Void-Homelab/lib/icons/ingest.js

// lib/icons/ingest.js
import path from 'node:path';
import dns from 'node:dns';
import AdmZip from 'adm-zip';
import { sanitizeSvg } from './sanitize.js';

export const MAX_FILE = 256 * 1024;          // 256 KB per icon
export const MAX_ZIP_ENTRIES = 200;
export const MAX_ZIP_TOTAL = 5 * 1024 * 1024; // 5 MB uncompressed
export const MAX_URL_BYTES = 5 * 1024 * 1024;

const EXT = { '.svg': 'image/svg+xml', '.png': 'image/png' };
const PNG_SIG = [0x89,0x50,0x4e,0x47];

function slugBase(name) {
  return path.basename(name, path.extname(name)).toLowerCase().replace(/[^a-z0-9]+/g, '-').replace(/(^-|-$)/g, '');
}
function magicOk(ext, buf) {
  if (ext === '.png') return PNG_SIG.every((b, i) => buf[i] === b);
  if (ext === '.svg') return buf.toString('utf8', 0, 400).includes('<svg');
  return false;
}

// Validate + normalize one icon. Returns { name, buffer, ext, contentType }. Throws on invalid.
export function processFile({ name, buffer }) {
  const ext = path.extname(name).toLowerCase();
  if (!EXT[ext]) throw new Error('unsupported_type');
  if (!buffer || buffer.length === 0) throw new Error('empty');
  if (buffer.length > MAX_FILE) throw new Error('too_large');
  if (!magicOk(ext, buffer)) throw new Error('bad_magic');
  const base = slugBase(name);
  if (!base) throw new Error('bad_name');
  const out = ext === '.svg' ? Buffer.from(sanitizeSvg(buffer)) : buffer;
  return { name: `${base}${ext}`, buffer: out, ext, contentType: EXT[ext] };
}

// Extract image entries from a zip buffer; flatten basenames, skip traversal/junk.
export function unpackZip(buffer) {
  const zip = new AdmZip(buffer);
  const entries = zip.getEntries();
  if (entries.length > MAX_ZIP_ENTRIES) throw new Error('too_many_entries');
  const out = []; let total = 0;
  for (const e of entries) {
    if (e.isDirectory) continue;
    const ext = path.extname(e.entryName).toLowerCase();
    if (!EXT[ext]) continue;                              // skip non-images
    if (/(^|[\\/])\.\.([\\/]|$)/.test(e.entryName)) continue; // skip traversal
    const data = e.getData();
    total += data.length;
    if (total > MAX_ZIP_TOTAL) throw new Error('zip_too_big');
    try { out.push(processFile({ name: path.basename(e.entryName), buffer: data })); }
    catch { /* skip individually-invalid entries */ }
  }
  return out;
}

const PRIVATE_HOST = /^(localhost|127\.|0\.0\.0\.0|10\.|192\.168\.|169\.254\.|172\.(1[6-9]|2\d|3[01])\.|\[?::1\]?)/i;

/**
 * Returns true if the given IP string is a blocked (loopback, private,
 * link-local, or ULA) address that should not be fetched.
 * Handles both IPv4 and IPv6.
 */
export function isBlockedAddress(ip) {
  if (!ip) return true;
  // IPv6 loopback
  if (ip === '::1') return true;
  // IPv4-mapped loopback ::ffff:127.x.x.x
  const v4mapped = ip.match(/^::ffff:(\d+\.\d+\.\d+\.\d+)$/i);
  const v4 = v4mapped ? v4mapped[1] : ip;

  if (/^\d+\.\d+\.\d+\.\d+$/.test(v4)) {
    const parts = v4.split('.').map(Number);
    const [a, b] = parts;
    // 0.0.0.0
    if (a === 0) return true;
    // 127.0.0.0/8 — loopback
    if (a === 127) return true;
    // 10.0.0.0/8 — private
    if (a === 10) return true;
    // 172.16.0.0/12 — private
    if (a === 172 && b >= 16 && b <= 31) return true;
    // 192.168.0.0/16 — private
    if (a === 192 && b === 168) return true;
    // 169.254.0.0/16 — link-local
    if (a === 169 && b === 254) return true;
    return false;
  }

  // IPv6 checks (expand to lower-case for prefix matching)
  const lower = ip.toLowerCase();
  // ::1 is caught above; handle full-form loopback
  if (lower === '0:0:0:0:0:0:0:1') return true;
  // fe80::/10 — link-local (fe80 – febf)
  if (/^fe[89ab][0-9a-f]:/i.test(lower)) return true;
  // fc00::/7 — ULA (fc00 – fdff)
  if (/^f[cd][0-9a-f]{2}:/i.test(lower)) return true;

  return false;
}

function dnsLookupAll(hostname) {
  return new Promise((resolve, reject) =>
    dns.lookup(hostname, { all: true }, (err, addrs) => err ? reject(err) : resolve(addrs))
  );
}

// Fetch a remote icon or zip. SSRF guard: http/https only, no localhost/private,
// DNS-resolved address check, size + timeout caps. `fetcher` injectable for tests.
export async function fetchUrl(url, { fetcher } = {}) {
  let u;
  try { u = new URL(url); } catch { throw new Error('bad_url'); }
  if (u.protocol !== 'http:' && u.protocol !== 'https:') throw new Error('bad_scheme');
  // Fast literal-hostname guard (catches raw IP strings and 'localhost' without DNS)
  if (PRIVATE_HOST.test(u.hostname)) throw new Error('blocked_host');
  // DNS resolution guard — only when using the real fetcher (not in tests)
  if (!fetcher) {
    const addrs = await dnsLookupAll(u.hostname).catch(() => []);
    if (addrs.some(a => isBlockedAddress(a.address))) throw new Error('blocked_host');
  }
  const realFetcher = fetcher ?? fetch;
  const res = await realFetcher(url, { signal: AbortSignal.timeout(8000), redirect: 'error' });
  if (!res.ok) throw new Error('fetch_failed');
  const ab = await res.arrayBuffer();
  if (ab.byteLength > MAX_URL_BYTES) throw new Error('too_large');
  return { buffer: Buffer.from(ab) };
}

export function isZip(buf) { return buf && buf.length > 4 && buf[0] === 0x50 && buf[1] === 0x4b; }