@ontrails/trails 1.0.0-beta.5 → 1.0.0-beta.50

package/src/regrade/live-api-preserve.ts

@@ -0,0 +1,123 @@
+import { Result, ValidationError, escapeRegExp } from '@ontrails/core';
+import type { Result as TrailsResult } from '@ontrails/core';
+import type {
+  VocabularyPreserveInventoryEntry,
+  VocabularyRegradePlan,
+} from '@ontrails/regrade';
+import { deriveVocabularyFormProposals } from '@ontrails/regrade';
+import { existsSync, readFileSync } from 'node:fs';
+import { join } from 'node:path';
+import { z } from 'zod';
+
+const lockShapeSchema = z
+  .object({
+    topoGraph: z
+      .object({
+        entries: z.array(z.object({ id: z.string() }).passthrough()),
+        library: z
+          .object({
+            exports: z.array(
+              z.object({ exportName: z.string() }).passthrough()
+            ),
+          })
+          .passthrough()
+          .optional(),
+      })
+      .passthrough(),
+  })
+  .passthrough();
+
+type LockShape = z.output<typeof lockShapeSchema>;
+
+const liveApiValues = (
+  lock: LockShape
+): readonly {
+  readonly evidence: string;
+  readonly value: string;
+}[] => {
+  const values = new Map<string, string>();
+  for (const entry of lock.topoGraph?.entries ?? []) {
+    if (typeof entry.id === 'string') {
+      values.set(entry.id, `topo.entry:${entry.id}`);
+    }
+  }
+  for (const item of lock.topoGraph?.library?.exports ?? []) {
+    if (typeof item.exportName === 'string') {
+      values.set(item.exportName, `topo.library:${item.exportName}`);
+    }
+  }
+  return [...values.entries()]
+    .map(([value, evidence]) => ({ evidence, value }))
+    .toSorted((left, right) => left.value.localeCompare(right.value));
+};
+
+const identifierSegments = (value: string): readonly string[] =>
+  value
+    .split(/[^A-Za-z0-9]+/u)
+    .flatMap(
+      (part) => part.match(/[A-Z]+(?=[A-Z][a-z]|\d|$)|[A-Z]?[a-z]+|\d+/gu) ?? []
+    );
+
+const liveApiValueContainsSource = (value: string, source: string): boolean => {
+  if (/^[A-Za-z][A-Za-z0-9]*$/u.test(source)) {
+    const normalized = source.toLowerCase();
+    return identifierSegments(value).some(
+      (segment) => segment.toLowerCase() === normalized
+    );
+  }
+  return new RegExp(
+    `(?<![A-Za-z0-9_$])${escapeRegExp(source)}(?![A-Za-z0-9_$])`,
+    'iu'
+  ).test(value);
+};
+
+export const deriveLiveApiPreserveInventory = async (
+  plan: VocabularyRegradePlan,
+  rootDir = process.cwd()
+): Promise<
+  TrailsResult<readonly VocabularyPreserveInventoryEntry[], ValidationError>
+> => {
+  const lockPath = join(rootDir, 'trails.lock');
+  if (!existsSync(lockPath)) {
+    return Result.ok([]);
+  }
+  let input: unknown;
+  try {
+    input = JSON.parse(readFileSync(lockPath, 'utf8'));
+  } catch (error) {
+    return Result.err(
+      new ValidationError(
+        'Unable to parse trails.lock for live API preserves.',
+        {
+          ...(error instanceof Error ? { cause: error } : {}),
+          context: { path: lockPath },
+        }
+      )
+    );
+  }
+  const parsed = lockShapeSchema.safeParse(input);
+  if (!parsed.success) {
+    return Result.err(
+      new ValidationError(
+        'trails.lock does not contain a compatible TopoGraph for live API preserves.',
+        { context: { issues: parsed.error.issues, path: lockPath } }
+      )
+    );
+  }
+  const sourceForms = deriveVocabularyFormProposals(plan).map(
+    (proposal) => proposal.from
+  );
+  return Result.ok(
+    liveApiValues(parsed.data)
+      .filter(({ value }) =>
+        sourceForms.some((source) => liveApiValueContainsSource(value, source))
+      )
+      .map(({ evidence, value }) => ({
+        disposition: 'preserve-current-live-api' as const,
+        evidence: [evidence],
+        pattern: escapeRegExp(value),
+        reason: 'current-live-topo-api',
+        source: 'derived-live-api' as const,
+      }))
+  );
+};

package/src/regrade/plan-artifact.ts

@@ -0,0 +1,461 @@
+/**
+ * Saved Regrade plan artifact shape plus the slug, path, and hash helpers
+ * shared by the `regrade` trails and the consolidated Regrade history module.
