{"db": "hard_retail", "question": "找出最近30天内下过单但从未有过退款记录的客户,列出他们的姓名和城市。", "sql": "SELECT c.customer_name, c.city FROM customers c WHERE EXISTS (SELECT 1 FROM sales_orders so WHERE so.customer_id = c.customer_id AND date(so.order_time) >= date('now', '-30 days')) AND NOT EXISTS (SELECT 1 FROM sales_orders so2 JOIN sales_items si ON so2.order_id = si.order_id JOIN returns r ON si.item_id = r.item_id WHERE so2.customer_id = c.customer_id)"} {"db": "hard_retail", "question": "列出每个品牌中售价最高的产品名称和价格,如果同品牌有多个最高价则全部列出。", "sql": "SELECT b.brand_name, p.product_name, p.list_price FROM products p JOIN brands b ON p.brand_id = b.brand_id WHERE p.list_price = (SELECT MAX(list_price) FROM products p2 WHERE p2.brand_id = p.brand_id)"} {"db": "hard_retail", "question": "计算每个仓库当前库存总量(考虑所有产品),并按照库存总量降序排名,给出仓库名称、城市和总量,以及排名。", "sql": "SELECT w.warehouse_name, w.city, SUM(i.quantity) AS total_quantity, RANK() OVER (ORDER BY SUM(i.quantity) DESC) AS rank FROM inventory i JOIN warehouses w ON i.warehouse_id = w.warehouse_id GROUP BY w.warehouse_id"} {"db": "hard_retail", "question": "找出过去一年内没有被任何客户购买过的产品(即没有出现在销售订单明细中的产品),列出产品名称和SKU。", "sql": "SELECT p.product_name, p.sku FROM products p WHERE NOT EXISTS (SELECT 1 FROM sales_items si JOIN sales_orders so ON si.order_id = so.order_id WHERE si.product_id = p.product_id AND date(so.order_time) >= date('now', '-1 year'))"} {"db": "hard_retail", "question": "找出订单总金额排名前三的订单。订单总金额 = 该订单所有明细的「数量×单价」之和 ×(1 − 折扣率 discount)+ 运费。列出订单ID、客户姓名、订单总金额和下单时间,按订单总金额从高到低排。", "sql": "SELECT so.order_id, c.customer_name, SUM(si.quantity * si.unit_price) * (1 - so.discount) + so.shipping_fee AS total_amount, so.order_time FROM sales_orders so JOIN customers c ON so.customer_id = c.customer_id JOIN sales_items si ON so.order_id = si.order_id GROUP BY so.order_id ORDER BY total_amount DESC LIMIT 3"} {"db": "hard_retail", "question": "列出所有供应商,以及他们供应的产品种类数量(通过采购订单关联),如果没有供应任何产品则显示0。", "sql": "SELECT s.supplier_name, COUNT(DISTINCT pi.product_id) AS product_count FROM suppliers s LEFT JOIN purchase_orders po ON s.supplier_id = po.supplier_id LEFT JOIN purchase_items pi ON po.po_id = pi.po_id GROUP BY s.supplier_id"} {"db": "hard_retail", "question": "每个下过单的客户,最近一次下单距今多少天?按日期算:今天的日期减去最近一次下单那天的日期。列出客户姓名和天数。", "sql": "SELECT c.customer_name, julianday(date('now')) - julianday(date(MAX(so.order_time))) AS days_since_last_order FROM customers c JOIN sales_orders so ON c.customer_id = so.customer_id GROUP BY c.customer_id"} {"db": "hard_retail", "question": "找出库存量低于安全库存的产品,列出产品名称、仓库名称和当前库存量。", "sql": "SELECT p.product_name, w.warehouse_name, i.quantity FROM inventory i JOIN products p ON i.product_id = p.product_id JOIN warehouses w ON i.warehouse_id = w.warehouse_id WHERE i.quantity < i.safety_stock"} {"db": "hard_retail", "question": "查询每个城市的客户数量,并按照客户数量降序排列,如果数量相同则按城市名称升序。", "sql": "SELECT city, COUNT(*) AS customer_count FROM customers GROUP BY city ORDER BY customer_count DESC, city ASC"} {"db": "hard_retail", "question": "2023 年及以后注册的客户里,谁下单次数最多?并列第一的都列出来,给出客户姓名和下单次数。", "sql": "WITH cnt AS (SELECT c.customer_id, c.customer_name, COUNT(*) AS order_count FROM customers c JOIN sales_orders so ON so.customer_id = c.customer_id WHERE strftime('%Y', c.register_date) >= '2023' GROUP BY c.customer_id) SELECT customer_name, order_count FROM cnt WHERE order_count = (SELECT MAX(order_count) FROM cnt)"} {"db": "hard_retail", "question": "每个产品类别里,净销售额最高的是哪个产品?净销售额 = 该产品所有销售明细的「数量×单价」之和,减去这些明细对应的退款金额之和。并列第一的都列出来,给出类别名称、产品名称和净销售额。", "sql": "WITH gross AS (SELECT product_id, SUM(quantity * unit_price) AS amt FROM sales_items GROUP BY product_id), refund AS (SELECT si.product_id, SUM(r.refund_amount) AS amt FROM returns r JOIN sales_items si ON si.item_id = r.item_id GROUP BY si.product_id), net AS (SELECT p.category_id, p.product_name, g.amt - COALESCE(rf.amt, 0) AS net_sales FROM gross g JOIN products p ON p.product_id = g.product_id LEFT JOIN refund rf ON rf.product_id = g.product_id) SELECT c.category_name, n.product_name, n.net_sales FROM net n JOIN categories c ON c.category_id = n.category_id WHERE n.net_sales = (SELECT MAX(n2.net_sales) FROM net n2 WHERE n2.category_id = n.category_id)"} {"db": "hard_retail", "question": "找出每个仓库中库存金额(数量 * 成本价)最高的产品,列出仓库名称、产品名称和库存金额。", "sql": "SELECT w.warehouse_name, p.product_name, i.quantity * p.cost_price AS inventory_value FROM inventory i JOIN products p ON i.product_id = p.product_id JOIN warehouses w ON i.warehouse_id = w.warehouse_id WHERE (i.warehouse_id, i.quantity * p.cost_price) IN (SELECT warehouse_id, MAX(quantity * cost_price) FROM inventory i2 JOIN products p2 ON i2.product_id = p2.product_id GROUP BY warehouse_id)"} {"db": "hard_retail", "question": "计算每个品牌的产品平均价格,并只显示平均价格高于所有产品平均价格的品牌。", "sql": "SELECT