sillyspec 3.20.2 → 3.20.4

package/test/scan-knowledge.test.mjs

@@ -1,175 +1,175 @@
-/**
- * scan-knowledge.test.mjs — knowledge 产物校验测试
- */
-
-import { join, resolve, dirname, basename } from 'path'
-import { existsSync, mkdirSync, writeFileSync, rmSync } from 'fs'
-import { fileURLToPath, pathToFileURL } from 'url'
-import { tmpdir } from 'os'
-
-const __filename = fileURLToPath(import.meta.url)
-const __dirname = dirname(__filename)
-const root = resolve(__dirname, '..')
-
-const { runScanPostCheck } = await import(pathToFileURL(join(root, 'src', 'scan-postcheck.js')).href)
-
-let passed = 0, failed = 0
-
-function assert(cond, msg) {
-  if (cond) { console.log(`  ✅ PASS: ${msg}`); passed++ }
-  else { console.log(`  ❌ FAIL: ${msg}`); failed++ }
-}
-
-function setup(name) {
-  const cwd = join(tmpdir(), `kn-${name}`)
-  mkdirSync(cwd, { recursive: true })
-  return cwd
-}
-function specSetup(name) {
-  const d = join(tmpdir(), `kn-${name}-spec`)
-  mkdirSync(d, { recursive: true })
-  return d
-}
-function clean(...dirs) { for (const d of dirs) try { rmSync(d, { recursive: true, force: true }) } catch {} }
-
-const DOCS = ['ARCHITECTURE.md','CONVENTIONS.md','STRUCTURE.md','INTEGRATIONS.md','TESTING.md','CONCERNS.md','PROJECT.md']
-
-function writeFull(cwd, specDir) {
-  const proj = basename(cwd)
-  for (const d of DOCS) {
-    const p = join(specDir, 'docs', proj, 'scan', d)
-    mkdirSync(dirname(p), { recursive: true })
-    writeFileSync(p, 'author: bot\ncreated_at: 2026-06-19 10:00:00\n# doc\n')
-  }
-}
-
-// ── 1: knowledge INDEX.md 存在，引用真实文件 → 通过 ──
-console.log('\n=== Test 1: knowledge INDEX.md 存在且引用有效 → no knowledge warning ===')
-{
-  const cwd = setup('t1'), spec = specSetup('t1')
-  try {
-    writeFull(cwd, spec)
-    const knowledgeDir = join(spec, 'knowledge')
-    mkdirSync(knowledgeDir, { recursive: true })
-    writeFileSync(join(knowledgeDir, 'INDEX.md'),
-      '# Knowledge Index\n\n## Conventions\n- [kebab-case naming](conventions.md#kebab-case-naming)\n')
-    writeFileSync(join(knowledgeDir, 'conventions.md'),
-      '# Conventions\n\n## kebab-case-naming\nFiles use kebab-case.\n')
-
-    const result = runScanPostCheck({ cwd, specDir: spec })
-    const knowledgeChecks = result.checks.filter(c => c.name.startsWith('knowledge_'))
-    assert(knowledgeChecks.length === 0, 'no knowledge warnings when INDEX.md + referenced files exist')
-  } finally { clean(cwd, spec) }
-}
-
-// ── 2: knowledge 目录不存在 → dir_missing 警告 ──
-console.log('\n=== Test 2: knowledge 目录不存在 → knowledge_dir_missing ===')
-{
-  const cwd = setup('t2'), spec = specSetup('t2')
-  try {
-    writeFull(cwd, spec)
-
-    const result = runScanPostCheck({ cwd, specDir: spec })
-    const missing = result.checks.find(c => c.name === 'knowledge_dir_missing')
-    assert(!!missing, 'knowledge_dir_missing check exists')
-    assert(missing.detail.includes('knowledge/ 目录不存在'), 'detail text correct')
-  } finally { clean(cwd, spec) }
-}
-
-// ── 3: INDEX.md 不存在 → index_missing 警告 ──
