My Store: Add many books, with title, author, description and cover read out of each EPUB

Co-Authored-By: Claude Sonnet 5.5 <noreply@anthropic.com>

epubmeta.js

@@ -0,0 +1,46 @@
+// Reads the title, author, description and cover picture out of an EPUB (a zip with an OPF file inside), with no libraries.
+const zlib = require('zlib'), path = require('path');
+
+const MAX_PART = 8 * 1024 * 1024; // the biggest piece of an EPUB this reads (the OPF and the cover picture)
+const IMAGE = { jpg: 'image/jpeg', jpeg: 'image/jpeg', png: 'image/png', gif: 'image/gif', webp: 'image/webp' };
+
+/** The entries of a zip: name -> where to find it. */
+function entries(buf) {
+  let e = -1; for (let i = buf.length - 22; i >= Math.max(0, buf.length - 65557); i--) if (buf.readUInt32LE(i) === 0x06054b50) { e = i; break; }
+  if (e < 0) throw new Error('Not a zip');
+  const n = buf.readUInt16LE(e + 10); let p = buf.readUInt32LE(e + 16); const out = new Map();
+  for (let k = 0; k < n && p + 46 <= buf.length && buf.readUInt32LE(p) === 0x02014b50; k++) {
+    const method = buf.readUInt16LE(p + 10), csize = buf.readUInt32LE(p + 20), usize = buf.readUInt32LE(p + 24), nl = buf.readUInt16LE(p + 28), xl = buf.readUInt16LE(p + 30), cl = buf.readUInt16LE(p + 32), off = buf.readUInt32LE(p + 42);
+    out.set(buf.toString('utf8', p + 46, p + 46 + nl), { method, csize, usize, off }); p += 46 + nl + xl + cl;
+  }
+  return out;
+}
+function read(buf, ent) {
+  if (!ent || ent.usize > MAX_PART || buf.readUInt32LE(ent.off) !== 0x04034b50) return null;
+  const s = ent.off + 30 + buf.readUInt16LE(ent.off + 26) + buf.readUInt16LE(ent.off + 28), raw = buf.subarray(s, s + ent.csize);
+  return ent.method === 0 ? raw : ent.method === 8 ? zlib.inflateRawSync(raw, { maxOutputLength: MAX_PART }) : null;
+}
+
+const unxml = s => String(s).replace(/<!\[CDATA\[([\s\S]*?)\]\]>/g, '$1').replace(/<[^>]+>/g, ' ').replace(/&(#x[0-9a-f]+|#\d+|amp|lt|gt|quot|apos);/gi, (_, c) =>
+  ({ amp: '&', lt: '<', gt: '>', quot: '"', apos: "'" })[c.toLowerCase()] || String.fromCodePoint(c[1].toLowerCase() === 'x' ? parseInt(c.slice(2), 16) : +c.slice(1))).replace(/\s+/g, ' ').trim();
+const tagsOf = (xml, name) => [...xml.matchAll(new RegExp(`<(?:\\w+:)?${name}\\b([^>]*?)(?:/>|>([\\s\\S]*?)</(?:\\w+:)?${name}>)`, 'gi'))].map(m => ({ attrs: m[1] || '', text: m[2] == null ? '' : unxml(m[2]) }));
+const attr = (attrs, name) => { const m = new RegExp(`\\b${name}\\s*=\\s*("([^"]*)"|'([^']*)')`, 'i').exec(attrs); return m ? (m[2] != null ? m[2] : m[3]) : ''; };
+const resolve = (base, href) => { try { href = decodeURIComponent(href); } catch {} return path.posix.normalize(path.posix.join(path.posix.dirname(base), href.split('#')[0])).replace(/^\/+/, ''); };
+
+/** { title, author, description, cover: { type, data } | null } (any part may be missing). Throws if the file is not a readable EPUB. */
+function readEpub(buf) {
+  const z = entries(buf), container = read(buf, z.get('META-INF/container.xml')); if (!container) throw new Error('No container.xml');
+  const opfPath = attr((tagsOf(container.toString('utf8'), 'rootfile')[0] || {}).attrs || '', 'full-path'); if (!opfPath) throw new Error('No package file');
+  const opf = (read(buf, z.get(opfPath)) || Buffer.alloc(0)).toString('utf8');
+  const meta = { title: (tagsOf(opf, 'title')[0] || {}).text || '', author: tagsOf(opf, 'creator').map(c => c.text).filter(Boolean).join(', '), description: ((tagsOf(opf, 'description')[0] || {}).text || '').slice(0, 2000), cover: null };
+  // the cover picture: the item the package calls the cover-image, else the one <meta name="cover"> points at, else an image named "cover"
+  const items = tagsOf(opf, 'item').map(i => ({ id: attr(i.attrs, 'id'), href: attr(i.attrs, 'href'), type: attr(i.attrs, 'media-type').toLowerCase(), props: attr(i.attrs, 'properties') }));
+  const metaCover = tagsOf(opf, 'meta').find(m => attr(m.attrs, 'name').toLowerCase() === 'cover'), cid = metaCover && attr(metaCover.attrs, 'content');
+  const pick = items.find(i => /\bcover-image\b/.test(i.props)) || items.find(i => cid && i.id === cid) || items.find(i => /^image\//.test(i.type) && /cover/i.test(i.id + ' ' + i.href));
+  if (pick) {
+    const data = read(buf, z.get(resolve(opfPath, pick.href))), ext = path.extname(pick.href).slice(1).toLowerCase(), type = /^image\/(jpeg|png|gif|webp)$/.test(pick.type) ? pick.type : IMAGE[ext];
+    if (data && type && data.length > 500) meta.cover = { type, data };
+  }
+  return meta;
+}
+module.exports = { readEpub };