"""
retrieve_scrape_data 入库语句。
列顺序须与 DrugPipeline.storge_data 中构建的 row 字段一致。
"""

# 与 pipelines/drug_pipelines.py 里 row 的键顺序一致（便于核对与扩展）
RETRIEVE_SCRAPE_INSERT_COLUMNS = (
    "platform_id",
    "platform_item_id",
    "enterprise_id",
    "product_name",
    "product_specs",
    "product_brand",
    "one_box_price",
    "link_url",
    "store_name",
    "store_url",
    "shipment_province_id",
    "shipment_province_name",
    "shipment_city_id",
    "shipment_city_name",
    "manufacturer",
    "company_name",
    "scrape_date",
    "is_sold_out",
    "min_price",
    "sales",
    "inventory",
    "snapshot_url",
    "approval_number",
    "expiry_date",
    "update_time",
    "insert_time",
    "number",
    "task_id",
    "anonymous_store_name",
    "search_name",
    "collect_config_info",
    "area_info",
    "city_name",
    "city_id",
    "province_name",
    "province_id",
    "collect_equipment_account_id",
    "collect_region_id",
    "collect_round",
)


def _retrieve_scrape_insert_if_absent_sql() -> str:
    cols = RETRIEVE_SCRAPE_INSERT_COLUMNS
    column_list = ", ".join(f"`{c}`" for c in cols)
    inner_select = ",\n            ".join(f"%s AS `{c}`" for c in cols)
    return f"""INSERT INTO `retrieve_scrape_data` ({column_list})
                SELECT * FROM (
                    SELECT
                            {inner_select}
                ) AS `t`
                WHERE NOT EXISTS (
                    SELECT 1 FROM `retrieve_scrape_data` AS `p`
                    WHERE `p`.`platform_id` <=> %s
                      AND `p`.`scrape_date` <=> %s
                      AND `p`.`platform_item_id` <=> %s
                      AND `p`.`collect_equipment_account_id` <=> %s
                      AND `p`.`collect_round` <=> %s
                );"""
sql_map = {
    # 仅插入：同 platform_id, scrape_date, platform_item_id, collect_equipment_account_id, collect_round 已存在则跳过
    "retrieve_scrape_insert_if_absent_sql": _retrieve_scrape_insert_if_absent_sql(),
}