-console.log('\n=== Test 3: INDEX.md 不存在 → knowledge_index_missing ===')
-{
-  const cwd = setup('t3'), spec = specSetup('t3')
-  try {
-    writeFull(cwd, spec)
-    const knowledgeDir = join(spec, 'knowledge')
-    mkdirSync(knowledgeDir, { recursive: true })
-    // INDEX.md not created
-
-    const result = runScanPostCheck({ cwd, specDir: spec })
-    const missing = result.checks.find(c => c.name === 'knowledge_index_missing')
-    assert(!!missing, 'knowledge_index_missing check exists')
-    assert(missing.detail.includes('INDEX.md 不存在'), 'detail text correct')
-  } finally { clean(cwd, spec) }
-}
-
-// ── 4: INDEX.md 引用不存在文件 → broken_refs 警告 ──
-console.log('\n=== Test 4: INDEX.md 引用不存在的文件 → knowledge_broken_refs ===')
-{
-  const cwd = setup('t4'), spec = specSetup('t4')
-  try {
-    writeFull(cwd, spec)
-    const knowledgeDir = join(spec, 'knowledge')
-    mkdirSync(knowledgeDir, { recursive: true })
-    writeFileSync(join(knowledgeDir, 'INDEX.md'),
-      '# Knowledge Index\n\n## Conventions\n- [naming](conventions.md#naming)\n\n## Patterns\n- [auth pattern](patterns.md#auth-pattern)\n')
-    // conventions.md exists, patterns.md does not
-    writeFileSync(join(knowledgeDir, 'conventions.md'), '# Conventions\n')
-
-    const result = runScanPostCheck({ cwd, specDir: spec })
-    const broken = result.checks.find(c => c.name === 'knowledge_broken_refs')
-    assert(!!broken, 'knowledge_broken_refs check exists')
-    assert(broken.detail.includes('patterns.md'), 'reports missing patterns.md')
-  } finally { clean(cwd, spec) }
-}
-
-// ── 5: uncategorized.md 为空但存在 → 通过 ──
-console.log('\n=== Test 5: uncategorized.md 空但存在 → no knowledge warning ===')
-{
-  const cwd = setup('t5'), spec = specSetup('t5')
-  try {
-    writeFull(cwd, spec)
-    const knowledgeDir = join(spec, 'knowledge')
-    mkdirSync(knowledgeDir, { recursive: true })
-    writeFileSync(join(knowledgeDir, 'INDEX.md'),
-      '# Knowledge Index\n\n(no entries yet)\n')
-    writeFileSync(join(knowledgeDir, 'uncategorized.md'), '# Uncategorized\n')
-
-    const result = runScanPostCheck({ cwd, specDir: spec })
-    const knowledgeChecks = result.checks.filter(c => c.name.startsWith('knowledge_'))
-    assert(knowledgeChecks.length === 0, 'empty uncategorized.md passes')
-  } finally { clean(cwd, spec) }
-}
-
-// ── 6: categorized 文件存在但 INDEX.md 无条目 → 通过（不强求 categorized 非空）──
-console.log('\n=== Test 6: categorized 文件存在 + INDEX.md 有效 → no knowledge warning ===')
-{
-  const cwd = setup('t6'), spec = specSetup('t6')
-  try {
-    writeFull(cwd, spec)
-    const knowledgeDir = join(spec, 'knowledge')
-    mkdirSync(knowledgeDir, { recursive: true })
-    writeFileSync(join(knowledgeDir, 'INDEX.md'),
-      '# Knowledge Index\n\n## Known Issues\n- [GLM proxy](known-issues.md#glm-proxy)\n')
-    writeFileSync(join(knowledgeDir, 'known-issues.md'),
-      '# Known Issues\n\n## glm-proxy\nGLM proxy has usage metadata limits.\n')
