before claude

2026-01-17 13:40:26 +01:00
parent d0b73b9319
commit cf7c415e22
35 changed files with 3411 additions and 221 deletions
--- a/scripts/missing_data_by_store.py
+++ b/scripts/missing_data_by_store.py
@@ -0,0 +1,121 @@
+import os
+from typing import Dict, Optional
+
+import psycopg2
+from psycopg2.extras import RealDictCursor
+
+
+def _env_str(name: str, default: str) -> str:
+    return os.environ.get(name, default)
+
+
+def _env_int(name: str, default: int) -> int:
+    try:
+        return int(os.environ.get(name, default))
+    except ValueError:
+        return default
+
+
+def get_connection():
+    return psycopg2.connect(
+        host=_env_str("PW_DB_HOST", "localhost"),
+        port=_env_int("PW_DB_PORT", 5432),
+        dbname=_env_str("PW_DB_NAME", "pricewatch"),
+        user=_env_str("PW_DB_USER", "pricewatch"),
+        password=_env_str("PW_DB_PASSWORD", "pricewatch"),
+    )
+
+
+def gather(limit: Optional[int] = None):
+    query = """
+    SELECT
+        COALESCE(p.source, 'unknown') AS source,
+        p.id,
+        p.reference,
+        p.title,
+        p.description,
+        p.category,
+        p.msrp,
+        EXISTS (
+            SELECT 1 FROM product_images WHERE product_id = p.id LIMIT 1
+        ) AS has_image,
+        EXISTS (
+            SELECT 1 FROM product_specs WHERE product_id = p.id LIMIT 1
+        ) AS has_specs,
+        ph.price,
+        ph.stock_status
+    FROM products p
+    LEFT JOIN LATERAL (
+        SELECT price, stock_status
+        FROM price_history
+        WHERE product_id = p.id
+        ORDER BY fetched_at DESC
+        LIMIT 1
+    ) ph ON TRUE
+    ORDER BY p.last_updated_at DESC
+    """
+    if limit:
+        query += f" LIMIT {limit}"
+
+    with get_connection() as conn:
+        with conn.cursor(cursor_factory=RealDictCursor) as cur:
+            cur.execute(query)
+            return cur.fetchall()
+
+
+def summarize(rows):
+    stores: Dict[str, Dict[str, object]] = {}
+    fields = [
+        ("price", "Prix absent"),
+        ("stock_status", "Statut stock manquant"),
+        ("description", "Description manquante"),
+        ("category", "Catégorie manquante"),
+        ("msrp", "Prix conseillé absent"),
+        ("has_image", "Images absentes"),
+        ("has_specs", "Caractéristiques absentes"),
+    ]
+    for row in rows:
+        store = row["source"] or "unknown"
+        entry = stores.setdefault(
+            store,
+            {
+                "total": 0,
+                "details": {field: [] for field, _ in fields},
+            },
+        )
+        entry["total"] += 1
+        for field, label in fields:
+            value = row.get(field)
+            if field in ("has_image", "has_specs"):
+                missing = not value
+            else:
+                missing = value in (None, "", [])
+            if missing:
+                entry["details"][field].append(
+                    {
+                        "id": row["id"],
+                        "reference": row["reference"],
+                        "title": row["title"] or "Sans titre",
+                    }
+                )
+    return fields, stores
+
+
+def pretty_print(fields, stores):
+    for store, data in stores.items():
+        print(f"\n=== Store: {store} ({data['total']} produits) ===")
+        for field, label in fields:
+            unit = len(data["details"][field])
+            print(f"  {label}: {unit}")
+            for item in data["details"][field][:5]:
+                print(f"    - [{item['id']}] {item['reference']} · {item['title']}")
+
+
+def main():
+    rows = gather(limit=1000)
+    fields, stores = summarize(rows)
+    pretty_print(fields, stores)
+
+
+if __name__ == "__main__":
+    main()