+ */
+
+import { isPlainObject } from '@ontrails/core';
+import { vocabularyRegradePlanSchema } from '@ontrails/regrade';
+import type {
+  RegradeReport,
+  RegradeReportEntry,
+  VocabularyRegradePlan,
+} from '@ontrails/regrade';
+import { getGovernedVocabularyTransition } from '@ontrails/warden';
+import { createHash } from 'node:crypto';
+import { join, normalize, relative } from 'node:path';
+import { z } from 'zod';
+
+export const REGRADE_PLAN_SCHEMA_VERSION = 1;
+
+const regradePlanProvenanceValueSchema = z.enum(['authored', 'derived']);
+
+const regradePlanDerivationSchema = z
+  .object({
+    fileRenames: z.array(
+      z
+        .object({
+          evidence: z.array(z.string()),
+          from: z.string(),
+          provenance: z.literal('derived'),
+          status: z.literal('pending'),
+          to: z.string(),
+        })
+        .strict()
+    ),
+    forms: z.array(
+      z
+        .object({
+          from: z.string(),
+          kind: z.enum(['review', 'safe-rewrite']),
+          provenance: regradePlanProvenanceValueSchema,
+          reason: z.string(),
+          source: z.enum([
+            'default-morphology',
+            'plan-defer',
+            'plan-override',
+            'seed',
+          ]),
+          to: z.string().optional(),
+        })
+        .strict()
+    ),
+    namespaces: z.array(
+      z
+        .object({
+          inScope: z.number().int().nonnegative(),
+          namespace: z.string(),
+          policyClassified: z.number().int().nonnegative(),
+          provenance: z.literal('derived'),
+        })
+        .strict()
+    ),
+    preserves: z.array(
+      z
+        .object({
+          disposition: z.literal('preserve-current-live-api'),
+          evidence: z.array(z.string()),
+          pattern: z.string(),
+          provenance: z.literal('derived'),
+          reason: z.string().optional(),
+        })
+        .strict()
+    ),
+    referenceClosure: z
+      .object({
+        entries: z.array(
+          z
+            .object({
+              outcome: z.enum(['needs-review', 'rewrite']),
+              path: z.string(),
+              provenance: z.literal('derived'),
+              reason: z.string().optional(),
+            })
+            .strict()
+        ),
+        issue: z.string().optional(),
+        moves: z.array(
+          z
+            .object({
+              deferred: z.number().int().nonnegative(),
+              from: z.string(),
+              historical: z.number().int().nonnegative(),
+              preserved: z.number().int().nonnegative(),
+              provenance: z.literal('derived'),
+              rewritten: z.number().int().nonnegative(),
+              skipped: z.number().int().nonnegative(),
+              to: z.string(),
+            })
+            .strict()
+        ),
+      })
+      .strict(),
+    reviews: z.array(
+      z
+        .object({
+          evidence: z.array(
+            z
+              .object({
+                column: z.number().int().positive().optional(),
+                line: z.number().int().positive().optional(),
+                path: z.string(),
+              })
+              .strict()
+          ),
+          provenance: z.literal('derived'),
+          reason: z.string(),
+          status: z.literal('pending'),
+          value: z.string(),
+        })
+        .strict()
+    ),
+  })
+  .strict();
+
+const classRegradePlanScopeSchema = z
+  .object({
+    exclude: z
+      .array(z.string())
+      .optional()
+      .describe('Root-relative path globs excluded from the class run'),
+    extensions: z
+      .array(z.string())
+      .optional()
+      .describe('Source file extensions scanned by the class run'),
+    include: z
+      .array(z.string())
+      .optional()
+      .describe('Root-relative path globs collected during the class run'),
+  })
+  .strict();
+
+/**
+ * A saved class-mode Regrade plan: which classes run, over what scope, and
+ * why. The parallel payload to {@link vocabularyRegradePlanSchema} — the
+ * `kind` discriminant keeps existing vocabulary plan artifacts
+ * byte-compatible.