-    writeFileSync(join(knowledgeDir, 'uncategorized.md'), '# Uncategorized\n')
-
-    const result = runScanPostCheck({ cwd, specDir: spec })
-    const knowledgeChecks = result.checks.filter(c => c.name.startsWith('knowledge_'))
-    assert(knowledgeChecks.length === 0, 'at least one categorized file + valid INDEX passes')
-  } finally { clean(cwd, spec) }
-}
-
-// ── 7: source_root knowledge 泄漏检测（平台模式已有检查）──
-console.log('\n=== Test 7: source_root knowledge 泄漏 → source_root_leak ===')
-{
-  const cwd = setup('t7'), spec = specSetup('t7')
-  try {
-    writeFull(cwd, spec)
-    // 写入 source_root/.sillyspec/knowledge/ 泄漏
-    const leakedKnowledgeDir = join(cwd, '.sillyspec', 'knowledge')
-    mkdirSync(leakedKnowledgeDir, { recursive: true })
-    writeFileSync(join(leakedKnowledgeDir, 'INDEX.md'), '# leaked\n')
-
-    const result = runScanPostCheck({ cwd, specDir: spec })
-    const leak = result.checks.find(c => c.name === 'source_root_leak')
-    assert(!!leak, 'source_root knowledge leak detected')
-    assert(leak.severity === 'failed', 'leak severity is failed')
-  } finally { clean(cwd, spec) }
-}
-
-// ── 汇总 ──
-console.log(`\n${'='.repeat(40)}`)
-console.log(`scan-knowledge tests: ${passed} passed, ${failed} failed, ${passed + failed} total`)
-if (failed > 0) process.exit(1)
+/**
+ * scan-knowledge.test.mjs — knowledge 产物校验测试
+ */
+
+import { join, resolve, dirname, basename } from 'path'
+import { existsSync, mkdirSync, writeFileSync, rmSync } from 'fs'
+import { fileURLToPath, pathToFileURL } from 'url'
+import { tmpdir } from 'os'
+
+const __filename = fileURLToPath(import.meta.url)
+const __dirname = dirname(__filename)
+const root = resolve(__dirname, '..')
+
+const { runScanPostCheck } = await import(pathToFileURL(join(root, 'src', 'scan-postcheck.js')).href)
+
+let passed = 0, failed = 0
+
+function assert(cond, msg) {
+  if (cond) { console.log(`  ✅ PASS: ${msg}`); passed++ }
+  else { console.log(`  ❌ FAIL: ${msg}`); failed++ }
+}
+
+function setup(name) {
+  const cwd = join(tmpdir(), `kn-${name}`)
+  mkdirSync(cwd, { recursive: true })
+  return cwd
+}
+function specSetup(name) {
+  const d = join(tmpdir(), `kn-${name}-spec`)
+  mkdirSync(d, { recursive: true })
+  return d
+}
+function clean(...dirs) { for (const d of dirs) try { rmSync(d, { recursive: true, force: true }) } catch {} }
+
+const DOCS = ['ARCHITECTURE.md','CONVENTIONS.md','STRUCTURE.md','INTEGRATIONS.md','TESTING.md','CONCERNS.md','PROJECT.md']
+
+function writeFull(cwd, specDir) {
+  const proj = basename(cwd)
+  for (const d of DOCS) {
+    const p = join(specDir, 'docs', proj, 'scan', d)
+    mkdirSync(dirname(p), { recursive: true })
+    writeFileSync(p, 'author: bot\ncreated_at: 2026-06-19 10:00:00\n# doc\n')
+  }
+}
+
+// ── 1: knowledge INDEX.md 存在，引用真实文件 → 通过 ──
+console.log('\n=== Test 1: knowledge INDEX.md 存在且引用有效 → no knowledge warning ===')
+{
+  const cwd = setup('t1'), spec = specSetup('t1')
+  try {
+    writeFull(cwd, spec)
+    const knowledgeDir = join(spec, 'knowledge')
