package-gutenberg-runtime/test/catalog.test.mjs

import assert from "node:assert/strict";
import { test, after } from "node:test";
import { stripTypeScriptTypes } from "node:module";
import { readFile, mkdtemp, writeFile, rm } from "node:fs/promises";
import { tmpdir } from "node:os";
import { join } from "node:path";
import { pathToFileURL } from "node:url";
const directory = await mkdtemp(join(tmpdir(), "gutenberg-test-"));
for (const name of ["catalog", "download"]) {
  const source = await readFile(new URL("../src/" + name + ".ts", import.meta.url), "utf8");
  await writeFile(
    join(directory, name + ".mjs"),
    stripTypeScriptTypes(source).replace("./catalog.js", "./catalog.mjs"),
  );
}
after(() => rm(directory, { recursive: true, force: true }));
const { parseIndex, searchIndex, openingExcerpt } = await import(pathToFileURL(join(directory, "catalog.mjs")));
const { createBookLoader, bookURL, fetchBook } = await import(pathToFileURL(join(directory, "download.mjs")));
test("CSV preserves quoted/multiline titles and selects text editions", () => {
  const books = parseIndex(
    'Text#,Type,Title,Language,Authors\r\n1342,Text,"Pride, and\r\nPrejudice",en,"Austen, Jane"\r\n11,Text,"Alice and ""friends""",en,Carroll\r\n12,Sound,Audio,en,Someone\r\n',
  );
  assert.deepEqual(books, [
    { ebook: 1342, title: "Pride, and Prejudice", language: "en", author: "Austen, Jane" },
    { ebook: 11, title: 'Alice and "friends"', language: "en", author: "Carroll" },
  ]);
  assert.throws(() => parseIndex('Text#,Type,Title,Language,Authors\n1,Text,"broken'), /Unclosed/);
});
test("fuzzy titles and diacritics work, with a bounded result list", () => {
  const books = [
    { ebook: 1342, title: "Pride and Prejudice", author: "Austen, Jane", language: "en" },
    { ebook: 11, title: "Alice in Wonderland", author: "Carroll", language: "en" },
    ...Array.from({ length: 40 }, (_, i) => ({
      ebook: 100 + i,
      title: "Émile " + i,
      author: "Rousseau",
      language: "fr",
    })),
  ];
  const search = searchIndex(books);
  assert.equal(search("prde and prejudi").at(0).ebook, 1342);
  assert.equal(search("austen").at(0).ebook, 1342);
  assert.equal(search("emile").length, 20);
  assert.deepEqual(search("a"), []);
});
test("opening excerpt keeps paragraph boundaries and excludes Gutenberg framing", () => {
  const body =
    "license\n*** START OF THE PROJECT GUTENBERG EBOOK TEST ***\n\nFirst\nline\n\nSecond\n\n*** END OF THE PROJECT GUTENBERG EBOOK TEST ***\nlicense";
  assert.deepEqual(openingExcerpt(body), { text: "First\nline\n\nSecond", words: 3 });
  const first = "word ".repeat(600).trim(),
    second = "more ".repeat(450).trim();
  const text = "*** START OF THE PROJECT GUTENBERG EBOOK TEST ***\n\n" + first + "\n\n" + second;
  assert.equal(openingExcerpt(text), null, "Wait for the whole final paragraph");
  assert.deepEqual(openingExcerpt(text + "\n\nignored"), { text: first + "\n\n" + second, words: 1050 });
  assert.deepEqual(openingExcerpt("short book", true), { text: "short book", words: 2 });
  const long = "😀".repeat(30000);
  const bounded = openingExcerpt(long, true);
  assert(bounded.text.length <= 56000);
  assert(!/[\uD800-\uDBFF]$/.test(bounded.text));
});
test("streaming fetch stops at the excerpt and releases the remaining download", async () => {
  const original = globalThis.fetch;
  let cancelled = false,
    reads = 0;
  const chunks = [
    "licence\n*** START OF THE PROJECT GUTEN",
    "BERG EBOOK TEST ***\n\n",
    "word ".repeat(600) + "\n\n",
    "more ".repeat(450) + "\n\n",
    "unused rest of the book",
  ];
  globalThis.fetch = async () =>
    new Response(
      new ReadableStream(
        {
          pull(controller) {
            reads++;
            const chunk = chunks.shift();
            if (chunk) controller.enqueue(new TextEncoder().encode(chunk));
            else controller.close();
          },
          cancel() {
            cancelled = true;
          },
        },
        { highWaterMark: 0 },
      ),
    );
  try {
    const text = await fetchBook(1342);
    assert.equal(reads, 4);
    assert.equal(cancelled, true);
    assert.equal(openingExcerpt(text, true).words, 1050);
    assert(!text.includes("unused"));
  } finally {
    globalThis.fetch = original;
  }
});
test("downloads coalesce, cache, serialize, and permit retries after failure", async () => {
  let calls = 0,
    active = 0,
    maximum = 0,
    fail = true;
  const load = createBookLoader(async (ebook) => {
    calls++;
    active++;
    maximum = Math.max(maximum, active);
    await new Promise((r) => setTimeout(r, 2));
    active--;
    if (ebook === 3 && fail) {
      fail = false;
      throw Error("offline");
    }
    return "Book " + ebook;
  }, 1);
  assert.deepEqual(await Promise.all([load(1), load(1), load(2)]), [
    { text: "Book 1", words: 2 },
    { text: "Book 1", words: 2 },
    { text: "Book 2", words: 2 },
  ]);
  assert.equal(calls, 2);
  assert.equal(maximum, 1);
  await load(1);
  assert.equal(calls, 2);
  await assert.rejects(load(3), /offline/);
  assert.deepEqual(await load(3), { text: "Book 3", words: 2 });
  assert.equal(bookURL(1342), "https://gutenberg.pglaf.org/cache/epub/1342/pg1342.txt");
});