() => dropRedundantBoldField(noProof, 'Dato', 'Last updated'), /no bold/i);
});

test('dropRedundantBoldField: no-op when the drop field is absent (idempotent)', () => {
  const once = dropRedundantBoldField(FIXTURE_IDENTICAL, 'Dato', 'Last updated');
  assert.equal(dropRedundantBoldField(once, 'Dato', 'Last updated'), once);
});

test('dropRedundantBoldField: never touches a **Dato:** in the body (below the header region)', () => {
  // FIXTURE has a distinct-value dual header (would throw) so build an identical-value variant
  // that ALSO carries a body **Dato:** template line — only the header one may be removed.
  const withBody = FIXTURE_IDENTICAL.replace('---', '## Innhold\n\n**Dato:** [YYYY-MM-DD] template\n\n---');
  const out = dropRedundantBoldField(withBody, 'Dato', 'Last updated');
  assert.ok(out.includes('**Dato:** [YYYY-MM-DD] template'), 'body template **Dato:** untouched');
  assert.ok(!/\*\*Dato:\*\*/.test(out.split('## Innhold')[0]), 'header **Dato:** removed');
});

test('dropRedundantBoldField: blank labels throw', () => {
  assert.throws(() => dropRedundantBoldField(FIXTURE_IDENTICAL, '', 'Last updated'), /required/);
  assert.throws(() => dropRedundantBoldField(FIXTURE_IDENTICAL, 'Dato', ''), /required/);
});

// ── relabel **Dato:** → **Created:** (reused relabelHeaderDialect + CREATED_MAP) ──────────────
test('CREATED_MAP: relabels **Dato:** → **Created:**, value byte-preserved, one line, body identical', () => {
  const { content: out, applied } = relabelHeaderDialect(FIXTURE, CREATED_MAP);
  assert.equal(applied.length, 1);
  assert.ok(/\*\*Created:\*\* 2026-02-04/.test(out), 'Created line present with preserved value');
  assert.ok(!/\*\*Dato:\*\*/.test(out.split('## Innhold')[0]), 'header **Dato:** gone');
  const bodyOf = (c) => c.slice(c.indexOf('## Innhold'));
  assert.equal(bodyOf(out), bodyOf(FIXTURE), 'body byte-identical (template **Dato:** untouched)');
  assert.equal(out.split('\n').length, FIXTURE.split('\n').length, 'line count unchanged');
});

test('CREATED_MAP: idempotent — a file with no **Dato:** in header is unchanged', () => {
  const { content: once } = relabelHeaderDialect(FIXTURE, CREATED_MAP);
  const { content: twice, applied } = relabelHeaderDialect(once, CREATED_MAP);
  assert.equal(twice, once);
  assert.equal(applied.length, 0);
});

// ── applyOp dispatch ─────────────────────────────────────────────────────────
test('applyOp: relabel dispatches to Created relabel', () => {
  assert.ok(/\*\*Created:\*\* 2026-02-04/.test(applyOp(FIXTURE, 'relabel')));
});

test('applyOp: drop dispatches to redundant-bold drop', () => {
  const out = applyOp(FIXTURE_IDENTICAL, 'drop');
  assert.ok(!/\*\*Dato:\*\*/.test(out.split('\n---')[0]));
});

test('applyOp: unknown action throws', () => {
  assert.throws(() => applyOp(FIXTURE, 'nonsense'), /unknown action/i);
});

// ── MANIFEST ─────────────────────────────────────────────────────────────────
test('MANIFEST: exactly the 5 dual-**Dato:** mlops-genaiops files, 4 relabel + 1 drop', () => {
  assert.equal(MANIFEST.length, 5);
  assert.equal(new Set(MANIFEST.map((m) => m.rel)).size, 5, 'no duplicate paths');
  const relabels = MANIFEST.filter((m) => m.action === 'relabel');
  const drops = MANIFEST.filter((m) => m.action === 'drop');
  assert.equal(relabels.length, 4);
  assert.equal(drops.length, 1);
  assert.equal(drops[0].rel, `${DIR}/mlops-security-access-control.md`, 'the identical-date file is the drop');
  for (const r of relabels) {
    assert.match(r.created, /^\d{4}-\d{2}-\d{2}$/, `${r.rel}: frozen created date must be YYYY-MM-DD`);
  }
  const expected = [
    `${DIR}/feedback-loops-continuous-improvement.md`,
    `${DIR}/genaiops-llm-specific-practices.md`,
    `${DIR}/model-deployment-strategies-azure.md`,
    `${DIR}/prompt-flow-production-deployment.md`,
    `${DIR}/mlops-security-access-control.md`,
  ];
  for (const p of expected) assert.ok(MANIFEST.some((m) => m.rel === p), `missing ${p}`);
});

// ── Corpus-state guard (gap-discipline lock over the committed files) ─────────
test('corpus guard: the 4 distinct files carry **Created:** with the frozen date and no **Dato:** in the header', () => {
  for (const m of MANIFEST.filter((x) => x.action === 'relabel')) {
    const content = readFileSync(join(PLUGIN_ROOT, m.rel), 'utf8');
    const header = content.split(/\n(?:---|## )/)[0];
    assert.match(header, new RegExp(`\\*\\*Created:\\*\\* ${m.created}`), `${m.rel}: missing **Created:** ${m.created}`);
    assert.ok(!/\*\*Dato:\*\*/.test(header), `${m.rel}: header still carries **Dato:**`);
    assert.match(header, /\*\*Last updated:\*\*/, `${m.rel}: **Last updated:** must survive`);
  }
});

test('corpus guard: mlops-security-access-control has no **Dato:** and keeps **Last updated:** 2026-06-19', () => {
  const rel = `${DIR}/mlops-security-access-control.md`;
  const content = readFileSync(join(PLUGIN_ROOT, rel), 'utf8');
  const header = content.split(/\n(?:---|## )/)[0];
  assert.ok(!/\*\*Dato:\*\*/.test(header), 'redundant **Dato:** must be gone');
  assert.ok(!/\*\*Created:\*\*/.test(header), 'no fabricated **Created:** (drop, not relabel)');
  assert.match(header, /\*\*Last updated:\*\* 2026-06-19/, '**Last updated:** 2026-06-19 must survive');
});