+    mkdirSync(knowledgeDir, { recursive: true })
+    writeFileSync(join(knowledgeDir, 'INDEX.md'),
+      '# Knowledge Index\n\n## Conventions\n- [kebab-case naming](conventions.md#kebab-case-naming)\n')
+    writeFileSync(join(knowledgeDir, 'conventions.md'),
+      '# Conventions\n\n## kebab-case-naming\nFiles use kebab-case.\n')
+
+    const result = runScanPostCheck({ cwd, specDir: spec })
+    const knowledgeChecks = result.checks.filter(c => c.name.startsWith('knowledge_'))
+    assert(knowledgeChecks.length === 0, 'no knowledge warnings when INDEX.md + referenced files exist')
+  } finally { clean(cwd, spec) }
+}
+
+// ── 2: knowledge 目录不存在 → dir_missing 警告 ──
+console.log('\n=== Test 2: knowledge 目录不存在 → knowledge_dir_missing ===')
+{
+  const cwd = setup('t2'), spec = specSetup('t2')
+  try {
+    writeFull(cwd, spec)
+
+    const result = runScanPostCheck({ cwd, specDir: spec })
+    const missing = result.checks.find(c => c.name === 'knowledge_dir_missing')
+    assert(!!missing, 'knowledge_dir_missing check exists')
+    assert(missing.detail.includes('knowledge/ 目录不存在'), 'detail text correct')
+  } finally { clean(cwd, spec) }
+}
+
+// ── 3: INDEX.md 不存在 → index_missing 警告 ──
+console.log('\n=== Test 3: INDEX.md 不存在 → knowledge_index_missing ===')
+{
+  const cwd = setup('t3'), spec = specSetup('t3')
+  try {
+    writeFull(cwd, spec)
+    const knowledgeDir = join(spec, 'knowledge')
+    mkdirSync(knowledgeDir, { recursive: true })
+    // INDEX.md not created
+
+    const result = runScanPostCheck({ cwd, specDir: spec })
+    const missing = result.checks.find(c => c.name === 'knowledge_index_missing')
+    assert(!!missing, 'knowledge_index_missing check exists')
+    assert(missing.detail.includes('INDEX.md 不存在'), 'detail text correct')
+  } finally { clean(cwd, spec) }
+}
+
+// ── 4: INDEX.md 引用不存在文件 → broken_refs 警告 ──
+console.log('\n=== Test 4: INDEX.md 引用不存在的文件 → knowledge_broken_refs ===')
+{
+  const cwd = setup('t4'), spec = specSetup('t4')
+  try {
+    writeFull(cwd, spec)
+    const knowledgeDir = join(spec, 'knowledge')
+    mkdirSync(knowledgeDir, { recursive: true })
+    writeFileSync(join(knowledgeDir, 'INDEX.md'),
+      '# Knowledge Index\n\n## Conventions\n- [naming](conventions.md#naming)\n\n## Patterns\n- [auth pattern](patterns.md#auth-pattern)\n')
+    // conventions.md exists, patterns.md does not
+    writeFileSync(join(knowledgeDir, 'conventions.md'), '# Conventions\n')
+
+    const result = runScanPostCheck({ cwd, specDir: spec })
+    const broken = result.checks.find(c => c.name === 'knowledge_broken_refs')
+    assert(!!broken, 'knowledge_broken_refs check exists')
+    assert(broken.detail.includes('patterns.md'), 'reports missing patterns.md')
+  } finally { clean(cwd, spec) }
+}
+
+// ── 5: uncategorized.md 为空但存在 → 通过 ──
+console.log('\n=== Test 5: uncategorized.md 空但存在 → no knowledge warning ===')
+{
+  const cwd = setup('t5'), spec = specSetup('t5')
+  try {
+    writeFull(cwd, spec)