b.brand_name, AVG(p.list_price) AS avg_price FROM brands b JOIN products p ON b.brand_id = p.brand_id GROUP BY b.brand_id HAVING AVG(p.list_price) > (SELECT AVG(list_price) FROM products)"} {"db": "hard_retail", "question": "找出 2024 年 1 月 1 日及以后下的、并且至少有一件商品在退货表 returns 里有记录的订单,列出订单ID和客户姓名。", "sql": "SELECT DISTINCT so.order_id, c.customer_name FROM sales_orders so JOIN customers c ON so.customer_id = c.customer_id JOIN sales_items si ON so.order_id = si.order_id JOIN returns r ON si.item_id = r.item_id WHERE date(so.order_time) >= '2024-01-01'"} {"db": "hard_retail", "question": "统计每个有采购单的供应商的采购订单总金额,并按照总金额降序排名,列出供应商名称、总金额和排名。", "sql": "SELECT s.supplier_name, SUM(po.total_amount) AS total_amount, RANK() OVER (ORDER BY SUM(po.total_amount) DESC) AS rank FROM suppliers s JOIN purchase_orders po ON s.supplier_id = po.supplier_id GROUP BY s.supplier_id"} {"db": "hard_retail", "question": "查询所有退货记录,显示退货ID、产品名称、退货数量、退货原因和退款金额,并按退款金额从高到低排序。", "sql": "SELECT r.return_id, p.product_name, r.quantity, r.reason, r.refund_amount FROM returns r JOIN sales_items si ON r.item_id = si.item_id JOIN products p ON si.product_id = p.product_id ORDER BY r.refund_amount DESC"} {"db": "hard_retail", "question": "算每个客户的净消费金额:他所有订单明细的「数量×单价」之和,减去这些明细的退款金额之和。列出净消费金额最高的前 10 名客户的姓名和金额,按金额从高到低。", "sql": "WITH gross AS (SELECT so.customer_id, SUM(si.quantity * si.unit_price) AS amt FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.customer_id), refund AS (SELECT so.customer_id, SUM(r.refund_amount) AS amt FROM returns r JOIN sales_items si ON si.item_id = r.item_id JOIN sales_orders so ON so.order_id = si.order_id GROUP BY so.customer_id) SELECT c.customer_name, g.amt - COALESCE(rf.amt, 0) AS net_spent FROM gross g JOIN customers c ON c.customer_id = g.customer_id LEFT JOIN refund rf ON rf.customer_id = g.customer_id ORDER BY net_spent DESC LIMIT 10"} {"db": "hard_retail", "question": "找出所有产品中,成本价高于其所属品牌所有产品平均成本价的产品,列出产品名称、品牌名称和成本价。", "sql": "SELECT p.product_name, b.brand_name, p.cost_price FROM products p JOIN brands b ON p.brand_id = b.brand_id WHERE p.cost_price > (SELECT AVG(p2.cost_price) FROM products p2 WHERE p2.brand_id = p.brand_id)"} {"db": "hard_retail", "question": "每个城市里消费总额最高的客户是谁?消费总额 = 他所有订单明细的「数量×单价」之和(不管折扣和运费)。并列的都列出,给出城市、客户姓名和消费总额。", "sql": "WITH spend AS (SELECT c.city, c.customer_id, c.customer_name, SUM(si.quantity * si.unit_price) AS total FROM customers c JOIN sales_orders so ON so.customer_id = c.customer_id JOIN sales_items si ON si.order_id = so.order_id GROUP BY c.customer_id) SELECT city, customer_name, total FROM spend s WHERE total = (SELECT MAX(total) FROM spend s2 WHERE s2.city = s.city)"} {"db": "hard_retail", "question": "算每个品牌的毛利率:毛利率 =(销售额 − 销售成本)÷ 销售额,销售额 = 该品牌所有销售明细的「数量×单价」之和,销售成本 = 「数量×产品成本价 cost_price」之和。只算有销售的品牌,列出品牌名称和毛利率,按毛利率从高到低。", "sql": "SELECT b.brand_name, (SUM(si.quantity * si.unit_price) - SUM(si.quantity * p.cost_price)) / SUM(si.quantity * si.unit_price) AS margin_rate FROM sales_items si JOIN products p ON p.product_id = si.product_id JOIN brands b ON b.brand_id = p.brand_id GROUP BY b.brand_id ORDER BY margin_rate DESC"} {"db": "hard_retail", "question": "每个渠道(channel)有多少笔订单、平均每单买几件?每单件数 = 该订单明细数量之和,只统计有明细的订单。列出渠道、订单数、平均件数。", "sql": "WITH per_order AS (SELECT so.order_id, so.channel, SUM(si.quantity) AS qty FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.order_id) SELECT channel, COUNT(*) AS order_count, AVG(qty) AS avg_qty FROM per_order GROUP BY channel"} {"db": "hard_retail", "question": "找出采购总量大于销售总量的产品:采购总量 = purchase_items 里数量之和,销售总量 = sales_items 里数量之和(没卖过按 0 算);没采购过的产品不算。列出产品名称、采购总量、销售总量。", "sql": "WITH p_in AS (SELECT product_id, SUM(quantity) AS bought FROM purchase_items GROUP BY product_id), p_out AS (SELECT product_id, SUM(quantity) AS sold FROM sales_items GROUP BY product_id) SELECT p.product_name, i.bought, COALESCE(o.sold, 0) AS sold FROM p_in i JOIN products p ON p.product_id = i.product_id LEFT JOIN p_out o ON o.product_id = i.product_id WHERE i.bought > COALESCE(o.sold, 0)"} {"db": "hard_retail", "question": "每个仓库的库存总量占它容量(capacity)的比例是多少?比例 = 库存数量之和 ÷ 容量,要小数。只列有库存记录的仓库,给出仓库名称、库存总量、容量和比例。", "sql": "SELECT w.warehouse_name, SUM(i.quantity) AS total_qty, w.capacity, SUM(i.quantity) * 1.0 / w.capacity AS ratio FROM warehouses w JOIN inventory i ON i.warehouse_id = w.warehouse_id GROUP BY w.warehouse_id"} {"db": "hard_retail", "question": "按会员等级(tier)比较订单金额:每单金额 = 该订单明细「数量×单价」之和(只算有明细的订单)。列出等级、该等级下过单的客户数、平均每单金额。", "sql": "WITH ord AS (SELECT so.order_id, so.customer_id, SUM(si.quantity * si.unit_price) AS amt FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.order_id) SELECT c.tier, COUNT(DISTINCT c.customer_id) AS customer_count, AVG(o.amt) AS avg_order_amount FROM ord o JOIN customers c ON c.customer_id = o.customer_id GROUP BY c.tier"} {"db": "hard_retail", "question": "每家供应商的采购单里,状态为「已完成」的占比是多少(0 到 1 的小数)?