citation-evidence/tests/integration/anchor-source-roundtrip.test.ts

/**
 * Integration round-trip: ingest → createSelectors → resolveSelectors.
 *
 * This test crosses the source ↔ anchor boundary, which the
 * `boundaries/element-types` lint rule (correctly) forbids inside `src/`.
 * It lives under `tests/integration/` so it can verify the end-to-end
 * MVP contract without weakening the production-code boundary.
 *
 * CE-WP-0002-T04 contract:
 *   For each fixture+known-quote pair, create selectors then immediately
 *   resolve them; resolution must succeed with confidence ≥ 0.9.
 */

import { readFileSync } from "node:fs";
import { dirname, resolve } from "node:path";
import { createRequire } from "node:module";
import { fileURLToPath } from "node:url";
import { beforeAll, describe, expect, it } from "vitest";

import { ingestPdf } from "@source/pdf/ingest";
import { createSelectors, resolveSelectors } from "@anchor/selectors";
import type { PdfSelectionCapture } from "@anchor/types";
import manifest from "../../fixtures/pdfs/manifest.json" with { type: "json" };

const __dirname = dirname(fileURLToPath(import.meta.url));
const FIXTURE_DIR = resolve(__dirname, "../../fixtures/pdfs");

interface Fixture {
  id: string;
  filename: string;
  known_good_quote: string;
  known_good_quote_page: number;
}

const FIXTURES: readonly Fixture[] = manifest.fixtures;

beforeAll(async () => {
  const pdfjs = await import("pdfjs-dist");
  const require = createRequire(import.meta.url);
  pdfjs.GlobalWorkerOptions.workerSrc = require.resolve(
    "pdfjs-dist/legacy/build/pdf.worker.mjs",
  );
});

describe("create + resolve round-trip — fixture corpus", { timeout: 30_000 }, () => {
  for (const fixture of FIXTURES) {
    it(`${fixture.id}: known-good quote round-trips with confidence ≥ 0.9`, async () => {
      const bytes = new Uint8Array(readFileSync(resolve(FIXTURE_DIR, fixture.filename)));
      const { representation } = await ingestPdf(bytes, { filename: fixture.filename });

      const capture: PdfSelectionCapture = {
        kind: "pdf",
        text: fixture.known_good_quote,
        page: fixture.known_good_quote_page,
        rects: [{ x: 0.1, y: 0.2, width: 0.5, height: 0.04 }],
      };

      const selectors = createSelectors(capture, representation);
      const resolution = resolveSelectors(selectors, representation);

      expect(resolution.status).toBe("resolved");
      expect(resolution.confidence).toBeGreaterThanOrEqual(0.9);

      const span = resolution.candidates[0]?.textPosition;
      expect(span).toBeDefined();
      const text = representation.canonicalText ?? "";
      expect(text.slice(span!.start, span!.end)).toBe(fixture.known_good_quote);

      expect(resolution.candidates[0]?.page).toBe(fixture.known_good_quote_page);
    });
  }
});