+    const knowledgeDir = join(spec, 'knowledge')
+    mkdirSync(knowledgeDir, { recursive: true })
+    writeFileSync(join(knowledgeDir, 'INDEX.md'),
+      '# Knowledge Index\n\n(no entries yet)\n')
+    writeFileSync(join(knowledgeDir, 'uncategorized.md'), '# Uncategorized\n')
+
+    const result = runScanPostCheck({ cwd, specDir: spec })
+    const knowledgeChecks = result.checks.filter(c => c.name.startsWith('knowledge_'))
+    assert(knowledgeChecks.length === 0, 'empty uncategorized.md passes')
+  } finally { clean(cwd, spec) }
+}
+
+// ── 6: categorized 文件存在但 INDEX.md 无条目 → 通过（不强求 categorized 非空）──
+console.log('\n=== Test 6: categorized 文件存在 + INDEX.md 有效 → no knowledge warning ===')
+{
+  const cwd = setup('t6'), spec = specSetup('t6')
+  try {
+    writeFull(cwd, spec)
+    const knowledgeDir = join(spec, 'knowledge')
+    mkdirSync(knowledgeDir, { recursive: true })
+    writeFileSync(join(knowledgeDir, 'INDEX.md'),
+      '# Knowledge Index\n\n## Known Issues\n- [GLM proxy](known-issues.md#glm-proxy)\n')
+    writeFileSync(join(knowledgeDir, 'known-issues.md'),
+      '# Known Issues\n\n## glm-proxy\nGLM proxy has usage metadata limits.\n')
+    writeFileSync(join(knowledgeDir, 'uncategorized.md'), '# Uncategorized\n')
+
+    const result = runScanPostCheck({ cwd, specDir: spec })
+    const knowledgeChecks = result.checks.filter(c => c.name.startsWith('knowledge_'))
+    assert(knowledgeChecks.length === 0, 'at least one categorized file + valid INDEX passes')
+  } finally { clean(cwd, spec) }
+}
+
+// ── 7: source_root knowledge 泄漏检测（平台模式已有检查）──
+console.log('\n=== Test 7: source_root knowledge 泄漏 → source_root_leak ===')
+{
+  const cwd = setup('t7'), spec = specSetup('t7')
+  try {
+    writeFull(cwd, spec)
+    // 写入 source_root/.sillyspec/knowledge/ 泄漏
+    const leakedKnowledgeDir = join(cwd, '.sillyspec', 'knowledge')
+    mkdirSync(leakedKnowledgeDir, { recursive: true })
+    writeFileSync(join(leakedKnowledgeDir, 'INDEX.md'), '# leaked\n')
+
+    const result = runScanPostCheck({ cwd, specDir: spec })
+    const leak = result.checks.find(c => c.name === 'source_root_leak')
+    assert(!!leak, 'source_root knowledge leak detected')
+    assert(leak.severity === 'failed', 'leak severity is failed')
+  } finally { clean(cwd, spec) }
+}
+
+// ── 汇总 ──
+console.log(`\n${'='.repeat(40)}`)
+console.log(`scan-knowledge tests: ${passed} passed, ${failed} failed, ${passed + failed} total`)
+if (failed > 0) process.exit(1)

package/test/scan-paths.test.mjs

@@ -1,68 +1,68 @@
-/**
- * 防回归测试：scan.js 中不允许硬编码 .sillyspec/docs/<project>/ 作为写入路径
- * 所有正式文档路径必须使用 {DOCS_ROOT} 占位符
- */
-import { readFileSync } from 'fs'
-import { join, dirname } from 'path'
-import { fileURLToPath } from 'url'
-
-const __dirname = dirname(fileURLToPath(import.meta.url))
-const scanPath = join(__dirname, '..', 'src', 'stages', 'scan.js')
-const content = readFileSync(scanPath, 'utf8')
-
-const banned = [
-  '.sillyspec/docs/<project>/scan/',
