ConceptARC db upload

scripts/verify_conceptarc_upload.py

@@ -0,0 +1,86 @@
+#!/usr/bin/env python3
+"""
+Verify ConceptArc data was uploaded correctly to the database
+"""
+
+import os
+import json
+import pymysql
+from dotenv import load_dotenv
+
+def load_env_config():
+    """Load database configuration from .env file"""
+    load_dotenv()
+    return {
+        'host': os.getenv('DB_HOST'),
+        'user': os.getenv('DB_USER'),
+        'password': os.getenv('DB_PASSWORD'),
+        'database': os.getenv('DB_NAME'),
+        'port': int(os.getenv('DB_PORT', 3306)),
+        'charset': 'utf8mb4'
+    }
+
+def main():
+    print("Verifying ConceptArc Upload")
+    print("=" * 50)
+
+    config = load_env_config()
+    connection = pymysql.connect(**config)
+    cursor = connection.cursor()
+
+    # Check arc_puzzles table
+    cursor.execute("""
+        SELECT `set`, COUNT(*)
+        FROM arc_puzzles
+        WHERE corpora = 'ConceptArc'
+        GROUP BY `set`
+        ORDER BY `set`
+    """)
+
+    print("\nConceptArc puzzles by category (from arc_puzzles):")
+    total = 0
+    for category, count in cursor.fetchall():
+        print(f"  {category}: {count} puzzles")
+        total += count
+    print(f"  TOTAL: {total} puzzles")
+
+    # Check arc_jsons table
+    cursor.execute("""
+        SELECT COUNT(*)
+        FROM arc_jsons aj
+        JOIN arc_puzzles ap ON aj.arc_puzzle_id = ap.id
+        WHERE ap.corpora = 'ConceptArc'
+    """)
+    json_count = cursor.fetchone()[0]
+    print(f"\nConceptArc entries in arc_jsons: {json_count}")
+
+    # Sample some puzzles
+    cursor.execute("""
+        SELECT ap.id, ap.corpora, ap.`set`, aj.json, aj.solution
+        FROM arc_puzzles ap
+        JOIN arc_jsons aj ON ap.id = aj.arc_puzzle_id
+        WHERE ap.corpora = 'ConceptArc'
+        LIMIT 3
+    """)
+
+    print("\n" + "=" * 50)
+    print("Sample ConceptArc entries:")
+    print("=" * 50)
+
+    for row in cursor.fetchall():
+        puzzle_id, corpora, category, json_data, solution = row
+        print(f"\nPuzzle ID: {puzzle_id}")
+        print(f"  Corpora: {corpora}")
+        print(f"  Category: {category}")
+        print(f"  JSON length: {len(json_data)} chars")
+        print(f"  Has solution: {'Yes' if solution else 'No'}")
+
+        if solution:
+            sol = json.loads(solution)
+            print(f"  Solution dimensions: {len(sol)}x{len(sol[0]) if sol else 0}")
+
+    connection.close()
+    print("\n✓ Verification complete")
+
+if __name__ == '__main__':
+    main()