只列有采购单的供应商,给出供应商名称、采购单总数、已完成单数和占比。", "sql": "SELECT s.supplier_name, COUNT(*) AS po_count, SUM(CASE WHEN po.status = '已完成' THEN 1 ELSE 0 END) AS done_count, SUM(CASE WHEN po.status = '已完成' THEN 1 ELSE 0 END) * 1.0 / COUNT(*) AS done_ratio FROM suppliers s JOIN purchase_orders po ON po.supplier_id = s.supplier_id GROUP BY s.supplier_id"} {"db": "hard_retail", "question": "哪些产品至少在 3 个不同的仓库里有库存记录?列出产品名称和有库存的仓库数。", "sql": "SELECT p.product_name, COUNT(DISTINCT i.warehouse_id) AS warehouse_count FROM products p JOIN inventory i ON i.product_id = p.product_id GROUP BY p.product_id HAVING COUNT(DISTINCT i.warehouse_id) >= 3"} {"db": "hard_retail", "question": "哪个退货原因(reason)出现的退货记录最多?并列的都列出,给出原因和记录数。", "sql": "WITH c AS (SELECT reason, COUNT(*) AS cnt FROM returns GROUP BY reason) SELECT reason, cnt FROM c WHERE cnt = (SELECT MAX(cnt) FROM c)"} {"db": "hard_retail", "question": "每个一级类别(parent_id 为空的类别)下面直接挂着几个子类别?列出一级类别名称和子类别数,没有子类别的显示 0。", "sql": "SELECT p.category_name, COUNT(c.category_id) AS child_count FROM categories p LEFT JOIN categories c ON c.parent_id = p.category_id WHERE p.parent_id IS NULL GROUP BY p.category_id"} {"db": "hard_retail", "question": "按类别统计在售产品(is_active = 1)的数量和平均标价,同时给出这个类别的父类别名称(没有父类别显示 NULL)。只列有在售产品的类别,给出类别名称、父类别名称、在售产品数、平均标价。", "sql": "SELECT c.category_name, pc.category_name AS parent_name, COUNT(*) AS active_count, AVG(p.list_price) AS avg_price FROM products p JOIN categories c ON c.category_id = p.category_id LEFT JOIN categories pc ON pc.category_id = c.parent_id WHERE p.is_active = 1 GROUP BY c.category_id"} {"db": "hard_retail", "question": "每个客户第一次下单是在注册后的第几天?天数 = 第一笔订单的日期(只取日期部分)减去注册日期。只列天数 ≥ 0 的客户,给出客户姓名和天数。", "sql": "SELECT c.customer_name, julianday(date(MIN(so.order_time))) - julianday(c.register_date) AS days FROM customers c JOIN sales_orders so ON so.customer_id = c.customer_id GROUP BY c.customer_id HAVING days >= 0"} {"db": "hard_retail", "question": "等级为「金卡」和「钻石」的客户,在各个渠道分别下了多少单?列出等级、渠道、订单数。", "sql": "SELECT c.tier, so.channel, COUNT(*) AS order_count FROM sales_orders so JOIN customers c ON c.customer_id = so.customer_id WHERE c.tier IN ('金卡', '钻石') GROUP BY c.tier, so.channel"} {"db": "hard_retail", "question": "核对采购单金额:明细合计 = 该采购单 purchase_items 的「数量×单位成本 unit_cost」之和。找出明细合计和采购单上 total_amount 相差超过 100 的采购单,列出采购单ID、供应商名称、total_amount 和明细合计。", "sql": "SELECT po.po_id, s.supplier_name, po.total_amount, SUM(pi.quantity * pi.unit_cost) AS items_total FROM purchase_orders po JOIN suppliers s ON s.supplier_id = po.supplier_id JOIN purchase_items pi ON pi.po_id = po.po_id GROUP BY po.po_id HAVING ABS(po.total_amount - SUM(pi.quantity * pi.unit_cost)) > 100"} {"db": "hard_retail", "question": "每个客户最常用的下单渠道是哪个(按订单数)?并列的渠道都列出,给出客户姓名、渠道和订单数。", "sql": "WITH c AS (SELECT customer_id, channel, COUNT(*) AS cnt FROM sales_orders GROUP BY customer_id, channel) SELECT cu.customer_name, c.channel, c.cnt FROM c JOIN customers cu ON cu.customer_id = c.customer_id WHERE c.cnt = (SELECT MAX(c2.cnt) FROM c c2 WHERE c2.customer_id = c.customer_id)"} {"db": "hard_retail", "question": "每个品牌的退款率:退款率 = 该品牌所有退款金额之和 ÷ 该品牌所有销售明细「数量×单价」之和;只算有销售的品牌,没有退货按 0。列出品牌名称和退款率,按退款率从高到低,相同按品牌名称排。", "sql": "WITH s AS (SELECT p.brand_id, SUM(si.quantity * si.unit_price) AS sales FROM sales_items si JOIN products p ON p.product_id = si.product_id GROUP BY p.brand_id), r AS (SELECT p.brand_id, SUM(r.refund_amount) AS refunds FROM returns r JOIN sales_items si ON si.item_id = r.item_id JOIN products p ON p.product_id = si.product_id GROUP BY p.brand_id) SELECT b.brand_name, COALESCE(r.refunds, 0) / s.sales AS refund_ratio FROM s JOIN brands b ON b.brand_id = s.brand_id LEFT JOIN r ON r.brand_id = s.brand_id ORDER BY refund_ratio DESC, b.brand_name"} {"db": "hard_retail", "question": "按下单年份统计销售额(明细「数量×单价」之和)和同比增长率:增长率 =(今年 − 上一年)÷ 上一年,最早那年显示 NULL。按年份排序,列出年份、销售额、增长率。", "sql": "WITH y AS (SELECT strftime('%Y', so.order_time) AS yr, SUM(si.quantity * si.unit_price) AS sales FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY yr) SELECT yr, sales, (sales - LAG(sales) OVER (ORDER BY yr)) / LAG(sales) OVER (ORDER BY yr) AS growth FROM y ORDER BY yr"} {"db": "hard_retail", "question": "每个仓库里库存最多的前 2 个产品是什么?同一仓库同一产品有多条库存记录的,先把数量加起来;数量相同时产品ID小的排前面。列出仓库名称、产品名称和库存量。", "sql": "WITH t AS (SELECT warehouse_id, product_id, SUM(quantity) AS qty FROM inventory GROUP BY warehouse_id, product_id), r AS (SELECT warehouse_id, product_id, qty, ROW_NUMBER() OVER (PARTITION BY warehouse_id ORDER BY qty DESC, product_id) AS rn FROM t) SELECT w.warehouse_name, p.product_name, r.qty FROM r JOIN warehouses w ON w.warehouse_id = r.warehouse_id JOIN products p ON p.product_id = r.product_id WHERE r.rn <= 2"} {"db": "hard_retail", "question": "下过至少 2 单的客户,相邻两次下单平均间隔多少天?