+ */
+const classRegradePlanSchema = z.object({
+  classIds: z
+    .array(z.string().min(1))
+    .min(1)
+    .describe('Regrade class ids this plan runs'),
+  id: z.string().min(1).describe('Stable Regrade plan identifier'),
+  intent: z
+    .string()
+    .optional()
+    .describe('Human-authored migration intent for the class run'),
+  kind: z.literal('class').describe('Regrade plan kind'),
+  name: z
+    .string()
+    .min(1)
+    .optional()
+    .describe(
+      'Authored transition name; keys the saved plan and consolidated history filenames'
+    ),
+  scope: classRegradePlanScopeSchema
+    .optional()
+    .describe('Collection scope for the class run'),
+});
+
+export type ClassRegradePlan = z.output<typeof classRegradePlanSchema>;
+
+const regradePlanBodySchema = z.discriminatedUnion('kind', [
+  vocabularyRegradePlanSchema,
+  classRegradePlanSchema,
+]);
+
+export type RegradePlanBody = VocabularyRegradePlan | ClassRegradePlan;
+
+const regradeExpansionCandidateSchema = z.union([
+  z.object({
+    evidence: z
+      .array(
+        z.object({
+          column: z.number().optional(),
+          detail: z.string().optional(),
+          line: z.number().optional(),
+          path: z.string(),
+        })
+      )
+      .default([]),
+    kind: z.enum(['file-rename', 'form', 'namespace', 'preserve']),
+    provenance: regradePlanProvenanceValueSchema.default('derived'),
+    reason: z.string().optional(),
+    status: z.enum(['pending', 'rejected']).default('pending'),
+    suggestedClassification: z.string(),
+    value: z.string(),
+  }),
+  z
+    .object({
+      detail: z.string().optional(),
+      path: z.string(),
+      status: z.enum(['pending', 'rejected']).default('pending'),
+    })
+    .transform((candidate) => ({
+      evidence: [
+        {
+          ...(candidate.detail === undefined
+            ? {}
+            : { detail: candidate.detail }),
+          path: candidate.path,
+        },
+      ],
+      kind: 'file-rename' as const,
+      provenance: 'derived' as const,
+      ...(candidate.detail === undefined ? {} : { reason: candidate.detail }),
+      status: candidate.status,
+      suggestedClassification: 'legacy-path-candidate',
+      value: candidate.path,
+    })),
+]);
+
+export const regradePlanArtifactSchema = z
+  .object({
+    derivation: regradePlanDerivationSchema.optional(),
+    expansion: z
+      .object({
+        candidates: z.array(regradeExpansionCandidateSchema).default([]),
+      })
+      .optional(),
+    kind: z.literal('regrade-plan'),
+    path: z.string(),
+    plan: regradePlanBodySchema,
+    provenance: z.object({
+      fields: z.record(z.string(), regradePlanProvenanceValueSchema),
+    }),
+    schemaVersion: z.literal(REGRADE_PLAN_SCHEMA_VERSION),
+    sourceHash: z.string(),
+    transitionId: z
+      .string()
+      .min(1)
+      .optional()
+      .describe(
+        'Stable transition identity this plan re-runs; preserves the consolidated history spine'
+      ),
+  })
+  .strict();
+
+export interface RegradePlanExpansion {
+  readonly candidates: readonly {
+    readonly evidence: readonly {
+      readonly column?: number | undefined;
+      readonly detail?: string | undefined;
+      readonly line?: number | undefined;
+      readonly path: string;
+    }[];
+    readonly kind: 'file-rename' | 'form' | 'namespace' | 'preserve';
+    readonly provenance: 'authored' | 'derived';