-  '.sillyspec/docs/<project>/modules/',
-  '.sillyspec/docs/<project>/flows/',
-  '.sillyspec/docs/<project>/glossary.md',
-]
-
-const required = [
-  '{DOCS_ROOT}/scan/',
-  '{DOCS_ROOT}/modules/',
-  '{DOCS_ROOT}/flows/',
-]
-
-let failed = false
-
-// 禁止硬编码路径
-for (const pattern of banned) {
-  if (content.includes(pattern)) {
-    console.error(`❌ FAIL: scan.js 仍包含硬编码路径 "${pattern}"`)
-    failed = true
-  } else {
-    console.log(`✅ PASS: 不包含 "${pattern}"`)
-  }
-}
-
-// 必须包含占位符
-for (const pattern of required) {
-  if (content.includes(pattern)) {
-    console.log(`✅ PASS: 包含占位符 "${pattern}"`)
-  } else {
-    console.error(`❌ FAIL: scan.js 缺少占位符 "${pattern}"`)
-    failed = true
-  }
-}
-
-// 禁止硬编码 projects 路径
-if (content.includes('.sillyspec/projects/')) {
-  console.error('❌ FAIL: scan.js 仍包含硬编码 ".sillyspec/projects/"')
-  failed = true
-} else {
-  console.log('✅ PASS: 不包含 ".sillyspec/projects/"')
-}
-
-if (content.includes('{PROJECTS_ROOT}/')) {
-  console.log('✅ PASS: 包含占位符 "{PROJECTS_ROOT}/"')
-} else {
-  console.error('❌ FAIL: scan.js 缺少占位符 "{PROJECTS_ROOT}/"')
-  failed = true
-}
-
-if (failed) {
-  console.error('\n💥 有测试失败！scan.js 路径占位符可能被回退为硬编码。')
-  throw new Error("test failed")
-} else {
-  console.log('\n✅ 全部通过 — scan.js 路径占位符防回归测试 OK')
-}
+/**
+ * 防回归测试：scan.js 中不允许硬编码 .sillyspec/docs/<project>/ 作为写入路径
+ * 所有正式文档路径必须使用 {DOCS_ROOT} 占位符
+ */
+import { readFileSync } from 'fs'
+import { join, dirname } from 'path'
+import { fileURLToPath } from 'url'
+
+const __dirname = dirname(fileURLToPath(import.meta.url))
+const scanPath = join(__dirname, '..', 'src', 'stages', 'scan.js')
+const content = readFileSync(scanPath, 'utf8')
+
+const banned = [
+  '.sillyspec/docs/<project>/scan/',
+  '.sillyspec/docs/<project>/modules/',
+  '.sillyspec/docs/<project>/flows/',
+  '.sillyspec/docs/<project>/glossary.md',
+]
+
+const required = [
+  '{DOCS_ROOT}/scan/',
+  '{DOCS_ROOT}/modules/',
+  '{DOCS_ROOT}/flows/',
+]
+
+let failed = false
+
+// 禁止硬编码路径
+for (const pattern of banned) {
+  if (content.includes(pattern)) {
+    console.error(`❌ FAIL: scan.js 仍包含硬编码路径 "${pattern}"`)
+    failed = true
+  } else {
+    console.log(`✅ PASS: 不包含 "${pattern}"`)
+  }
+}
+
+// 必须包含占位符
+for (const pattern of required) {
+  if (content.includes(pattern)) {
+    console.log(`✅ PASS: 包含占位符 "${pattern}"`)
+  } else {
+    console.error(`❌ FAIL: scan.js 缺少占位符 "${pattern}"`)
+    failed = true
+  }
+}
+
+// 禁止硬编码 projects 路径
+if (content.includes('.sillyspec/projects/')) {
+  console.error('❌ FAIL: scan.js 仍包含硬编码 ".sillyspec/projects/"')
+  failed = true
+} else {
+  console.log('✅ PASS: 不包含 ".sillyspec/projects/"')
+}
+
+if (content.includes('{PROJECTS_ROOT}/')) {
+  console.log('✅ PASS: 包含占位符 "{PROJECTS_ROOT}/"')
+} else {
+  console.error('❌ FAIL: scan.js 缺少占位符 "{PROJECTS_ROOT}/"')
+  failed = true
+}
+
+if (failed) {
+  console.error('\n💥 有测试失败！scan.js 路径占位符可能被回退为硬编码。')
+  throw new Error("test failed")
+} else {
+  console.log('\n✅ 全部通过 — scan.js 路径占位符防回归测试 OK')
+}