间隔用下单时间的 julianday 相减。列出客户姓名和平均间隔天数。", "sql": "WITH g AS (SELECT customer_id, julianday(order_time) - julianday(LAG(order_time) OVER (PARTITION BY customer_id ORDER BY order_time)) AS gap FROM sales_orders) SELECT c.customer_name, AVG(g.gap) AS avg_gap FROM g JOIN customers c ON c.customer_id = g.customer_id WHERE g.gap IS NOT NULL GROUP BY g.customer_id"} {"db": "hard_retail", "question": "每个月新增了多少下单客户?客户第一次下单所在的月份('YYYY-MM')算他的新增月份。列出月份和新增客户数,按月份排序。", "sql": "WITH f AS (SELECT customer_id, MIN(order_time) AS first_time FROM sales_orders GROUP BY customer_id) SELECT strftime('%Y-%m', first_time) AS month, COUNT(*) AS new_customers FROM f GROUP BY month ORDER BY month"} {"db": "hard_retail", "question": "给每个客户的订单按下单时间编号(第 1 单、第 2 单……),列出客户姓名、订单ID、下单时间和序号。", "sql": "SELECT c.customer_name, so.order_id, so.order_time, ROW_NUMBER() OVER (PARTITION BY so.customer_id ORDER BY so.order_time) AS seq FROM sales_orders so JOIN customers c ON c.customer_id = so.customer_id"} {"db": "hard_retail", "question": "按下单日期统计每天的订单数,以及截至当天的累计订单数。按日期排序,列出日期、当天订单数、累计订单数。", "sql": "SELECT d, cnt, SUM(cnt) OVER (ORDER BY d) AS cum_cnt FROM (SELECT date(order_time) AS d, COUNT(*) AS cnt FROM sales_orders GROUP BY d) ORDER BY d"} {"db": "hard_retail", "question": "每个产品的标价在它所属类别里排第几?标价从高到低,同价同名次(RANK)。列出类别名称、产品名称、标价和名次。", "sql": "SELECT c.category_name, p.product_name, p.list_price, RANK() OVER (PARTITION BY p.category_id ORDER BY p.list_price DESC) AS price_rank FROM products p JOIN categories c ON c.category_id = p.category_id"} {"db": "hard_retail", "question": "每家供应商的采购单按下单日期排好(同一天的按采购单ID),每一单的 total_amount 比他上一单多了多少(本单 − 上一单,第一单显示 NULL)?列出供应商名称、采购单ID、下单日期和变化额。", "sql": "SELECT s.supplier_name, po.po_id, po.order_date, po.total_amount - LAG(po.total_amount) OVER (PARTITION BY po.supplier_id ORDER BY po.order_date, po.po_id) AS diff FROM purchase_orders po JOIN suppliers s ON s.supplier_id = po.supplier_id"} {"db": "hard_retail", "question": "每个城市注册最早的 3 位客户是谁?注册日期相同的客户ID小的优先。列出城市、客户姓名、注册日期。", "sql": "WITH r AS (SELECT customer_name, city, register_date, ROW_NUMBER() OVER (PARTITION BY city ORDER BY register_date, customer_id) AS rn FROM customers) SELECT city, customer_name, register_date FROM r WHERE rn <= 3"} {"db": "hard_retail", "question": "每个品牌的销售额(明细「数量×单价」之和)占全部销售额的百分之多少(乘以 100)?列出品牌名称、销售额、百分比。", "sql": "SELECT b.brand_name, SUM(si.quantity * si.unit_price) AS sales, SUM(si.quantity * si.unit_price) * 100.0 / SUM(SUM(si.quantity * si.unit_price)) OVER () AS pct FROM sales_items si JOIN products p ON p.product_id = si.product_id JOIN brands b ON b.brand_id = p.brand_id GROUP BY b.brand_id"} {"db": "hard_retail", "question": "把下过单的客户按消费总额(明细「数量×单价」之和)从高到低平均分成 4 组(NTILE(4)),列出客户姓名、消费总额和组号。", "sql": "SELECT c.customer_name, t.total, NTILE(4) OVER (ORDER BY t.total DESC) AS grp FROM (SELECT so.customer_id, SUM(si.quantity * si.unit_price) AS total FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.customer_id) t JOIN customers c ON c.customer_id = t.customer_id"} {"db": "hard_retail", "question": "每个仓库最近一次更新(updated_at 最晚)的库存记录是哪个产品?列出仓库名称、产品名称、更新时间。", "sql": "WITH r AS (SELECT warehouse_id, product_id, updated_at, ROW_NUMBER() OVER (PARTITION BY warehouse_id ORDER BY updated_at DESC) AS rn FROM inventory) SELECT w.warehouse_name, p.product_name, r.updated_at FROM r JOIN warehouses w ON w.warehouse_id = r.warehouse_id JOIN products p ON p.product_id = r.product_id WHERE r.rn = 1"} {"db": "hard_retail", "question": "每笔订单的金额(明细「数量×单价」之和)比这个客户自己的平均每单金额高多少?列出客户姓名、订单ID、订单金额、差值(订单金额 − 该客户平均每单金额),只算有明细的订单。", "sql": "WITH o AS (SELECT so.order_id, so.customer_id, SUM(si.quantity * si.unit_price) AS amt FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.order_id) SELECT c.customer_name, o.order_id, o.amt, o.amt - AVG(o.amt) OVER (PARTITION BY o.customer_id) AS diff FROM o JOIN customers c ON c.customer_id = o.customer_id"} {"db": "hard_retail", "question": "每个月('YYYY-MM')销售额最高的是哪一天?销售额 = 当天订单明细「数量×单价」之和。并列的都列出,给出月份、日期、当天销售额。", "sql": "WITH d AS (SELECT strftime('%Y-%m', so.order_time) AS month, date(so.order_time) AS day, SUM(si.quantity * si.unit_price) AS sales FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY day) SELECT month, day, sales FROM d WHERE sales = (SELECT MAX(d2.sales) FROM d d2 WHERE d2.month = d.month)"} {"db": "hard_retail", "question": "产品按上市日期(launch_date)排好(同一天的按产品ID),每款和前一款隔了多少天?第一款显示 NULL。列出产品名称、上市日期、间隔天数。", "sql": "SELECT product_name, launch_date, julianday(launch_date) - julianday(LAG(launch_date) OVER (ORDER BY launch_date, product_id)) AS gap_days FROM products"} {"db": "hard_retail", "question": "每个下过单的客户,金额最大的那一单占他全部消费的比例是多少?