Before updates to NHTSA

2025-12-14 14:53:45 -06:00
parent 61e87bb9ad
commit 1fc69b7779
12 changed files with 1680503 additions and 1156458 deletions
--- a/data/make-model-import/qa_validate.py
+++ b/data/make-model-import/qa_validate.py
@@ -0,0 +1,112 @@
+#!/usr/bin/env python3
+"""
+Post-import QA validation for vehicle dropdown data.
+Runs basic duplicate and range checks against the motovaultpro Postgres container.
+"""
+
+import os
+import subprocess
+import sys
+
+
+def run_psql(query: str) -> str:
+    cmd = [
+        "docker",
+        "exec",
+        "mvp-postgres",
+        "psql",
+        "-U",
+        "postgres",
+        "-d",
+        "motovaultpro",
+        "-At",
+        "-c",
+        query,
+    ]
+    return subprocess.check_output(cmd, text=True)
+
+
+def check_container():
+    try:
+        subprocess.check_output(["docker", "ps"], text=True)
+    except Exception:
+        print("❌ Docker not available.")
+        sys.exit(1)
+
+    try:
+        containers = subprocess.check_output(
+            ["docker", "ps", "--filter", "name=mvp-postgres", "--format", "{{.Names}}"],
+            text=True,
+        ).strip()
+        if not containers:
+            print("❌ mvp-postgres container not running.")
+            sys.exit(1)
+    except Exception as exc:
+        print(f"❌ Failed to check containers: {exc}")
+        sys.exit(1)
+
+
+def main():
+    check_container()
+
+    print("🔍 Running QA checks...\n")
+
+    queries = {
+        "engine_duplicate_names": """
+            SELECT COUNT(*) FROM (
+                SELECT LOWER(name) as n, COUNT(*) c
+                FROM engines
+                GROUP BY 1 HAVING COUNT(*) > 1
+            ) t;
+        """,
+        "transmission_duplicate_types": """
+            SELECT COUNT(*) FROM (
+                SELECT LOWER(type) as t, COUNT(*) c
+                FROM transmissions
+                GROUP BY 1 HAVING COUNT(*) > 1
+            ) t;
+        """,
+        "vehicle_option_duplicates": """
+            SELECT COUNT(*) FROM (
+                SELECT year, make, model, trim, engine_id, transmission_id, COUNT(*) c
+                FROM vehicle_options
+                GROUP BY 1,2,3,4,5,6 HAVING COUNT(*) > 1
+            ) t;
+        """,
+        "year_range": """
+            SELECT MIN(year) || ' - ' || MAX(year) FROM vehicle_options;
+        """,
+        "counts": """
+            SELECT
+              (SELECT COUNT(*) FROM engines) AS engines,
+              (SELECT COUNT(*) FROM transmissions) AS transmissions,
+              (SELECT COUNT(*) FROM vehicle_options) AS vehicle_options;
+        """,
+    }
+
+    results = {}
+    for key, query in queries.items():
+        try:
+            results[key] = run_psql(query).strip()
+        except subprocess.CalledProcessError as exc:
+            print(f"❌ Query failed ({key}): {exc}")
+            sys.exit(1)
+
+    print(f"Engine duplicate names: {results['engine_duplicate_names']}")
+    print(f"Transmission duplicate types: {results['transmission_duplicate_types']}")
+    print(f"Vehicle option duplicates: {results['vehicle_option_duplicates']}")
+    print(f"Year range: {results['year_range']}")
+    print(f"Counts (engines, transmissions, vehicle_options): {results['counts']}")
+
+    if (
+        results["engine_duplicate_names"] == "0"
+        and results["transmission_duplicate_types"] == "0"
+        and results["vehicle_option_duplicates"] == "0"
+    ):
+        print("\n✅ QA checks passed.")
+    else:
+        print("\n❌ QA checks found issues.")
+
+
+if __name__ == "__main__":
+    main()