+    readonly reason?: string | undefined;
+    readonly status: 'pending' | 'rejected';
+    readonly suggestedClassification: string;
+    readonly value: string;
+  }[];
+}
+
+export interface RegradePlanArtifact {
+  readonly derivation?: RegradePlanDerivation | undefined;
+  readonly expansion?: RegradePlanExpansion | undefined;
+  readonly kind: 'regrade-plan';
+  readonly path: string;
+  readonly plan: RegradePlanBody;
+  readonly provenance: {
+    readonly fields: Readonly<Record<string, 'authored' | 'derived'>>;
+  };
+  readonly schemaVersion: typeof REGRADE_PLAN_SCHEMA_VERSION;
+  readonly sourceHash: string;
+  readonly transitionId?: string | undefined;
+}
+
+export type RegradePlanDerivation = z.output<
+  typeof regradePlanDerivationSchema
+>;
+
+/** A plan artifact narrowed to a vocabulary plan body. */
+export type VocabularyRegradePlanArtifact = RegradePlanArtifact & {
+  readonly plan: VocabularyRegradePlan;
+};
+
+const regradeSlugText = (text: string): string =>
+  text
+    .toLowerCase()
+    .replaceAll(/[^a-z0-9]+/g, '-')
+    .replaceAll(/^-|-$/g, '');
+
+const regradePlanSlug = (plan: Pick<VocabularyRegradePlan, 'from' | 'to'>) =>
+  regradeSlugText(`${plan.from}-to-${plan.to}`);
+
+export const regradePlanSlugForBody = (plan: RegradePlanBody): string =>
+  plan.kind === 'class'
+    ? regradeSlugText(plan.name ?? plan.classIds.join('-'))
+    : (() => {
+        const transition =
+          plan.id === undefined
+            ? undefined
+            : getGovernedVocabularyTransition(plan.id);
+        return transition?.target.kind === 'classified'
+          ? regradeSlugText(transition.id)
+          : regradePlanSlug(plan);
+      })();
+
+const normalizeRelativePath = (path: string): string =>
+  normalize(path).replaceAll('\\', '/');
+
+export const rootRelativePath = (
+  rootDir: string,
+  absolutePath: string
+): string => normalizeRelativePath(relative(rootDir, absolutePath));
+
+export const regradePlanDirectory = (rootDir: string): string =>
+  join(rootDir, '.trails', 'regrade');
+
+export const regradePlanPathForPlan = (
+  rootDir: string,
+  plan: RegradePlanBody
+): string =>
+  join(regradePlanDirectory(rootDir), `${regradePlanSlugForBody(plan)}.json`);
+
+const sourceHashEntryFacts = (
+  entries: readonly RegradeReportEntry[]
+): readonly Pick<
+  RegradeReportEntry,
+  'classId' | 'notes' | 'outcome' | 'path' | 'reason' | 'reviewDetails'
+>[] =>
+  entries
+    .filter(
+      (entry) => entry.outcome === 'rewrite' || entry.outcome === 'needs-review'
+    )
+    .map(({ classId, notes, outcome, path, reason, reviewDetails }) => ({
+      ...(classId === undefined ? {} : { classId }),
+      ...(notes === undefined ? {} : { notes }),
+      outcome,
+      path,
+      ...(reason === undefined ? {} : { reason }),
+      ...(reviewDetails === undefined ? {} : { reviewDetails }),
+    }));
+
+const canonicalizeJsonValue = (value: unknown): unknown => {
+  if (Array.isArray(value)) {
+    return value.map((entry) => canonicalizeJsonValue(entry));
+  }
+  if (isPlainObject(value)) {
+    return Object.fromEntries(
+      Object.keys(value)
+        .toSorted()
+        .map((key) => [key, canonicalizeJsonValue(value[key])])
+    );
+  }
+  return value;
+};
+
+/**
+ * JSON.stringify with recursively sorted object keys so structurally equal
+ * values serialize identically regardless of key insertion order. Arrays keep
+ * their authored order.