每单金额 = 明细「数量×单价」之和。列出客户姓名和比例。", "sql": "WITH o AS (SELECT so.order_id, so.customer_id, SUM(si.quantity * si.unit_price) AS amt FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.order_id) SELECT c.customer_name, MAX(o.amt) / SUM(o.amt) AS max_share FROM o JOIN customers c ON c.customer_id = o.customer_id GROUP BY o.customer_id"} {"db": "hard_retail", "question": "哪些客户下过单,但从来没在「线上」渠道下过单?列出客户姓名和城市。", "sql": "SELECT c.customer_name, c.city FROM customers c WHERE EXISTS (SELECT 1 FROM sales_orders so WHERE so.customer_id = c.customer_id) AND NOT EXISTS (SELECT 1 FROM sales_orders so WHERE so.customer_id = c.customer_id AND so.channel = '线上')"} {"db": "hard_retail", "question": "哪些产品从来没被采购过(purchase_items 里没有),却有库存记录?列出产品名称和 SKU。", "sql": "SELECT p.product_name, p.sku FROM products p WHERE NOT EXISTS (SELECT 1 FROM purchase_items pi WHERE pi.product_id = p.product_id) AND EXISTS (SELECT 1 FROM inventory i WHERE i.product_id = p.product_id)"} {"db": "hard_retail", "question": "哪些客户既买过「中国」品牌的产品,也买过非「中国」品牌的产品?列出客户姓名。", "sql": "SELECT c.customer_name FROM customers c WHERE EXISTS (SELECT 1 FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id JOIN products p ON p.product_id = si.product_id JOIN brands b ON b.brand_id = p.brand_id WHERE so.customer_id = c.customer_id AND b.country = '中国') AND EXISTS (SELECT 1 FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id JOIN products p ON p.product_id = si.product_id JOIN brands b ON b.brand_id = p.brand_id WHERE so.customer_id = c.customer_id AND b.country <> '中国')"} {"db": "hard_retail", "question": "哪些在售产品(is_active = 1)在所有位于「北京」的仓库里都没有库存记录?列出产品名称。", "sql": "SELECT p.product_name FROM products p WHERE p.is_active = 1 AND NOT EXISTS (SELECT 1 FROM inventory i JOIN warehouses w ON w.warehouse_id = i.warehouse_id WHERE i.product_id = p.product_id AND w.city = '北京')"} {"db": "hard_retail", "question": "有退货记录的订单里,哪些订单的状态不是「已退款」?列出订单ID和状态(每单一行),按订单ID排序。", "sql": "SELECT DISTINCT so.order_id, so.status FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id JOIN returns r ON r.item_id = si.item_id WHERE so.status <> '已退款' ORDER BY so.order_id"} {"db": "hard_retail", "question": "哪些客户只在一个渠道下过单(他所有订单的 channel 都一样)?列出客户姓名和那个渠道。", "sql": "SELECT c.customer_name, MIN(so.channel) AS channel FROM customers c JOIN sales_orders so ON so.customer_id = c.customer_id GROUP BY c.customer_id HAVING COUNT(DISTINCT so.channel) = 1"} {"db": "hard_retail", "question": "状态为「已完成」、有明细、而且没有任何一件商品退过货(明细在 returns 里都没记录)的订单有哪些?列出订单ID和下单时间。", "sql": "SELECT so.order_id, so.order_time FROM sales_orders so WHERE so.status = '已完成' AND EXISTS (SELECT 1 FROM sales_items si WHERE si.order_id = so.order_id) AND NOT EXISTS (SELECT 1 FROM sales_items si JOIN returns r ON r.item_id = si.item_id WHERE si.order_id = so.order_id)"} {"db": "hard_retail", "question": "哪些品牌的产品全部已经下架(is_active = 0)?只看至少有一款产品的品牌,列出品牌名称。", "sql": "SELECT b.brand_name FROM brands b WHERE EXISTS (SELECT 1 FROM products p WHERE p.brand_id = b.brand_id) AND NOT EXISTS (SELECT 1 FROM products p WHERE p.brand_id = b.brand_id AND p.is_active = 1)"} {"db": "hard_retail", "question": "哪些客户在至少两笔不同的订单里买过同一款产品?列出客户姓名和产品名称。", "sql": "SELECT c.customer_name, p.product_name FROM sales_items si JOIN sales_orders so ON so.order_id = si.order_id JOIN customers c ON c.customer_id = so.customer_id JOIN products p ON p.product_id = si.product_id GROUP BY so.customer_id, si.product_id HAVING COUNT(DISTINCT si.order_id) >= 2"} {"db": "hard_retail", "question": "有没有采购单的期望到货日期(expected_date)比下单日期(order_date)还早?列出这些采购单的ID、下单日期和期望到货日期。", "sql": "SELECT po_id, order_date, expected_date FROM purchase_orders WHERE date(expected_date) < date(order_date)"} {"db": "hard_retail", "question": "找出等级为「普通」、但消费总额超过「金卡」客户平均消费总额的客户。消费总额 = 所有订单明细「数量×单价」之和,金卡平均只算下过单的金卡客户。列出客户姓名和消费总额。", "sql": "WITH t AS (SELECT so.customer_id, SUM(si.quantity * si.unit_price) AS total FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.customer_id) SELECT c.customer_name, t.total FROM t JOIN customers c ON c.customer_id = t.customer_id WHERE c.tier = '普通' AND t.total > (SELECT AVG(t2.total) FROM t t2 JOIN customers c2 ON c2.customer_id = t2.customer_id WHERE c2.tier = '金卡')"} {"db": "hard_retail", "question": "哪些仓库收过采购单,但一条库存记录都没有?列出仓库名称和城市。", "sql": "SELECT w.warehouse_name, w.city FROM warehouses w WHERE EXISTS (SELECT 1 FROM purchase_orders po WHERE po.warehouse_id = w.warehouse_id) AND NOT EXISTS (SELECT 1 FROM inventory i WHERE i.warehouse_id = w.warehouse_id)"} {"db": "hard_retail", "question": "列出每一笔退货对应的客户姓名、产品名称、品牌名称、退货日期和退款金额。", "sql": "SELECT c.customer_name, p.product_name, b.brand_name, r.return_date, r.refund_amount FROM returns r JOIN sales_items si ON si.item_id = r.item_id JOIN sales_orders so ON so.order_id = si.order_id JOIN customers c ON c.customer_id = so.customer_id JOIN products p ON p.product_id = si.product_id JOIN brands b ON b.brand_id = p.brand_id"} {"db": "hard_retail", "question": "每家供应商供过的货(采购明细里的产品)涉及几个不同的品牌?