printhive/lib/flaresolverr.js

// Minimal FlareSolverr client. FlareSolverr runs a headless browser that solves
// Cloudflare's anti-bot challenge and returns the resolved page HTML.
// https://github.com/FlareSolverr/FlareSolverr

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

// opts: { maxTimeout, cookies: [{ name, value, domain }], retries }
async function solve(flaresolverrUrl, targetUrl, opts = {}) {
  const endpoint = flaresolverrUrl.replace(/\/+$/, '') + '/v1';
  const body = { cmd: 'request.get', url: targetUrl, maxTimeout: opts.maxTimeout || 60000 };
  if (opts.cookies) body.cookies = opts.cookies;

  // FlareSolverr serializes browser sessions and returns 500 on back-to-back
  // requests, so retry transient failures with a growing backoff.
  const retries = opts.retries ?? 4;
  let lastErr;
  for (let attempt = 0; attempt <= retries; attempt++) {
    if (attempt > 0) await sleep(2000 * attempt);
    try {
      const res = await fetch(endpoint, {
        method: 'POST',
        headers: { 'Content-Type': 'application/json' },
        body: JSON.stringify(body),
      });
      if (res.status >= 500) { lastErr = new Error(`FlareSolverr HTTP ${res.status}`); continue; }
      if (!res.ok) throw new Error(`FlareSolverr HTTP ${res.status}`);
      const data = await res.json();
      if (data.status !== 'ok' || !data.solution) throw new Error(`FlareSolverr: ${data.message || 'no solution'}`);
      return data.solution.response;
    } catch (err) {
      lastErr = err;
      if (!/HTTP 5\d\d|fetch failed|ECONNRESET|network/i.test(err.message)) throw err;
    }
  }
  throw lastErr;
}

// Like solve(), but for endpoints that return JSON. A real browser wraps JSON
// in a <pre> tag, so strip tags before parsing.
async function solveJson(flaresolverrUrl, targetUrl, opts = {}) {
  const html = await solve(flaresolverrUrl, targetUrl, opts);
  // The browser wraps JSON in <pre> and HTML-encodes entities (notably & -> &amp;
  // inside signed URLs), so strip tags and decode before parsing.
  const text = html.replace(/<[^>]*>/g, '')
    .replace(/&quot;/g, '"').replace(/&#39;/g, "'").replace(/&lt;/g, '<')
    .replace(/&gt;/g, '>').replace(/&amp;/g, '&')
    .trim();
  try {
    return JSON.parse(text);
  } catch {
    throw new Error('FlareSolverr did not return parseable JSON.');
  }
}

module.exports = { solve, solveJson };