n8n-flows/flows/paywall-bypass.json

{
  "updatedAt": "2026-05-27T23:47:30.991Z",
  "createdAt": "2026-05-27T22:09:54.846Z",
  "id": "JqX7ejCYX1oAHvQ2",
  "name": "paywall-bypass",
  "description": null,
  "active": true,
  "isArchived": false,
  "nodes": [
    {
      "id": "86fa097b-c230-484d-806d-45b421635c05",
      "name": "Every 5 Minutes",
      "type": "n8n-nodes-base.scheduleTrigger",
      "position": [
        -800,
        0
      ],
      "parameters": {
        "rule": {
          "interval": [
            {
              "field": "minutes",
              "minutesInterval": 5
            }
          ]
        }
      },
      "typeVersion": 1.2
    },
    {
      "id": "eb644570-cca3-4464-9b99-02084f199078",
      "name": "Get Miniflux Entries",
      "type": "n8n-nodes-base.httpRequest",
      "position": [
        -360,
        0
      ],
      "parameters": {
        "url": "http://192.168.1.40:17002/v1/entries",
        "method": "GET",
        "options": {},
        "sendQuery": true,
        "sendHeaders": true,
        "queryParameters": {
          "parameters": [
            {
              "name": "status",
              "value": "unread"
            },
            {
              "name": "limit",
              "value": "100"
            },
            {
              "name": "order",
              "value": "published_at"
            },
            {
              "name": "direction",
              "value": "desc"
            }
          ]
        },
        "headerParameters": {
          "parameters": [
            {
              "name": "X-Auth-Token",
              "value": "3f439f2db55967ed5acd83398214d2aa76124927cde07115d945b9282597fecb"
            }
          ]
        }
      },
      "typeVersion": 4.2
    },
    {
      "id": "f8024262-3163-4cd9-8233-3c91bfddcb01",
      "name": "Process Entries",
      "type": "n8n-nodes-base.code",
      "position": [
        -120,
        0
      ],
      "parameters": {
        "mode": "runOnceForAllItems",
        "jsCode": "const MINIFLUX_TOKEN = '3f439f2db55967ed5acd83398214d2aa76124927cde07115d945b9282597fecb';\nconst MINIFLUX_URL = 'http://192.168.1.40:17002';\n\nconst PAYWALL_NEWS = [\n  'spiegel.de','zeit.de','faz.net','sueddeutsche.de','handelsblatt.com',\n  'nytimes.com','wsj.com','ft.com','washingtonpost.com','thetimes.co.uk',\n  'wired.com','theatlantic.com','economist.com','bloomberg.com',\n  'hbr.org','foreignpolicy.com','newyorker.com','taz.de',\n  'telegraph.co.uk','businessinsider.com'\n];\nconst PAYWALL_SCIENCE = [\n  'nature.com','science.org','cell.com','nejm.org','thelancet.com',\n  'link.springer.com','springer.com','wiley.com','onlinelibrary.wiley.com',\n  'sciencedirect.com','pubs.acs.org','rsc.org','pubs.rsc.org',\n  'bmj.com','annualreviews.org','pnas.org',\n  'jamanetwork.com','ahajournals.org','tandfonline.com','academic.oup.com'\n];\n\nconst CONTENT_MIN = 1500;\nconst RETRY_LIMIT_MS = 3 * 60 * 60 * 1000;\n\nconst ALREADY_PROCESSED = /[\\u{1F513}\\u{1F4D6}\\u{1F4F0}\\u{1F6AB}\\u{274C}]$/u;\nconst ANY_STATUS = /\\s[\\u{1F513}\\u{1F4D6}\\u{1F4F0}\\u{1F6AB}\\u{274C}⏳]$/u;\n\nconst PAYWALL_MARKERS = [\n  'Sie können den Artikel leider nicht mehr aufrufen',\n  'Diesen Artikel weiterlesen mit',\n  'SPIEGEL plus',\n  'Jetzt abonnieren',\n  'Bereits Abonnent',\n  'Subscribe to read',\n  'Become a subscriber',\n  'To continue reading',\n  'Get full access',\n  'um diesen Artikel zu lesen',\n  'Mit Digital-Abo',\n];\n\nfunction hostname(url) {\n  const m = String(url || '').match(/^[a-z]+:\\/\\/([^/?