只列有采购明细的供应商,给出供应商名称和品牌数。", "sql": "SELECT s.supplier_name, COUNT(DISTINCT p.brand_id) AS brand_count FROM suppliers s JOIN purchase_orders po ON po.supplier_id = s.supplier_id JOIN purchase_items pi ON pi.po_id = po.po_id JOIN products p ON p.product_id = pi.product_id GROUP BY s.supplier_id"} {"db": "hard_retail", "question": "每个城市的客户在「线上」渠道一共花了多少钱(明细「数量×单价」之和)?列出城市和金额,按金额从高到低。", "sql": "SELECT c.city, SUM(si.quantity * si.unit_price) AS total FROM customers c JOIN sales_orders so ON so.customer_id = c.customer_id JOIN sales_items si ON si.order_id = so.order_id WHERE so.channel = '线上' GROUP BY c.city ORDER BY total DESC"} {"db": "hard_retail", "question": "哪些采购明细的进货单位成本(unit_cost)高于这个产品现在的成本价(products.cost_price)?列出供应商名称、产品名称、unit_cost 和 cost_price。", "sql": "SELECT s.supplier_name, p.product_name, pi.unit_cost, p.cost_price FROM purchase_items pi JOIN purchase_orders po ON po.po_id = pi.po_id JOIN suppliers s ON s.supplier_id = po.supplier_id JOIN products p ON p.product_id = pi.product_id WHERE pi.unit_cost > p.cost_price"} {"db": "hard_retail", "question": "卖出价(sales_items.unit_price)低于产品标价(list_price)的销售明细,按品牌统计有多少条?列出品牌名称和条数。", "sql": "SELECT b.brand_name, COUNT(*) AS cnt FROM sales_items si JOIN products p ON p.product_id = si.product_id JOIN brands b ON b.brand_id = p.brand_id WHERE si.unit_price < p.list_price GROUP BY b.brand_id"} {"db": "hard_retail", "question": "每个仓库里的库存产品分属几个不同的品牌?只列有库存记录的仓库,给出仓库名称和品牌数,按品牌数从多到少、相同按仓库ID排。", "sql": "SELECT w.warehouse_name, COUNT(DISTINCT p.brand_id) AS brand_count FROM warehouses w JOIN inventory i ON i.warehouse_id = w.warehouse_id JOIN products p ON p.product_id = i.product_id GROUP BY w.warehouse_id ORDER BY brand_count DESC, w.warehouse_id"} {"db": "hard_retail", "question": "等级为「钻石」的客户买得最多的是哪个产品类别?按购买件数(明细数量之和)算,并列的都列出,给出类别名称和件数。", "sql": "WITH t AS (SELECT p.category_id, SUM(si.quantity) AS qty FROM customers c JOIN sales_orders so ON so.customer_id = c.customer_id JOIN sales_items si ON si.order_id = so.order_id JOIN products p ON p.product_id = si.product_id WHERE c.tier = '钻石' GROUP BY p.category_id) SELECT cat.category_name, t.qty FROM t JOIN categories cat ON cat.category_id = t.category_id WHERE t.qty = (SELECT MAX(qty) FROM t)"} {"db": "hard_retail", "question": "哪些订单里的商品来自 2 个及以上不同的品牌?列出订单ID、客户姓名和品牌数。", "sql": "SELECT so.order_id, c.customer_name, COUNT(DISTINCT p.brand_id) AS brand_count FROM sales_orders so JOIN customers c ON c.customer_id = so.customer_id JOIN sales_items si ON si.order_id = so.order_id JOIN products p ON p.product_id = si.product_id GROUP BY so.order_id HAVING COUNT(DISTINCT p.brand_id) >= 2"} {"db": "hard_retail", "question": "评分(rating)4 分及以上的供应商,他们供过货的产品(出现在他的采购明细里,同一产品只算一次)现在一共有多少库存(所有仓库加起来)?这些产品都没库存的供应商不用列。列出供应商名称和库存总量。", "sql": "WITH sp AS (SELECT DISTINCT po.supplier_id, pi.product_id FROM purchase_orders po JOIN purchase_items pi ON pi.po_id = po.po_id), stock AS (SELECT product_id, SUM(quantity) AS qty FROM inventory GROUP BY product_id) SELECT s.supplier_name, SUM(st.qty) AS total_stock FROM suppliers s JOIN sp ON sp.supplier_id = s.supplier_id JOIN stock st ON st.product_id = sp.product_id WHERE s.rating >= 4 GROUP BY s.supplier_id"} {"db": "hard_retail", "question": "在有明细的订单里,每个客户最近的一笔订单(下单时间最晚)买了几件商品(明细数量之和)?列出客户姓名、订单ID和件数。", "sql": "WITH o AS (SELECT so.customer_id, so.order_id, so.order_time, SUM(si.quantity) AS qty FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.order_id), r AS (SELECT customer_id, order_id, qty, ROW_NUMBER() OVER (PARTITION BY customer_id ORDER BY order_time DESC) AS rn FROM o) SELECT c.customer_name, r.order_id, r.qty FROM r JOIN customers c ON c.customer_id = r.customer_id WHERE r.rn = 1"} {"db": "hard_retail", "question": "下单时间在 2024 年内的订单里,每个品牌卖出了多少件(明细数量之和)?列出品牌名称和件数,按件数从多到少、相同按品牌名称排。", "sql": "SELECT b.brand_name, SUM(si.quantity) AS qty FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id JOIN products p ON p.product_id = si.product_id JOIN brands b ON b.brand_id = p.brand_id WHERE strftime('%Y', so.order_time) = '2024' GROUP BY b.brand_id ORDER BY qty DESC, b.brand_name"} {"db": "hard_retail", "question": "每个仓库从几家不同的供应商进过货(看 purchase_orders 里这个仓库的采购单)?只列有采购单的仓库,给出仓库名称、城市和供应商数。", "sql": "SELECT w.warehouse_name, w.city, COUNT(DISTINCT po.supplier_id) AS supplier_count FROM warehouses w JOIN purchase_orders po ON po.warehouse_id = w.warehouse_id GROUP BY w.warehouse_id"} {"db": "hard_retail", "question": "每个二级类别(parent_id 不为空的类别)下面产品的平均成本价是多少?没有产品的类别不列。