feat: add integration tests and seed script for Azure Storage emulator

- scripts/seed_test_data.py: seeds 45 orders (30 Recived + 15 Payed) across
  5 users and 3 days into the emulator
- tests/test_integration.py: full pipeline test (seed → scan → backup → delete)
  with table cleanup before/after
- pytest marker "integration" to separate from unit tests

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

scripts/seed_test_data.py

@@ -0,0 +1,115 @@
+"""Seed test data into Azure Table emulator for integration testing."""
+
+import sys
+from datetime import datetime, timezone
+
+from azure.data.tables import TableServiceClient
+
+DEFAULT_CONNECTION_STRING = (
+    "AccountName=devstoreaccount1;"
+    "AccountKey=Eby8vdM02xNOcqFlqUwJPLlmEtlCDXJ1OUzFT50uSRZ6IFsuFq2UVErCz4I6tq/"
+    "K1SZFPTOtr/KBHBeksoGMGw==;"
+    "DefaultEndpointsProtocol=http;"
+    "BlobEndpoint=http://192.168.9.101:10000/devstoreaccount1;"
+    "QueueEndpoint=http://192.168.9.101:10001/devstoreaccount1;"
+    "TableEndpoint=http://192.168.9.101:10002/devstoreaccount1;"
+)
+
+TABLE_NAME = "order"
+
+# Test users
+USERS = ["user001", "user002", "user003", "user004", "user005"]
+
+# 3 days of data: Jan 1-3, 2025
+DATES = [
+    datetime(2025, 1, 1, tzinfo=timezone.utc),
+    datetime(2025, 1, 2, tzinfo=timezone.utc),
+    datetime(2025, 1, 3, tzinfo=timezone.utc),
+]
+
+
+def seed(connection_string: str) -> None:
+    service = TableServiceClient.from_connection_string(connection_string)
+    service.create_table_if_not_exists(table_name=TABLE_NAME)
+    client = service.get_table_client(table_name=TABLE_NAME)
+
+    # Clear existing data
+    existing = list(client.list_entities(select=["PartitionKey", "RowKey"]))
+    by_partition: dict[str, list[dict]] = {}
+    for e in existing:
+        pk = e["PartitionKey"]
+        by_partition.setdefault(pk, []).append(
+            {"PartitionKey": pk, "RowKey": e["RowKey"]}
+        )
+    for pk, entities in by_partition.items():
+        for i in range(0, len(entities), 100):
+            batch = entities[i : i + 100]
+            ops = [("delete", e) for e in batch]
+            client.submit_transaction(ops)
+    if existing:
+        print(f"Cleared {len(existing)} existing records")
+
+    # Seed new data
+    entities = []
+    order_idx = 0
+    for date in DATES:
+        for user in USERS:
+            # 2 Recived orders per user per day
+            for _ in range(2):
+                hour = 8 + (order_idx % 12)
+                entities.append({
+                    "PartitionKey": user,
+                    "RowKey": f"test-order-{order_idx:04d}",
+                    "OrderTime": date.replace(hour=hour, minute=order_idx % 60),
+                    "ReceiveTime": date.replace(hour=hour + 1, minute=order_idx % 60),
+                    "revenue": 10.0 + order_idx,
+                    "platformOrder": f"plat-{order_idx:04d}",
+                    "prodCount": 1 + (order_idx % 5),
+                    "pfId": user,
+                    "prodId": f"prod-{order_idx % 3}",
+                    "prodPrice": 10.0 + (order_idx % 10),
+                    "state": 1,  # Recived
+                    "ver": 3,
+                    "customData": None,
+                })
+                order_idx += 1
+
+            # 1 Payed order per user per day (should be skipped by backup)
+            entities.append({
+                "PartitionKey": user,
+                "RowKey": f"test-order-{order_idx:04d}",
+                "OrderTime": date.replace(hour=14, minute=30),
+                "ReceiveTime": None,
+                "revenue": 5.0,
+                "platformOrder": f"plat-payed-{order_idx:04d}",
+                "prodCount": 1,
+                "pfId": user,
+                "prodId": "prod-pay",
+                "prodPrice": 5.0,
+                "state": 0,  # Payed
+                "ver": 3,
+                "customData": None,
+            })
+            order_idx += 1
+
+    # Insert in batches grouped by PartitionKey
+    total = 0
+    by_pk: dict[str, list[dict]] = {}
+    for e in entities:
+        by_pk.setdefault(e["PartitionKey"], []).append(e)
+
+    for pk, pk_entities in by_pk.items():
+        for i in range(0, len(pk_entities), 100):
+            batch = pk_entities[i : i + 100]
+            ops = [("create", e) for e in batch]
+            client.submit_transaction(ops)
+            total += len(batch)
+
+    print(f"Seeded {total} orders across {len(by_pk)} partitions, {len(DATES)} days")
+    print(f"  Recived (state=1): {len([e for e in entities if e['state'] == 1])}")
+    print(f"  Payed  (state=0): {len([e for e in entities if e['state'] == 0])}")
+
+
+if __name__ == "__main__":
+    conn_str = sys.argv[1] if len(sys.argv) > 1 else DEFAULT_CONNECTION_STRING
+    seed(conn_str)