ScraperControl/scripts/normalize-country-codes.ts

/**
 * Normalize country codes in the database.
 * Converts full country names to ISO 3166-1 alpha-2 codes.
 *
 * Usage:
 *   npx tsx scripts/normalize-country-codes.ts --dry-run
 *   npx tsx scripts/normalize-country-codes.ts --execute
 */

import path from 'path';
import dotenv from 'dotenv';
dotenv.config({ path: path.resolve(process.cwd(), '.env') });

import { Pool } from 'pg';
import { PrismaPg } from '@prisma/adapter-pg';
import { PrismaClient } from '@prisma/client';
import { normalizeCountryCode } from '../src/lib/country-normalize';

const pool = new Pool({ connectionString: process.env.DATABASE_URL });
const adapter = new PrismaPg(pool);
const prisma = new PrismaClient({ adapter });

async function main() {
  const dryRun = !process.argv.includes('--execute');

  if (dryRun) {
    console.log('DRY RUN — no changes will be made. Use --execute to apply.\n');
  }

  // Get all distinct country values
  const countries = await prisma.church.findMany({
    select: { country: true },
    distinct: ['country'],
    where: { country: { not: null } },
  });

  const countryValues = countries
    .map(c => c.country)
    .filter((c): c is string => c !== null);

  console.log(`Found ${countryValues.length} distinct country values.\n`);

  // Group by normalization result
  const changes: { original: string; normalized: string; count?: number }[] = [];
  const alreadyNormalized: string[] = [];
  const unknown: string[] = [];

  for (const country of countryValues) {
    const normalized = normalizeCountryCode(country);

    if (normalized === country) {
      // Already correct or unknown
      if (country.length === 2 && country === country.toUpperCase()) {
        alreadyNormalized.push(country);
      } else {
        unknown.push(country);
      }
    } else {
      changes.push({ original: country, normalized });
    }
  }

  // Get counts for changes
  for (const change of changes) {
    const count = await prisma.church.count({
      where: { country: change.original },
    });
    change.count = count;
  }

  // Report
  console.log(`Already normalized (${alreadyNormalized.length}): ${alreadyNormalized.sort().join(', ')}\n`);

  if (changes.length > 0) {
    console.log(`Changes to apply (${changes.length}):`);
    for (const { original, normalized, count } of changes) {
      console.log(`  "${original}" → "${normalized}" (${count} churches)`);
    }
    console.log();
  } else {
    console.log('No changes needed — all country values are already normalized.\n');
  }

  if (unknown.length > 0) {
    console.log(`Unknown values (${unknown.length}): ${unknown.join(', ')}`);
    console.log('  These could not be mapped to ISO codes. Review manually.\n');
  }

  // Apply changes
  if (!dryRun && changes.length > 0) {
    let totalUpdated = 0;
    for (const { original, normalized } of changes) {
      const result = await prisma.church.updateMany({
        where: { country: original },
        data: { country: normalized },
      });
      totalUpdated += result.count;
      console.log(`Updated "${original}" → "${normalized}": ${result.count} churches`);
    }
    console.log(`\nTotal updated: ${totalUpdated} churches`);
  }

  await prisma.$disconnect();
  await pool.end();
}

main().catch(err => {
  console.error('Error:', err);
  process.exit(1);
});