给出父类别名称、类别名称和平均成本价。", "sql": "SELECT pc.category_name AS parent_name, c.category_name, AVG(p.cost_price) AS avg_cost FROM categories c JOIN categories pc ON pc.category_id = c.parent_id JOIN products p ON p.category_id = c.category_id GROUP BY c.category_id"} {"db": "hard_retail", "question": "每个品牌的产品分布在几个不同的仓库里(有库存记录就算)?列出品牌名称和仓库数,没有任何库存的品牌显示 0。", "sql": "SELECT b.brand_name, COUNT(DISTINCT i.warehouse_id) AS warehouse_count FROM brands b LEFT JOIN products p ON p.brand_id = b.brand_id LEFT JOIN inventory i ON i.product_id = p.product_id GROUP BY b.brand_id"} {"db": "hard_retail", "question": "把有明细的订单按金额(明细「数量×单价」之和)分档:小于 50000 为「小单」,50000 到 200000(含 50000、不含 200000)为「中单」,200000 及以上为「大单」。统计每档订单数,列出档位和订单数。", "sql": "WITH o AS (SELECT so.order_id, so.customer_id, SUM(si.quantity * si.unit_price) AS amt FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.order_id) SELECT CASE WHEN amt < 50000 THEN '小单' WHEN amt < 200000 THEN '中单' ELSE '大单' END AS bucket, COUNT(*) AS order_count FROM o GROUP BY bucket"} {"db": "hard_retail", "question": "按年龄段统计客户数:年龄 = 今年的年份 − 出生年份;小于 30 为「青年」,30 到 49 为「中年」,50 及以上为「老年」。列出年龄段和客户数。", "sql": "SELECT CASE WHEN CAST(strftime('%Y', 'now') AS INTEGER) - CAST(strftime('%Y', birthday) AS INTEGER) < 30 THEN '青年' WHEN CAST(strftime('%Y', 'now') AS INTEGER) - CAST(strftime('%Y', birthday) AS INTEGER) < 50 THEN '中年' ELSE '老年' END AS age_group, COUNT(*) AS cnt FROM customers GROUP BY age_group"} {"db": "hard_retail", "question": "每个渠道的订单里,状态为「已取消」的占多少(0 到 1 的小数)?列出渠道、订单总数、取消单数、取消占比。", "sql": "SELECT channel, COUNT(*) AS total, SUM(CASE WHEN status = '已取消' THEN 1 ELSE 0 END) AS cancelled, AVG(CASE WHEN status = '已取消' THEN 1.0 ELSE 0 END) AS cancel_rate FROM sales_orders GROUP BY channel"} {"db": "hard_retail", "question": "按季度统计订单数,季度写成「年份-Q季度号」,例如 '2024-Q1'(季度号 =(月份 + 2)÷ 3 取整)。按季度排序,列出季度和订单数。", "sql": "SELECT strftime('%Y', order_time) || '-Q' || ((CAST(strftime('%m', order_time) AS INTEGER) + 2) / 3) AS quarter, COUNT(*) AS order_count FROM sales_orders GROUP BY quarter ORDER BY quarter"} {"db": "hard_retail", "question": "每个仓库里,库存低于安全库存的记录有几条、总共几条?只列有库存记录的仓库,给出仓库名称、低库存记录数、总记录数。", "sql": "SELECT w.warehouse_name, SUM(CASE WHEN i.quantity < i.safety_stock THEN 1 ELSE 0 END) AS low_count, COUNT(*) AS total_count FROM warehouses w JOIN inventory i ON i.warehouse_id = w.warehouse_id GROUP BY w.warehouse_id"} {"db": "hard_retail", "question": "今年生日已经过了(含今天)的客户有哪些?只比月和日:生日的 'MM-DD' 小于等于今天的 'MM-DD'。列出客户姓名和生日。", "sql": "SELECT customer_name, birthday FROM customers WHERE strftime('%m-%d', birthday) <= strftime('%m-%d', 'now')"} {"db": "hard_retail", "question": "每个品牌在售(is_active = 1)和已下架(is_active = 0)的产品各有几个?列出品牌名称、在售数、下架数,没有产品的品牌两项都是 0。", "sql": "SELECT b.brand_name, COUNT(CASE WHEN p.is_active = 1 THEN 1 END) AS active_count, COUNT(CASE WHEN p.is_active = 0 THEN 1 END) AS inactive_count FROM brands b LEFT JOIN products p ON p.brand_id = b.brand_id GROUP BY b.brand_id"} {"db": "hard_retail", "question": "每个城市的客户里,等级是「金卡」或「钻石」的占该城市客户总数的比例(0 到 1 的小数)是多少?列出城市和比例,按比例从高到低、相同按城市名排。", "sql": "SELECT city, AVG(CASE WHEN tier IN ('金卡', '钻石') THEN 1.0 ELSE 0 END) AS high_ratio FROM customers GROUP BY city ORDER BY high_ratio DESC, city"} {"db": "hard_retail", "question": "最近 90 天(下单日期不早于今天往前 90 天)每周的订单数是多少?周用 strftime 的 '%Y-%W' 格式表示。按周排序,列出周和订单数。", "sql": "SELECT strftime('%Y-%W', order_time) AS week, COUNT(*) AS cnt FROM sales_orders WHERE date(order_time) >= date('now', '-90 days') GROUP BY week ORDER BY week"} {"db": "hard_retail", "question": "每家供应商合作了几年(今年的年份 − cooperate_since 的年份)?不满 3 年算「新」,3 到 5 年算「稳定」,超过 5 年算「老」。列出供应商名称、合作年数、分组。", "sql": "SELECT supplier_name, CAST(strftime('%Y', 'now') AS INTEGER) - CAST(strftime('%Y', cooperate_since) AS INTEGER) AS years, CASE WHEN CAST(strftime('%Y', 'now') AS INTEGER) - CAST(strftime('%Y', cooperate_since) AS INTEGER) < 3 THEN '新' WHEN CAST(strftime('%Y', 'now') AS INTEGER) - CAST(strftime('%Y', cooperate_since) AS INTEGER) <= 5 THEN '稳定' ELSE '老' END AS grp FROM suppliers"} {"db": "hard_retail", "question": "退货发生在下单后第几天?天数 = 退货日期 − 下单日期(都只取日期)。分成:负数为「异常」,0 到 7 天为「7天内」,8 到 30 天为「30天内」,超过 30 天为「30天以上」。列出区间和退货记录数。", "sql": "WITH d AS (SELECT julianday(r.return_date) - julianday(date(so.order_time)) AS days FROM returns r JOIN sales_items si ON si.item_id = r.item_id JOIN sales_orders so ON so.order_id = si.order_id) SELECT CASE WHEN days < 0 THEN '异常' WHEN days <= 7 THEN '7天内' WHEN days <= 30 THEN '30天内' ELSE '30天以上' END AS bucket, COUNT(*) AS cnt FROM d GROUP BY bucket"} {"db": "hard_retail", "question": "把每个月('YYYY-MM')的采购金额(purchase_orders 的 total_amount 之和,按 order_date)和销售金额(明细「数量×单价」之和,按下单时间)并排列出;某个月只有一边有金额,另一边显示 0,两边都没金额的月份不列。