com.gregkilmer.www/legacy/scripts/fetch-book-covers.mjs

#!/usr/bin/env node
/**
 * fetch-book-covers.mjs
 * ---------------------------------------------------------------------------
 * Support tool for the Reading page. For each favorite item (those with
 * `favorite: true`) in src/data/reading.json that is missing an `image`, it
 * searches the Open Library Search API (free, no key required) to resolve the
 * work's OLID, then downloads the cover via the Open Library Covers API to
 * public/images/reading/<slug>.jpg, and writes the local public path back into
 * the JSON.
 *
 * Idempotent: items that already have an `image` are skipped. Be polite —
 * there's a small delay between requests.
 *
 * The reading.json titles are formatted like "Author — Series"; this script
 * searches by the series/title portion (after the em dash) for better matches.
 *
 * USAGE
 *   node scripts/fetch-book-covers.mjs             # fetch missing book covers
 *   node scripts/fetch-book-covers.mjs --force     # re-fetch even if image set
 *   node scripts/fetch-book-covers.mjs --sleep 500 # ms between requests
 *   node scripts/fetch-book-covers.mjs --help
 */
import { mkdir, readFile, writeFile } from 'node:fs/promises';
import { dirname, join, resolve } from 'node:path';
import { fileURLToPath } from 'node:url';

const __dirname = dirname(fileURLToPath(import.meta.url));
const PROJECT_ROOT = resolve(__dirname, '..');
const DATA_FILE = join(PROJECT_ROOT, 'src', 'data', 'reading.json');
const OUT_DIR = join(PROJECT_ROOT, 'public', 'images', 'reading');
const PUBLIC_PREFIX = '/images/reading';

const SEARCH_API = 'https://openlibrary.org/search.json';
const COVERS_API = 'https://covers.openlibrary.org/b';

// --- args -------------------------------------------------------------------
const args = process.argv.slice(2);
if (args.includes('-h') || args.includes('--help')) {
  console.log(`Usage: node scripts/fetch-book-covers.mjs [options]

Reads reading.json favorites and fills in the "image" field for any favorite
missing one, using the Open Library APIs (free, no key required).

Options:
  --force        re-fetch even for items that already have an image
  --sleep <ms>   delay between Open Library requests (default 500)
  -h, --help     show this help`);
  process.exit(0);
}

const force = args.includes('--force');
const sleepIdx = args.indexOf('--sleep');
const sleepMs = sleepIdx !== -1 ? Number(args[sleepIdx + 1]) || 500 : 500;

// --- helpers ----------------------------------------------------------------
const sleep = (ms) => new Promise((r) => setTimeout(r, ms));

const slugify = (s) =>
  s
    .toLowerCase()
    .replace(/['’.:()\u2014-]/g, ' ')
    .replace(/[^a-z0-9]+/g, '-')
    .replace(/^-+|-+$/g, '');

/** Strip an "Author — Series" prefix, leaving the searchable title/series. */
const searchableTitle = (raw) => {
  // Split on em dash (—) or en dash; take the last non-empty segment.
  const parts = raw.split(/[\u2014\u2013\u2012-]/).map((p) => p.trim()).filter(Boolean);
  return parts.length > 1 ? parts[parts.length - 1] : raw;
};

/**
 * Search Open Library for a title and return the best cover URL. Prefers an
 * edition with a cover ID; falls back to a work-level cover key.
 */
async function fetchCover(rawTitle) {
  const q = searchableTitle(rawTitle);
  const params = new URLSearchParams({ q, limit: '5', fields: 'title,cover_edition_key,edition_key,cover_i' });
  const url = `${SEARCH_API}?${params}`;
  const res = await fetch(url, { headers: { Accept: 'application/json' } });
  if (res.status === 429) {
    console.warn('  ! rate-limited by Open Library (429). Wait and re-run; skipping rest.');
    return { rateLimited: true };
  }
  if (!res.ok) throw new Error(`search ${res.status} ${res.statusText}`);
  const data = await res.json();
  const docs = data?.docs ?? [];
  // Prefer a doc with a cover_edition_key (best cover source).
  const doc =
    docs.find((d) => d.cover_edition_key) ?? docs.find((d) => d.cover_i) ?? docs[0];
  if (!doc) return null;

  if (doc.cover_edition_key) {
    return { url: `${COVERS_API}/olid/${doc.cover_edition_key}-L.jpg`, name: doc.title };
  }
  if (doc.cover_i) {
    return { url: `${COVERS_API}/id/${doc.cover_i}-L.jpg`, name: doc.title };
  }
  return null;
}

// --- main -------------------------------------------------------------------
let items;
try {
  items = JSON.parse(await readFile(DATA_FILE, 'utf8'));
} catch (err) {
  if (err?.code === 'ENOENT') {
    console.error(`✖ ${DATA_FILE} not found.`);
    process.exit(1);
  }
  throw err;
}

const favorites = items.filter((i) => i.favorite);
console.log(
  `→ ${favorites.length} reading favorite(s) in reading.json` +
    (force ? ' (force re-fetch)' : ''),
);

await mkdir(OUT_DIR, { recursive: true });
let changed = false;

for (let i = 0; i < favorites.length; i++) {
  const item = favorites[i];
  if (item.image && !force) {
    console.log(`  ✓ ${item.title} — already has image, skipping.`);
    continue;
  }
  if (i > 0) await sleep(sleepMs);

  console.log(`→ [${i + 1}/${favorites.length}] ${item.title}`);
  let result;
  try {
    result = await fetchCover(item.title);
  } catch (err) {
    console.warn(`  ! network error: ${err?.message ?? err}; skipping.`);
    continue;
  }
  if (result?.rateLimited) break;
  if (!result?.url) {
    console.warn(`  ! no cover found for "${item.title}"; skipping.`);
    continue;
  }

  const fileName = `${slugify(item.title)}.jpg`;
  let buf;
  try {
    const imgRes = await fetch(result.url);
    if (!imgRes.ok) throw new Error(`${imgRes.status} ${imgRes.statusText}`);
    buf = Buffer.from(await imgRes.arrayBuffer());
    // Open Library returns a 1x1 transparent placeholder when no cover exists.
    if (buf.length < 1024) {
      console.warn(`  ! Open Library has no real cover for "${item.title}"; skipping.`);
      continue;
    }
  } catch (err) {
    console.warn(`  ! image download failed (${err?.message ?? err}); skipping.`);
    continue;
  }
  await writeFile(join(OUT_DIR, fileName), buf);
  item.image = `${PUBLIC_PREFIX}/${fileName}`;
  const orig = items.find((o) => o.title === item.title && o.favorite);
  if (orig) orig.image = item.image;
  changed = true;
  console.log(`  ✔ ${item.title} → ${item.image}`);
}

if (changed) {
  await writeFile(DATA_FILE, JSON.stringify(items, null, 2) + '\n', 'utf8');
  console.log(`\n✔ Updated ${DATA_FILE}`);
} else {
  console.log('\n— Nothing to update (all reading favorites already had images).');
}