feat: iframe CAPTCHA-Flow, PDF von archive.ph Snapshot-URL
This commit is contained in:
@@ -6,48 +6,29 @@ const path = require('path');
|
||||
const fs = require('fs');
|
||||
const crypto = require('crypto');
|
||||
|
||||
// ── Einfacher HTTPS GET Helper ─────────────────────────────────────────────
|
||||
function httpsGet(url, timeoutMs = 20000) {
|
||||
return new Promise((resolve, reject) => {
|
||||
const req = https.get(url, {
|
||||
headers: {
|
||||
'User-Agent': 'Mozilla/5.0 (compatible; DickenDock/1.0)',
|
||||
'Accept': '*/*',
|
||||
},
|
||||
timeout: timeoutMs,
|
||||
}, (res) => {
|
||||
let body = '';
|
||||
res.setEncoding('utf8');
|
||||
res.on('data', d => { if (body.length < 200000) body += d; });
|
||||
res.on('end', () => resolve({ status: res.statusCode, body, headers: res.headers }));
|
||||
res.on('error', reject);
|
||||
});
|
||||
req.on('error', reject);
|
||||
req.on('timeout', () => { req.destroy(); reject(new Error('Timeout')); });
|
||||
});
|
||||
}
|
||||
|
||||
// ── GET /api/tools/paywallkiller/cdx?url=... ──────────────────────────────
|
||||
// Proxy für Wayback CDX API – löst CORS/Origin-Probleme der PWA
|
||||
// Wayback CDX Proxy (für spätere Verwendung)
|
||||
router.get('/cdx', authenticate, async (req, res) => {
|
||||
const { url } = req.query;
|
||||
if (!url) return res.status(400).json({ error: 'url fehlt' });
|
||||
|
||||
let targetUrl = url.trim();
|
||||
if (!targetUrl.startsWith('http')) targetUrl = 'https://' + targetUrl;
|
||||
|
||||
const cdxUrl = `https://web.archive.org/cdx/search/cdx?url=${encodeURIComponent(targetUrl)}&output=json&limit=1&fl=timestamp&filter=statuscode:200&fastLatest=true`;
|
||||
|
||||
try {
|
||||
const result = await httpsGet(cdxUrl);
|
||||
if (result.status !== 200) return res.status(502).json({ error: `CDX Status: ${result.status}` });
|
||||
|
||||
let data;
|
||||
try { data = JSON.parse(result.body); } catch { return res.status(502).json({ error: 'Ungültige CDX-Antwort' }); }
|
||||
|
||||
if (!Array.isArray(data) || data.length < 2 || !data[1]?.[0]) {
|
||||
return res.json({ found: false });
|
||||
}
|
||||
const result = await new Promise((resolve, reject) => {
|
||||
const req2 = https.get(cdxUrl, { headers: { 'User-Agent': 'Mozilla/5.0' }, timeout: 20000 }, (res2) => {
|
||||
let body = '';
|
||||
res2.setEncoding('utf8');
|
||||
res2.on('data', d => body += d);
|
||||
res2.on('end', () => resolve({ status: res2.statusCode, body }));
|
||||
res2.on('error', reject);
|
||||
});
|
||||
req2.on('error', reject);
|
||||
req2.on('timeout', () => { req2.destroy(); reject(new Error('Timeout')); });
|
||||
});
|
||||
if (result.status !== 200) return res.status(502).json({ error: `CDX ${result.status}` });
|
||||
const data = JSON.parse(result.body);
|
||||
if (!Array.isArray(data) || data.length < 2 || !data[1]?.[0]) return res.json({ found: false });
|
||||
return res.json({ found: true, waybackUrl: `https://web.archive.org/web/${data[1][0]}/${targetUrl}` });
|
||||
} catch (err) {
|
||||
return res.status(500).json({ error: err.message });
|
||||
@@ -55,13 +36,11 @@ router.get('/cdx', authenticate, async (req, res) => {
|
||||
});
|
||||
|
||||
// ── POST /api/tools/paywallkiller/pdf ─────────────────────────────────────
|
||||
// Lädt via 12ft.io (Paywall-Bypasser, kein CAPTCHA) → PDF
|
||||
// Lädt archive.ph Snapshot-URL mit Puppeteer → PDF
|
||||
// archive.ph Snapshot-URLs (hash-basiert) benötigen kein CAPTCHA
|
||||
router.post('/pdf', authenticate, async (req, res) => {
|
||||
const { originalUrl, archiveUrl } = req.body;
|
||||
if (!originalUrl && !archiveUrl) return res.status(400).json({ error: 'URL fehlt' });
|
||||
// 12ft.io umgeht Paywalls indem es die Googlebot-Version abruft
|
||||
const targetUrl = `https://12ft.io/proxy?q=${encodeURIComponent(originalUrl || archiveUrl)}`;
|
||||
if (!targetUrl || typeof targetUrl !== 'string') return res.status(400).json({ error: 'URL fehlt' });
|
||||
const { archiveUrl } = req.body;
|
||||
if (!archiveUrl || typeof archiveUrl !== 'string') return res.status(400).json({ error: 'archiveUrl fehlt' });
|
||||
|
||||
const tmpFile = path.join('/tmp', `pwk_${crypto.randomBytes(8).toString('hex')}.pdf`);
