feat: add forward-geocode to scheduler pipeline

- Wire enrich-with-forward-geocode.ts as scheduler job type
- Add geocode-enrichment pipeline group (500/cycle, post-imports)
- Harden transfer script: skip churches at (0,0) coordinates
- Rewrite dedup-mass-schedules.ts with raw SQL to avoid Prisma 7 stack overflow

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

scripts/dedup-mass-schedules.ts

@@ -3,12 +3,13 @@ import dotenv from 'dotenv';
 import path from 'path';
 dotenv.config({ path: path.resolve(process.cwd(), '.env.local') });
 import { Pool } from 'pg';
-import { PrismaPg } from '@prisma/adapter-pg';
-import { PrismaClient } from '@prisma/client';
 
 const pool = new Pool({ connectionString: process.env.DATABASE_URL });
-const adapter = new PrismaPg(pool);
+
-const prisma = new PrismaClient({ adapter });
+interface CountResult {
+  churches_with_dups: string;
+  duplicate_rows: string;
+}
 
 async function main() {
   const dryRun = !process.argv.includes('--execute');
@@ -17,44 +18,52 @@ async function main() {
     console.log('DRY RUN - pass --execute to actually delete duplicates\n');
   }
 
-  const churches = await prisma.church.findMany({
+  const client = await pool.connect();
-    where: { massSchedules: { some: { isActive: true } } },
-    include: { massSchedules: { where: { isActive: true }, orderBy: { createdAt: 'asc' } } },
-  });
 
-  let churchesFixed = 0;
+  try {
-  let rowsDeleted = 0;
+    const countResult = await client.query<CountResult>(`
+      WITH ranked AS (
+        SELECT church_id,
+               ROW_NUMBER() OVER (
+                 PARTITION BY church_id, day_of_week, time, language
+                 ORDER BY created_at ASC
+               ) AS rn
+        FROM mass_schedules
+        WHERE is_active = true
+      )
+      SELECT COUNT(DISTINCT church_id) AS churches_with_dups,
+             COUNT(*) AS duplicate_rows
+      FROM ranked
+      WHERE rn > 1;
+    `);
 
-  for (const church of churches) {
+    const { churches_with_dups, duplicate_rows } = countResult.rows[0];
-    const seen = new Map<string, string>();
+    console.log(`Churches with duplicate schedules: ${churches_with_dups}`);
-    const toDelete: string[] = [];
+    console.log(`Duplicate rows to ${dryRun ? 'delete' : 'delete'}: ${duplicate_rows}\n`);
 
-    for (const m of church.massSchedules) {
+    if (!dryRun && Number(duplicate_rows) > 0) {
-      const key = `${m.dayOfWeek}:${m.time}:${m.language}`;
+      console.log('Deleting duplicates (keeping oldest by created_at)...');
-      if (seen.has(key)) {
+
-        toDelete.push(m.id);
+      const deleteResult = await client.query(`
-      } else {
+        WITH ranked AS (
-        seen.set(key, m.id);
+          SELECT id,
-      }
+                 ROW_NUMBER() OVER (
-    }
+                   PARTITION BY church_id, day_of_week, time, language
-
+                   ORDER BY created_at ASC
-    if (toDelete.length > 0) {
+                 ) AS rn
-      churchesFixed++;
+          FROM mass_schedules
-      rowsDeleted += toDelete.length;
+          WHERE is_active = true
-
+        )
-      if (!dryRun) {
+        DELETE FROM mass_schedules
-        await prisma.massSchedule.deleteMany({
+        WHERE id IN (SELECT id FROM ranked WHERE rn > 1);
-          where: { id: { in: toDelete } },
+      `);
-        });
+
-      }
+      console.log(`Deleted ${deleteResult.rowCount} duplicate mass schedule rows.`);
     }
+  } finally {
+    client.release();
+    await pool.end();
   }
-
-  console.log(`Churches with duplicates: ${churchesFixed}`);
-  console.log(`Duplicate rows ${dryRun ? 'found' : 'deleted'}: ${rowsDeleted}`);
-
-  await prisma.$disconnect();
-  await pool.end();
 }
 
 main().catch((err) => {

scripts/scheduler.ts

@@ -59,6 +59,13 @@ const PIPELINE_GROUPS: PipelineGroup[] = [
       { name: 'masstimes-api-import', type: 'masstimes-api-import', config: {} },
     ],
   },
+  {
+    name: 'geocode-enrichment',
+    mode: 'sequential',
+    phases: [
+      { name: 'forward-geocode', type: 'forward-geocode-enrichment', config: { limit: 500 } },
+    ],
+  },
   {
     name: 'scrapers-batch-1',
     mode: 'parallel',
@@ -138,6 +145,12 @@ function getJobCommand(type: string, language?: string | null, config?: Record<s
       if (config?.country) args.push('--country', String(config.country));
       return { command: 'npx', args };
     }
+    case 'forward-geocode-enrichment': {
+      const args = ['tsx', 'scripts/enrich-with-forward-geocode.ts'];
+      if (limit) args.push('--limit', String(limit));
+      if (config?.country) args.push('--country', String(config.country));
+      return { command: 'npx', args };
+    }
     case 'match-search-results': {
       const args = ['tsx', 'scripts/match-search-results.ts'];
       if (limit) args.push('--limit', String(limit));

scripts/transfer-enriched-to-neon.ts

@@ -103,7 +103,8 @@ async function main() {
         { phone: { not: null } },
         { googlePlaceId: { not: null } },
         { massSchedules: { some: {} } },
-      ]
+      ],
+      NOT: { latitude: 0, longitude: 0 },
     };
 
     // Add incremental filter if applicable