#:]+)/i);\n  if (!m) return '';\n  return m[1].toLowerCase().replace(/^www\\./, '');\n}\n\nfunction isPaywalled(text) {\n  if (!text) return true;\n  const clean = text.replace(/<[^>]+>/g, '').trim();\n  if (clean.length < CONTENT_MIN) return true;\n  return PAYWALL_MARKERS.some(m => text.includes(m));\n}\n\nfunction extractDoi(url) {\n  let m = url.match(/doi\\.org\\/(10\\.\\d{4,}\\/[^\\s?#]+)/);\n  if (m) return decodeURIComponent(m[1]);\n  m = url.match(/\\/doi\\/(10\\.\\d{4,}\\/[^\\s?#]+)/);\n  if (m) return decodeURIComponent(m[1].replace(/[?#].*$/, ''));\n  return null;\n}\n\nfunction badge(source, type, status, url) {\n  const archiveLink = url ? '<a href=\"https://archive.ph/newest/' + encodeURI(url) + '\" target=\"_blank\" rel=\"noopener\">Check archive.ph manually →</a>' : '';\n  if (status === 'paywalled') return '<p style=\"font-size:0.85em;color:#555;border-left:3px solid #9E9E9E;padding:6px 10px;margin:0 0 1em;background:#fafafa\">⏳ Paywall detected — will retry archive.ph<br>' + archiveLink + '</p>';\n  if (status === 'giveup') return '<p style=\"font-size:0.85em;color:#555;border-left:3px solid #757575;padding:6px 10px;margin:0 0 1em;background:#fafafa\">🚫 Paywall — gave up after 3h<br>' + archiveLink + '</p>';\n  if (source === 'unpaywall') return '<p style=\"font-size:0.8em;color:#555;border-left:3px solid #4CAF50;padding:3px 8px;margin:0 0 1em\">🔓 Open access via Unpaywall</p>';\n  if (source === 'free') return '<p style=\"font-size:0.8em;color:#555;border-left:3px solid #4CAF50;padding:3px 8px;margin:0 0 1em\">📰 Free article — full text via crawl4ai</p>';\n  if (source === 'archive') return '<p style=\"font-size:0.8em;color:#555;border-left:3px solid #2196F3;padding:3px 8px;margin:0 0 1em\">📖 Full text via archive.ph</p>';\n  if (type === 'science') return '<p style=\"font-size:0.8em;color:#555;border-left:3px solid #2196F3;padding:3px 8px;margin:0 0 1em\">📖 Full text via crawl4ai (not OA)</p>';\n  return '<p style=\"font-size:0.8em;color:#555;border-left:3px solid #FF9800;padding:3px 8px;margin:0 0 1em\">📖 Paywall bypass via crawl4ai</p>';\n}\n\nfunction mdToHtml(md) {\n  if (!md) return '';\n  const lines = md.split('\\n').filter(l => {\n    const t = l.trim();\n    if (!t) return true;\n    if (/Zur Merkliste hinzufügen|Artikel anhören|Bild vergrößern|Bild schließen|Link kopieren|Weitere Optionen zum Teilen|Mehr lesen über|Verwandte Artikel/.test(t)) return false;\n    if (/^\\s*\\*\\s*\\[?\\s*(X\\.com|Facebook|Messenger|WhatsApp|E-Mail|Threads|Mastodon|Telegram)\\b/i.test(t)) return false;\n    return true;\n  });\n  let s = lines.join('\\n');\n  s = s.replace(/!\\[([^\\]]*)\\]\\(([^)\\s]+)[^)]*\\)/g, '<img alt=\"$1\" src=\"$2\" style=\"max-width:100%;height:auto\">');\n  s = s.replace(/\\[([^\\]]+)\\]\\(([^)\\s]+)[^)]*\\)/g, '<a href=\"$2\">$1</a>');\n  s = s.replace(/\\*\\*([^*\\n]+)\\*\\*/g, '<strong>$1</strong>');\n  s = s.replace(/`([^`\\n]+)`/g, '<code>$1</code>');\n  const out = [];\n  let inList = false, inPara = false;\n  for (const raw of s.split('\\n')) {\n    const line = raw.trim();\n    if (!line) {\n      if (inList) { out.push('</ul>'); inList = false; }\n      if (inPara) { out.push('</p>'); inPara = false; }\n      continue;\n    }\n    const h = line.match(/^(#{1,6})\\s+(.+)$/);\n    if (h) {\n      if (inList) { out.push('</ul>'); inList = false; }\n      if (inPara) { out.push('</p>'); inPara = false; }\n      const n = h[1].length;\n      out.push(`<h${n}>${h[2]}</h${n}>`);\n      continue;\n    }\n    const li = line.match(/^[*\\-]\\s+(.