|
||||
let browser = null;
|
||||
@@ -80,50 +59,46 @@ router.post('/pdf', authenticate, async (req, res) => {
|
||||
browser = await puppeteer.launch({
|
||||
executablePath,
|
||||
headless: true,
|
||||
args: ['--no-sandbox', '--disable-setuid-sandbox', '--disable-dev-shm-usage', '--disable-gpu', '--single-process', '--no-zygote'],
|
||||
args: ['--no-sandbox', '--disable-setuid-sandbox', '--disable-dev-shm-usage',
|
||||
'--disable-gpu', '--single-process', '--no-zygote'],
|
||||
timeout: 30000,
|
||||
});
|
||||
|
||||
const page = await browser.newPage();
|
||||
await page.setUserAgent('Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36');
|
||||
await page.setViewport({ width: 1280, height: 900 });
|
||||
await page.goto(targetUrl, { waitUntil: 'networkidle0', timeout: 30000 });
|
||||
|
||||
// UI-Störelemente entfernen (Wayback-Toolbar, Cookie-Banner, Werbung)
|
||||
// Der Artikel-Text ist im Wayback-Snapshot vollständig vorhanden
|
||||
await page.evaluate(() => {
|
||||
const remove = [
|
||||
// Wayback Machine Toolbar
|
||||
'#wm-ipp-base', '#wm-ipp', '#donato', '#wm-ipp-print',
|
||||
// Cookie/Consent Banner
|
||||
'#usercentrics-root', '.uc-banner', '#usercentrics-cmp-ui',
|
||||
'#onetrust-consent-sdk', '.onetrust-pc-dark-filter',
|
||||
'[id*="cookie"]', '[class*="cookie-banner"]', '[class*="cookieBanner"]',
|
||||
'[class*="consent-banner"]', '[id*="consent-banner"]',
|
||||
// Sticky Header/Footer
|
||||
'header[class*="sticky"]', '[class*="sticky-header"]',
|
||||
'[class*="stickyHeader"]', '[class*="fixed-header"]',
|
||||
// Werbung
|
||||
'[class*="advertisement"]', '[class*="adslot"]', '[id^="div-gpt-ad"]',
|
||||
// Newsletter-Popup, Abo-Hinweise (nicht die Paywall selbst)
|
||||
'[class*="newsletter"]', '[class*="teaser-abo"]',
|
||||
];
|
||||
remove.forEach(sel => {
|
||||
try { document.querySelectorAll(sel).forEach(el => el.remove()); } catch {}
|
||||
});
|
||||
// archive.ph laden – Snapshot-URL direkt (kein CAPTCHA bei direkten Snapshot-URLs)
|
||||
await page.goto(archiveUrl, { waitUntil: 'networkidle0', timeout: 30000 });
|
||||
|
||||
// body overflow freigeben
|
||||
document.body.style.overflow = 'visible';
|
||||
document.documentElement.style.overflow = 'visible';
|
||||
document.body.style.height = 'auto';
|
||||
// Prüfen ob CAPTCHA erschienen ist
|
||||
const hasCaptcha = await page.evaluate(() => {
|
||||
return document.title.includes('One more step') ||
|
||||
document.body.innerText.includes('complete a CAPTCHA') ||
|
||||
document.body.innerText.includes('security check');
|
||||
});
|
||||
|
||||
await new Promise(r => setTimeout(r, 300));
|
||||
if (hasCaptcha) {
|
||||
await browser.close(); browser = null;
|
||||
return res.status(403).json({ error: 'CAPTCHA – bitte im Browser das Archiv öffnen und CAPTCHA lösen, dann erneut versuchen.' });
|
||||
}
|
||||
|
||||
// UI-Störelemente entfernen
|
||||
await page.evaluate(() => {
|
||||
['#wm-ipp-base','#wm-ipp','#donato',
|
||||
'#usercentrics-root','.uc-banner',
|
||||
'#onetrust-consent-sdk','.onetrust-pc-dark-filter',
|
||||
'[id*="cookie"]','[class*="cookie-banner"]'].forEach(sel => {
|
||||
try { document.querySelectorAll(sel).forEach(el => el.remove()); } catch {}
|
||||
});
|
||||
document.body.style.overflow = 'visible';
|
||||
});
|
||||
|
||||
await new Promise(r => setTimeout(r, 500));
|
||||
|
||||
await page.pdf({
|
||||
path: tmpFile, format: 'A4', printBackground: true,
|
||||
printBackground: true,
|
||||
margin: { top: '12mm', bottom: '12mm', left: '12mm', right: '12mm' },
|
||||
margin: { top: '10mm', bottom: '10mm', left: '10mm', right: '10mm' },
|
||||
});
|
||||
await browser.close(); browser = null;
|
||||
|
||||
@@ -136,7 +111,7 @@ router.post('/pdf', authenticate, async (req, res) => {
|
||||
} catch (err) {
|
||||
if (browser) try { await browser.close(); } catch {}
|
||||
fs.unlink(tmpFile, () => {});
|
||||
return res.status(500).json({ error: 'PDF-Erstellung fehlgeschlagen: ' + err.message });
|
||||
return res.status(500).json({ error: 'PDF fehlgeschlagen: ' + err.message });
|
||||
}
|
||||
});
|
||||
|
||||
|
||||
Reference in New Issue
Block a user