Hankan/test/domain/dictionary.test.ts

/* Loading the shipped dictionary, and reloading it when it changes.

   Loaded bands used to be recorded by number alone, so a rebuilt dictionary
   never reached an existing install. Reloading only became safe once an id
   named a word instead of a position — these pin both halves. */

import { describe, it, expect, beforeEach, afterEach, vi } from "vitest";
import { readFileSync } from "node:fs";
import { SqliteWasmDb } from "@app/db/sqlite-wasm-core.js";
import { migrate } from "@app/db/migrate.js";
import { editCard, editAddCustomWord } from "@app/db/writes.js";
import { markKnown } from "@lib/srs.js";
import { lemmaId } from "@shared/lemma-id.mjs";
import type { Db } from "@app/db/types.js";

const DICT = new URL("../../app/public/dict/", import.meta.url);
const shipped = JSON.parse(readFileSync(new URL("manifest.json", DICT), "utf8")) as {
  bands: { band: number; file: string; sha256: string; lemmas: number }[];
};

/** Serve the committed files; `manifest` lets a test pretend to be a rebuild. */
function serve(manifest: unknown) {
  vi.stubGlobal("fetch", async (input: string) => {
    const name = String(input).split("/dict/")[1] ?? "";
    if (name === "manifest.json") return new Response(JSON.stringify(manifest));
    return new Response(readFileSync(new URL(name, DICT)));
  });
}

async function loader() {
  vi.resetModules(); // the manifest is cached per module instance
  return import("@app/domain/dictionary.js");
}

let db: Db;
beforeEach(async () => {
  db = await SqliteWasmDb.open({ memory: true });
  await migrate(db);
});
afterEach(async () => {
  vi.unstubAllGlobals();
  await db.close();
});

const count = async (sql: string, params: (string | number)[] = []) =>
  (await db.get<{ n: number }>(sql, params))!.n;

describe("loading a band", () => {
  it("derives every id from the word, so surfaces join their lemmas", async () => {
    serve(shipped);
    const { ensureBands } = await loader();
    await ensureBands(db, 0);

    expect(await count("SELECT count(*) AS n FROM lemma")).toBe(shipped.bands[0]!.lemmas);
    const rice = await db.get<{ id: number }>("SELECT id FROM lemma WHERE headword = '밥' AND pos = 'noun'");
    expect(rice?.id).toBe(lemmaId("밥", "noun"));
    expect(await count("SELECT count(*) AS n FROM surface s LEFT JOIN lemma l ON l.id = s.lemma_id WHERE l.id IS NULL")).toBe(0);
  });

  it("does not load a band twice", async () => {
    serve(shipped);
    const { ensureBands } = await loader();
    expect(await ensureBands(db, 0)).toEqual([0]);
    expect(await ensureBands(db, 0)).toEqual([]);
  });
});

describe("a rebuilt dictionary", () => {
  it("reloads, and every card and custom word still names its word", async () => {
    serve(shipped);
    let dict = await loader();
    await dict.ensureBands(db, 0);
    const rice = lemmaId("밥", "noun");
    await editCard(db, rice, markKnown(100));
    const custom = await editAddCustomWord(db, { headword: "던전돌", gloss: "dungeon stone", pos: "noun" });

    // A rebuild: band 0's file hash changes.
    const rebuilt = structuredClone(shipped);
    rebuilt.bands[0]!.sha256 = "f".repeat(64);
    serve(rebuilt);
    dict = await loader();
    await db.run("DELETE FROM lemma WHERE headword = '물'"); // prove the reload really re-inserts
    expect(await dict.ensureBands(db, 0)).toEqual([0]);

    expect(await count("SELECT count(*) AS n FROM lemma WHERE headword = '물'")).toBeGreaterThan(0);
    const card = await db.get<{ headword: string }>(
      "SELECT l.headword FROM card c JOIN lemma l ON l.id = c.lemma_id WHERE c.lemma_id = ?",
      [rice],
    );
    expect(card?.headword).toBe("밥");
    const mine = await db.get<{ headword: string; source: string }>(
      "SELECT headword, source FROM lemma WHERE id = ?",
      [custom.lemmaId],
    );
    expect(mine).toEqual({ headword: "던전돌", source: "custom" });
  });
});