+ */
+export const canonicalJsonStringify = (value: unknown): string =>
+  JSON.stringify(canonicalizeJsonValue(value));
+
+export const isGeneratedRegradeArtifactPath = (path: string): boolean =>
+  /(?:^|\/)\.trails\/regrade\/.+\.json$/u.test(path);
+
+const sourceHashLedgerFacts = (
+  report: RegradeReport,
+  policyMode: 'current' | 'legacy'
+): unknown => {
+  const ledger = report.run?.ledger;
+  if (ledger === undefined) {
+    return undefined;
+  }
+  // The active plan is written after its source hash and must not stale itself.
+  const occurrences = ledger.occurrences.filter(
+    (occurrence) =>
+      occurrence.scopeTier !== 'policy-classified' ||
+      (policyMode === 'current' &&
+        !isGeneratedRegradeArtifactPath(occurrence.path))
+  );
+  const forms = new Set(occurrences.map((occurrence) => occurrence.form));
+  return {
+    cycle: ledger.cycle,
+    forms: Object.fromEntries(
+      Object.entries(ledger.forms).filter(([form]) => forms.has(form))
+    ),
+    occurrences,
+  };
+};
+
+const regradeSourceHashFacts = (
+  report: RegradeReport,
+  policyMode: 'current' | 'legacy' = 'current',
+  fileEvidenceMode: 'current' | 'legacy' = 'current'
+): unknown => ({
+  entries: sourceHashEntryFacts(report.entries),
+  fileRenames: report.run?.report.fileRenames?.map(
+    ({ deferred, from, historical, preserved, rewritten, skipped, to }) => ({
+      deferred,
+      from,
+      ...(fileEvidenceMode === 'current' ? { historical, preserved } : {}),
+      rewritten,
+      ...(fileEvidenceMode === 'current' ? { skipped } : {}),
+      to,
+    })
+  ),
+  ledger: sourceHashLedgerFacts(report, policyMode),
+  selectedClassIds: report.selectedClassIds,
+});
+
+const hashSerializedSourceFacts = (serialized: string): string =>
+  createHash('sha256').update(serialized).digest('hex');
+
+export const regradeSourceHash = (report: RegradeReport): string =>
+  hashSerializedSourceFacts(
+    canonicalJsonStringify(regradeSourceHashFacts(report))
+  );
+
+/** Match only the complete current source-evidence shape used by active plans. */
+export const currentRegradeSourceHashMatches = (
+  stampedHash: string,
+  report: RegradeReport
+): boolean => stampedHash === regradeSourceHash(report);
+
+export const regradeSourceHashes = (
+  report: RegradeReport
+): readonly string[] => {
+  const facts = [
+    regradeSourceHashFacts(report),
+    regradeSourceHashFacts(report, 'current', 'legacy'),
+    regradeSourceHashFacts(report, 'legacy'),
+    regradeSourceHashFacts(report, 'legacy', 'legacy'),
+  ];
+  return [
+    ...new Set(
+      facts.flatMap((value) => [
+        hashSerializedSourceFacts(canonicalJsonStringify(value)),
+        hashSerializedSourceFacts(JSON.stringify(value)),
+      ])
+    ),
+  ];
+};
+
+/** Match source evidence written by any supported historical hash shape. */
+export const regradeSourceHashMatches = (
+  stampedHash: string,
+  report: RegradeReport
+): boolean => regradeSourceHashes(report).includes(stampedHash);
+
+/**
+ * Canonical content hash of a resolved Regrade plan body — the authored
+ * migration intent. Stable across key insertion order; changes on any edit to
+ * the plan contents.
+ */
+export const regradePlanContentHash = (plan: RegradePlanBody): string =>
+  createHash('sha256').update(canonicalJsonStringify(plan)).digest('hex');