Sql server 连接3个表并按最大列值分组

谢谢你回答我之前的问题。我希望这样做没问题，但我现在有另一个相关问题：

我现在被要求扩展我正在做的工作，包括将以当地货币持有的格罗斯科斯特转换为英镑的基础货币

我分别拥有这两部分的SQL，但不知道如何组合它们。这两部分是：

获取每家公司的总成本：

SELECT    *
    FROM (SELECT  CompanyName,CurrencyCode,ExchangeRate3,ActiveDate,  
    rank() OVER 
    (partition BY  CompanyName,dbo.Currencies.CurrencyId
    ORDER BY ActiveDate DESC) AS Rank
    FROM  dbo.company 
    INNER JOIN dbo.Currencies 
    ON dbo.company.CurrencyId = dbo.Currencies.CurrencyId 
    INNER JOIN   dbo.CurrencyRates 
    ON dbo.Currencies.CurrencyId = dbo.CurrencyRates.CurrencyFromId) ci
WHERE     ci.Rank = 1

选择Cast（ROW_NUMBER（）OVER（ORDER BY Sum（JobCosting.PurchaseOrderItems.GrossCost）Desc）作为Int）作为“RowNum”，
Admin.companys.UniqueName为“UN”，
总计（JobCosting.PurchaseOrderItems.GrossCost）为“DonutValue”
来自管理公司
Admin.companys.CompanyId上的内部联接JobCosting.PurchaseOrders=
作业成本计算.采购订单.供应商公司ID
内部联接JobCosting.PurchaseOrderItems
关于JobCosting.PurchaseOrders.PurchaseOrderId=
JobCosting.PurchaseOrderItems.PurchaseOrderId
其中UniqueName像“HH%”
按唯一名称分组
按“DonutValue”说明订购

获取每种货币的货币转换率。换算为相关货币的GrossCost*汇率。这是SQL的汇率（由SQLGRL提供-再次感谢）：

其中ci.Rank=1

我需要做的是得到一个组合SQL，它返回所有相同的行，并计算每个公司的Grossales总和的转换值

再说一次，任何帮助都会非常感激，因为我真的被卡住了

---

提前感谢

使用类似的方法-

WITH cte AS (
    SELECT c.*, --column_list
    ROW_NUMBER(ORDER BY cur_from.ActiveDate desc) row_num 
    FROM Company c 
    INNER JOIN Currencies cu ON c.CurrencyId = cu.CurrencyId
    INNER JOIN CurrencyRates cu_from ON cu.CurrencyId = cu_from.CurrencyFromId 
    -- WHERE CONDITION if any
)

SELECT * FROM cte WHERE row_num = 1

---

这将为您提供每家公司每种货币代码的最新汇率：

SELECT    *
    FROM (SELECT  CompanyName,CurrencyCode,ExchangeRate3,ActiveDate,  
    rank() OVER 
    (partition BY  CompanyName,dbo.Currencies.CurrencyId
    ORDER BY ActiveDate DESC) AS Rank
    FROM  dbo.company 
    INNER JOIN dbo.Currencies 
    ON dbo.company.CurrencyId = dbo.Currencies.CurrencyId 
    INNER JOIN   dbo.CurrencyRates 
    ON dbo.Currencies.CurrencyId = dbo.CurrencyRates.CurrencyFromId) ci
WHERE     ci.Rank = 1

编辑以包含请求的其他信息：

with cte as (
Select Cast(ROW_NUMBER() OVER 
(ORDER BY Sum(JobCosting.PurchaseOrderItems.GrossCost) Desc) AS Int) As 'RowNum', 
Admin.Companies.UniqueName As 'UN', 
Sum(JobCosting.PurchaseOrderItems.GrossCost) As 'DonutValue' 
From Admin.Companies 
Inner Join JobCosting.PurchaseOrders 
On Admin.Companies.CompanyId = JobCosting.PurchaseOrders.SupplierCompanyId 
Inner Join JobCosting.PurchaseOrderItems 
On JobCosting.PurchaseOrders.PurchaseOrderId  =JobCosting.PurchaseOrderItems.PurchaseOrderId 
Where UniqueName Like 'HH %' Group By UniqueName Order By 'DonutValue' Desc
)

SELECT    *
FROM (SELECT  CompanyId, UniqueName,CurrencyCode,ExchangeRate,FromCurrencyID, 
ToCurrencyId, ActiveDate,  
rank() OVER 
(partition BY  UniqueName,Admin.Currencies.CurrencyId
ORDER BY ActiveDate DESC) AS Rank
FROM  Admin.Companies
INNER JOIN Admin.Currencies 
ON Admin.Companies.CurrencyId = Admin.Currencies.CurrencyId 
INNER JOIN   Admin.CurrencyRates 
ON Admin.Currencies.CurrencyId = Admin.CurrencyRates.FromCurrencyId
WHERE Admin.CurrencyRates.ToCurrencyId = 47) as ci
INNER JOIN   cte
on cte.UN = ci.UniqueName
where    ci.Rank = 1;

---

此解决方案使用公共表表达式。但是，通过创建视图并将视图加入排名查询，您也可以获得相同的结果。

非常感谢，这正好满足了我的需要。在您的帮助下，挑战已经扩大，我再次陷入困境。任何帮助都将受到感激。