tractatus/scripts/check-missing-pdfs.js

#!/usr/bin/env node
/**
 * Check which of the 34 public documents are missing PDFs
 */
require('dotenv').config();

const { connect, close } = require('../src/utils/db.util');
const Document = require('../src/models/Document.model');
const fs = require('fs').promises;
const path = require('path');

const PUBLIC_CATEGORIES = [
  'getting-started',
  'technical-reference',
  'research-theory',
  'advanced-topics',
  'case-studies',
  'business-leadership',
  'archives'
];

async function main() {
  await connect();

  const { MongoClient } = require('mongodb');
  const mongoose = require('mongoose');
  const { getCollection } = require('../src/utils/db.util');

  const collection = await getCollection('documents');

  const docs = await collection.find({
    category: { $in: PUBLIC_CATEGORIES }
  }).sort({ category: 1, order: 1 }).toArray();

  console.log('\n' + '='.repeat(70));
  console.log('PDF AVAILABILITY CHECK - 34 Public Documents');
  console.log('='.repeat(70) + '\n');

  const pdfDir = '/home/theflow/projects/tractatus/public/downloads';

  let missingCount = 0;
  let presentCount = 0;
  const missing = [];

  for (const doc of docs) {
    const slug = doc.slug;
    const pdfPath = path.join(pdfDir, `${slug}.pdf`);

    try {
      await fs.access(pdfPath);
      presentCount++;
      // console.log(`✅ ${doc.title}`);
    } catch (err) {
      missingCount++;
      missing.push({ title: doc.title, slug, category: doc.category });
      console.log(`❌ MISSING PDF: ${doc.title}`);
      console.log(`   slug: ${slug}`);
      console.log(`   category: ${doc.category}`);
      console.log();
    }
  }

  console.log('='.repeat(70));
  console.log(`\n📊 Summary:`);
  console.log(`   ✅ PDFs present: ${presentCount}/${docs.length}`);
  console.log(`   ❌ PDFs missing: ${missingCount}/${docs.length}`);

  if (missing.length > 0) {
    console.log(`\n📋 Missing PDFs by category:`);
    const byCategory = {};
    missing.forEach(m => {
      if (!byCategory[m.category]) byCategory[m.category] = [];
      byCategory[m.category].push(m);
    });

    Object.keys(byCategory).forEach(cat => {
      console.log(`\n   ${cat}: ${byCategory[cat].length} missing`);
      byCategory[cat].forEach(m => {
        console.log(`      - ${m.slug}`);
      });
    });
  }

  console.log('\n' + '='.repeat(70) + '\n');

  await close();
}

main();