+)$/);\n    if (li) {\n      if (inPara) { out.push('</p>'); inPara = false; }\n      if (!inList) { out.push('<ul>'); inList = true; }\n      out.push(`<li>${li[1]}</li>`);\n      continue;\n    }\n    if (inList) { out.push('</ul>'); inList = false; }\n    if (!inPara) { out.push('<p>'); inPara = true; }\n    out.push(line);\n  }\n  if (inList) out.push('</ul>');\n  if (inPara) out.push('</p>');\n  return out.join('\\n');\n}\n\nconst self = this;\nasync function fetchCrawl4ai(url) {\n  try {\n    const cr = await self.helpers.httpRequest({\n      method: 'POST',\n      url: 'http://mcp-crawl4ai:11235/md',\n      headers: { 'Content-Type': 'application/json' },\n      body: { url, filter: 'fit' },\n      json: true,\n      timeout: 60000,\n    });\n    if (cr && cr.success && cr.markdown) return cr.markdown;\n  } catch (_) {}\n  return '';\n}\n\nasync function fetchUnpaywall(doi) {\n  try {\n    const up = await self.helpers.httpRequest({\n      method: 'GET',\n      url: `https://api.unpaywall.org/v2/${encodeURIComponent(doi)}?email=fkrebs@nucli.de`,\n      json: true,\n      timeout: 15000,\n    });\n    if (up.is_oa && up.best_oa_location) {\n      return up.best_oa_location.url || up.best_oa_location.url_for_pdf || '';\n    }\n  } catch (_) {}\n  return '';\n}\n\nasync function getArchiveSnapshotUrl(url) {\n  try {\n    const tm = await self.helpers.httpRequest({\n      method: 'GET',\n      url: 'https://archive.ph/timemap/' + url,\n      headers: { 'User-Agent': 'Mozilla/5.0 (X11; Linux x86_64) AppleWebKit/537.36 Chrome/120.0' },\n      timeout: 30000,\n    });\n    const txt = typeof tm === 'string' ? tm : (tm && tm.body) || '';\n    const matches = [...txt.matchAll(/<([^>]+)>;\\s*rel=\"[^\"]*memento[^\"]*\"/g)];\n    if (matches.length > 0) return matches[matches.length - 1][1];\n  } catch (_) {}\n  return '';\n}\n\nconst entries = $('Get Miniflux Entries').first().json.entries || [];\nconst results = [];\nconst stats = { skipped_done: 0, skipped_nopaywall: 0, attempted: 0, success: 0, retry: 0, giveup: 0, error: 0 };\n\nfor (const e of entries) {\n  if (ALREADY_PROCESSED.test(e.title)) { stats.skipped_done++; continue; }\n\n  const host = hostname(e.url);\n  const isScience = PAYWALL_SCIENCE.some(d => host === d || host.endsWith('.' + d));\n  const isNews = PAYWALL_NEWS.some(d => host === d || host.endsWith('.' + d));\n  const isShort = (e.content || '').replace(/<[^>]+>/g, '').trim().length < 800;\n\n  if (!isScience && !isNews && !isShort) { stats.skipped_nopaywall++; continue; }\n  stats.attempted++;\n\n  const type = isScience ? 'science' : isNews ? 'news' : 'short';\n  const publishedMs = new Date(e.published_at).getTime();\n  const ageMs = Date.now() - publishedMs;\n  const exhausted = ageMs >= RETRY_LIMIT_MS;\n\n  let markdown = '';\n  let source = '';\n\n  if (isScience) {\n    const doi = extractDoi(e.url);\n    if (doi) {\n      const oaUrl = await fetchUnpaywall(doi);\n      if (oaUrl) {\n        const md = await fetchCrawl4ai(oaUrl);\n        if (md && !isPaywalled(md)) { markdown = md; source = 'unpaywall'; }\n      }\n    }\n  }\n\n  // Try crawl4ai direct first to find out if it's actually paywalled\n  if (!source) {\n    const md = await fetchCrawl4ai(e.url);\n    if (md) {\n      if (!isPaywalled(md)) {\n        markdown = md;\n        source = 'free';\n      } else {\n        markdown = md; // keep for retry/giveup display\n      }\n    }\n  }\n\n  // If paywalled (or no content), try archive.ph for bypass\n  if (!source) {\n    const snapUrl = await getArchiveSnapshotUrl(e.url);\n    if (snapUrl) {\n      const md = await fetchCrawl4ai(snapUrl);\n      if (md && !isPaywalled(md)) { markdown = md; source = 'archive'; }\n    }\n  }\n\n  const cleanTitle = e.title.replace(ANY_STATUS, '');\n  let newTitle, content, status;\n\n  if (source) {\n    content = badge(source, type, 'ok') + mdToHtml(markdown);\n    const emoji = source === 'unpaywall' ? '🔓' : source === 'free' ? '📰' : '📖';\n    newTitle = cleanTitle + ' ' + emoji;\n    stats.success++;\n    status = 'updated';\n  } else if (markdown && exhausted) {\n    content = badge('', type, 'giveup', e.url) + mdToHtml(markdown);\n    newTitle = cleanTitle + ' 🚫';\n    stats.giveup++;\n    status = 'gaveup';\n  } else if (markdown) {\n    content = badge('', type, 'paywalled', e.url) + mdToHtml(markdown);\n    newTitle = cleanTitle + ' ⏳';\n    stats.retry++;\n    status = 'retry';\n  } else {\n    content = '<p style=\"font-size:0.85em;color:#555;border-left:3px solid #c62828;padding:6px 10px;margin:0 0 1em;background:#fafafa\">❌ Fetch error<br><a href=\"https://archive.ph/newest/' + encodeURI(e.url) + '\" target=\"_blank\" rel=\"noopener\">Check archive.ph manually →</a></p>';\n    newTitle = cleanTitle + ' ❌';\n    stats.error++;\n    status = 'error';\n  }\n\n  const body = content !== null ? { content, title: newTitle } : { title: newTitle };\n  try {\n    await self.helpers.httpRequest({\n      method: 'PUT',\n      url: `${MINIFLUX_URL}/v1/entries/${e.id}`,\n      headers: { 'X-Auth-Token': MINIFLUX_TOKEN, 'Content-Type': 'application/json' },\n      body,\n      json: true,\n      timeout: 15000,\n    });\n    results.push({ json: { entry_id: e.id, title: newTitle.slice(0, 80), status, source, type, host } });\n  } catch (err) {\n    results.push({ json: { entry_id: e.id, title: e.title.slice(0, 80), status: 'put_failed', error: String(err).slice(0, 200) } });\n  }\n}\n\nresults.push({ json: { stats } });\nreturn results;"
      },
      "typeVersion": 2
    }
  ],
  "connections": {
    "Every 5 Minutes": {
      "main": [
        [
          {
            "node": "Get Miniflux Entries",
            "type": "main",
            "index": 0
          }
        ]
      ]
    },
    "Get Miniflux Entries": {
      "main": [
        [
          {
            "node": "Process Entries",
            "type": "main",
            "index": 0
          }
        ]
      ]
    }
  },
  "settings": {
    "executionOrder": "v1"
  },
  "staticData": {
    "node:Every 5 Minutes": {
      "recurrenceRules": []
    }
  },
  "meta": null,
  "pinData": null,
  "versionId": "0a3f5003-653c-4494-96e9-a779815a1b9c",
  "activeVersionId": "0a3f5003-653c-4494-96e9-a779815a1b9c",
  "versionCounter": 17,
  "triggerCount": 1,
  "tags": [],
  "shared": [
    {
      "updatedAt": "2026-05-27T22:09:54.846Z",
      "createdAt": "2026-05-27T22:09:54.846Z",
      "role": "workflow:owner",
      "workflowId": "JqX7ejCYX1oAHvQ2",
      "projectId": "hkIaTyp1CKVS8ywT",
      "project": {
        "updatedAt": "2026-03-09T07:21:36.510Z",
        "createdAt": "2026-03-09T07:20:59.502Z",
        "id": "hkIaTyp1CKVS8ywT",
        "name": "Florian Krebs <fkrebs@nucli.de>",
        "type": "personal",
        "icon": null,
        "description": null,
        "creatorId": "8af5c813-f89c-4b83-bcf8-5d59864a038a"
      }
    }
  ],
  "versionMetadata": {
    "name": "paywall-bypass",
    "description": null
  }
}