ScraperControl/scripts/debug/check-enrichment-status.ts

import { config } from 'dotenv';
import { PrismaClient } from '@prisma/client';
import { Pool } from 'pg';
import { PrismaPg } from '@prisma/adapter-pg';

// Load .env.local first, then .env
config({ path: '.env.local' });
config({ path: '.env' });

const connectionString = process.env.DATABASE_URL;

if (!connectionString) {
  throw new Error('DATABASE_URL environment variable is not set');
}

const pool = new Pool({ connectionString });
const adapter = new PrismaPg(pool);
const prisma = new PrismaClient({ adapter });

async function checkEnrichmentStatus() {
  try {
    console.log('Checking enrichment status...\n');

    // Overall stats
    const totalOSM = await prisma.church.count({
      where: { source: 'osm' },
    });

    const enriched = await prisma.church.count({
      where: {
        source: 'osm',
        googlePlaceId: { not: null },
      },
    });

    const withWebsite = await prisma.church.count({
      where: {
        source: 'osm',
        hasWebsite: true,
      },
    });

    const needEnrichment = await prisma.church.count({
      where: {
        source: 'osm',
        hasWebsite: false,
        website: null,
      },
    });

    // Recently enriched (last 24 hours)
    const yesterday = new Date();
    yesterday.setDate(yesterday.getDate() - 1);

    const recentlyEnriched = await prisma.church.count({
      where: {
        source: 'osm',
        googlePlaceId: { not: null },
        updatedAt: { gte: yesterday },
      },
    });

    // Get top 10 priority countries status
    const PRIORITY_COUNTRIES = ['FR', 'DE', 'ES', 'PL', 'BR', 'PT', 'PH', 'CZ', 'MX', 'HU'];

    console.log('═══════════════════════════════════════════════════════════════');
    console.log('OVERALL ENRICHMENT STATUS');
    console.log('═══════════════════════════════════════════════════════════════');
    console.log(`Total OSM churches:           ${totalOSM.toLocaleString()}`);
    console.log(`Churches with Google Place ID: ${enriched.toLocaleString()} (${((enriched / totalOSM) * 100).toFixed(2)}%)`);
    console.log(`Churches with websites:        ${withWebsite.toLocaleString()} (${((withWebsite / totalOSM) * 100).toFixed(2)}%)`);
    console.log(`Need enrichment:               ${needEnrichment.toLocaleString()} (${((needEnrichment / totalOSM) * 100).toFixed(2)}%)`);
    console.log('');
    console.log(`Recently enriched (24h):       ${recentlyEnriched.toLocaleString()}`);
    console.log('');

    // Priority countries breakdown
    console.log('═══════════════════════════════════════════════════════════════');
    console.log('TOP 10 PRIORITY COUNTRIES STATUS');
    console.log('═══════════════════════════════════════════════════════════════');
    console.log('');

    for (const country of PRIORITY_COUNTRIES) {
      const total = await prisma.church.count({
        where: { source: 'osm', country },
      });

      const countryEnriched = await prisma.church.count({
        where: {
          source: 'osm',
          country,
          googlePlaceId: { not: null },
        },
      });

      const countryWithWebsite = await prisma.church.count({
        where: {
          source: 'osm',
          country,
          OR: [
            { hasWebsite: true },
            { googlePlaceId: { not: null } },
          ],
        },
      });

      const countryNeedEnrichment = await prisma.church.count({
        where: {
          source: 'osm',
          country,
          hasWebsite: false,
          website: null,
        },
      });

      const websitePercent = (countryWithWebsite / total) * 100;
      const enrichedPercent = (countryEnriched / total) * 100;

      console.log(`${country.padEnd(4)} | Total: ${String(total).padStart(6)} | Enriched: ${String(countryEnriched).padStart(5)} (${enrichedPercent.toFixed(1)}%) | With Website: ${String(countryWithWebsite).padStart(5)} (${websitePercent.toFixed(1)}%) | Need: ${String(countryNeedEnrichment).padStart(6)}`);
    }

    console.log('');

    // Estimate timeline
    const daysRemaining = Math.ceil(needEnrichment / 390);
    const monthsRemaining = (daysRemaining / 30).toFixed(1);

    console.log('═══════════════════════════════════════════════════════════════');
    console.log('TIMELINE ESTIMATE');
    console.log('═══════════════════════════════════════════════════════════════');
    console.log(`At 390 churches/day:`);
    console.log(`  Days remaining:   ${daysRemaining} days`);
    console.log(`  Months remaining: ~${monthsRemaining} months`);
    console.log(`  Estimated completion: ${new Date(Date.now() + daysRemaining * 24 * 60 * 60 * 1000).toLocaleDateString()}`);
    console.log('');

  } catch (error) {
    console.error('Error:', error);
    process.exit(1);
  } finally {
    await prisma.$disconnect();
    await pool.end();
  }
}

checkEnrichmentStatus();