按月份排序。", "sql": "WITH p AS (SELECT strftime('%Y-%m', order_date) AS m, SUM(total_amount) AS amt FROM purchase_orders GROUP BY m), s AS (SELECT strftime('%Y-%m', so.order_time) AS m, SUM(si.quantity * si.unit_price) AS amt FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY m), months AS (SELECT m FROM p UNION SELECT m FROM s) SELECT months.m AS month, COALESCE(p.amt, 0) AS purchase_amount, COALESCE(s.amt, 0) AS sales_amount FROM months LEFT JOIN p ON p.m = months.m LEFT JOIN s ON s.m = months.m ORDER BY month"} {"db": "hard_retail", "question": "每个客户在注册后 30 天内(订单日期不早于注册日期,且不晚于注册日期 + 30 天)下了几单?列出客户姓名和单数,没下的显示 0。", "sql": "SELECT c.customer_name, COUNT(so.order_id) AS cnt FROM customers c LEFT JOIN sales_orders so ON so.customer_id = c.customer_id AND date(so.order_time) BETWEEN c.register_date AND date(c.register_date, '+30 days') GROUP BY c.customer_id"} {"db": "hard_retail", "question": "2025 年每个月「门店」和「线上」两个渠道分别有多少单?只看这两个渠道的订单,列出月份('YYYY-MM')、门店单数、线上单数,按月份排序。", "sql": "SELECT strftime('%Y-%m', order_time) AS month, SUM(CASE WHEN channel = '门店' THEN 1 ELSE 0 END) AS store_count, SUM(CASE WHEN channel = '线上' THEN 1 ELSE 0 END) AS online_count FROM sales_orders WHERE strftime('%Y', order_time) = '2025' AND channel IN ('门店', '线上') GROUP BY month ORDER BY month"} {"db": "hard_retail", "question": "哪些产品的标价高于它所在类别的平均标价,同时成本价低于它所在品牌的平均成本价?列出产品名称、标价、成本价。", "sql": "SELECT p.product_name, p.list_price, p.cost_price FROM products p WHERE p.list_price > (SELECT AVG(list_price) FROM products x WHERE x.category_id = p.category_id) AND p.cost_price < (SELECT AVG(cost_price) FROM products y WHERE y.brand_id = p.brand_id)"} {"db": "hard_retail", "question": "哪些客户的下单次数超过了所有下过单客户的平均下单次数?列出客户姓名和下单次数,按次数从多到少、相同按客户ID排。", "sql": "WITH t AS (SELECT customer_id, COUNT(*) AS cnt FROM sales_orders GROUP BY customer_id) SELECT c.customer_name, t.cnt FROM t JOIN customers c ON c.customer_id = t.customer_id WHERE t.cnt > (SELECT AVG(cnt) FROM t) ORDER BY t.cnt DESC, t.customer_id"} {"db": "hard_retail", "question": "哪些仓库的库存总量高于所有有库存仓库的平均库存总量?列出仓库名称和库存总量。", "sql": "WITH t AS (SELECT warehouse_id, SUM(quantity) AS qty FROM inventory GROUP BY warehouse_id) SELECT w.warehouse_name, t.qty FROM t JOIN warehouses w ON w.warehouse_id = t.warehouse_id WHERE t.qty > (SELECT AVG(qty) FROM t)"} {"db": "hard_retail", "question": "卖出件数(明细数量之和)最多的前 5 个产品是哪些?件数相同时产品ID小的优先。列出产品名称、品牌名称和件数。", "sql": "SELECT p.product_name, b.brand_name, SUM(si.quantity) AS qty FROM sales_items si JOIN products p ON p.product_id = si.product_id JOIN brands b ON b.brand_id = p.brand_id GROUP BY p.product_id ORDER BY qty DESC, p.product_id LIMIT 5"} {"db": "hard_retail", "question": "有至少 2 条退货记录的客户是哪些?列出客户姓名和退货记录数,按记录数从多到少、相同按客户ID排。", "sql": "SELECT c.customer_name, COUNT(*) AS return_count FROM returns r JOIN sales_items si ON si.item_id = r.item_id JOIN sales_orders so ON so.order_id = si.order_id JOIN customers c ON c.customer_id = so.customer_id GROUP BY c.customer_id HAVING COUNT(*) >= 2 ORDER BY return_count DESC, c.customer_id"} {"db": "hard_retail", "question": "状态为「已完成」的订单里,哪个渠道的平均运费(shipping_fee)最高?列出渠道和平均运费。", "sql": "SELECT channel, AVG(shipping_fee) AS avg_fee FROM sales_orders WHERE status = '已完成' GROUP BY channel ORDER BY avg_fee DESC LIMIT 1"} {"db": "hard_retail", "question": "每个品牌最早上市(launch_date 最早)的产品是哪个?并列的都列出,给出品牌名称、产品名称、上市日期。", "sql": "SELECT b.brand_name, p.product_name, p.launch_date FROM products p JOIN brands b ON b.brand_id = p.brand_id WHERE p.launch_date = (SELECT MIN(launch_date) FROM products x WHERE x.brand_id = p.brand_id)"} {"db": "hard_retail", "question": "找出采购单数最多的仓库(并列第一的都算),它们所在的城市(去重)里一共有多少位客户?只输出客户数。", "sql": "WITH w AS (SELECT warehouse_id, COUNT(*) AS cnt FROM purchase_orders GROUP BY warehouse_id), top_city AS (SELECT DISTINCT wh.city FROM w JOIN warehouses wh ON wh.warehouse_id = w.warehouse_id WHERE w.cnt = (SELECT MAX(cnt) FROM w)) SELECT COUNT(*) AS customer_count FROM customers WHERE city IN (SELECT city FROM top_city)"} {"db": "hard_retail", "question": "哪些客户的每一笔订单金额都超过 50000?订单金额 = 明细「数量×单价」之和,只看有明细的订单。列出客户姓名和订单数。", "sql": "SELECT c.customer_name, COUNT(*) AS order_count FROM (SELECT so.order_id, so.customer_id, SUM(si.quantity * si.unit_price) AS amt FROM sales_orders so JOIN sales_items si ON si.order_id = so.order_id GROUP BY so.order_id) o JOIN customers c ON c.customer_id = o.customer_id GROUP BY o.customer_id HAVING MIN(o.amt) > 50000"} {"db": "hard_retail", "question": "退货原因为「质量问题」的退款总额最高的 3 个品牌是哪些?金额相同按品牌名称排。列出品牌名称和退款总额。", "sql": "SELECT b.brand_name, SUM(r.refund_amount) AS total_refund FROM returns r JOIN sales_items si ON si.item_id = r.item_id JOIN products p ON p.product_id = si.product_id JOIN brands b ON b.brand_id = p.brand_id WHERE r.reason = '质量问题' GROUP BY b.brand_id ORDER BY total_refund DESC, b.brand_name LIMIT 3"}