繁体   English   中英

sql查询获取汇率

[英]sql query to fetch exchange rate

给定一个以当地货币存储销售额的销售表和一个包含货币兑换率的汇率表,以获取每个销售日期的美元总销售额,我需要查询

销售表:

Sales Date  Sales Amount    Currency       
01-JAN-16            500    INR    
01-JAN-16            100    GBP    
02-JAN-16           1000    INR    
02-JAN-16            150    GBP    
03-JAN-16           1500    INR  

汇率表:

Source Currency  Target Currency  Exchange Rate  Effective Start Date
INR              USD                      0.014  31-DEC-15     
INR              USD                      0.015  02-JAN-16     
GBP              USD                       1.32  20-DEC-15     
GBP              USD                       1.30  01-JAN-16     
GBP              USD                       1.35  10-JAN-16   

不知道我应该怎么做

我必须做两件事来匹配货币,然后检查相同日期的汇率或在 sales_date 之前

尝试这个:

with SOURCE as
(
select s1.*, coalesce(e1.Rate, 1) as ExRate, row_number() over(partition by e2.Source order by e2.StartDate desc) as r_num
from Sales s1
left join Exchange e2
on s1.Currency = e2.Source
and e2.StartDate <= s1.SalesDate 
)
select SOURCE.*, SalesAmount*ExRate as USDAmount
from SOURCE 
where r_num = 1

很久以前,Oracle引入了用于这种处理的分析功能-避免通常很昂贵的联接(与查询中的其他所有事物相比,处理时间较长)。

在这种类型的问题中,最有效的方法是在需要的地方使用null列值来union all两个表。 然后使用last_value函数,忽略空值,然后仅收集结果。

假设

  1. 汇率表具有多种目标货币的转换率,而不仅仅是美元。 测试数据没有说明这一点,但是在查询中,我仅选择了target_currency = 'USD'来允许这样做。 (实际上,如果表中只有目标= USD的汇率,则不需要目标货币列)。
  2. 汇率栏不可为空。 (不应该!)
  3. 对于每个交易日期,对于相同的货币,汇率表中都有一行具有相同或更早的有效开始日期的行。 这是应该在数据库级别上维护的跨表约束。

查询 (包括with子句中的测试数据-使用基表时不需要)

with
     sales ( sales_date, sales_amount, currency ) as (       
       select to_date('01-JAN-16', 'dd-MON-rr'),  500, 'INR' from dual union all
       select to_date('01-JAN-16', 'dd-MON-rr'),  100, 'GBP' from dual union all
       select to_date('02-JAN-16', 'dd-MON-rr'), 1000, 'INR' from dual union all
       select to_date('02-JAN-16', 'dd-MON-rr'),  150, 'GBP' from dual union all
       select to_date('03-JAN-16', 'dd-MON-rr'), 1500, 'INR' from dual
     ),
     exch_rate ( source_currency, target_currency, exchange_rate, effective_date ) as (
       select 'INR', 'USD', 0.014, to_date('31-DEC-15', 'dd-MON-rr') from dual union all
       select 'INR', 'USD', 0.015, to_date('02-JAN-16', 'dd-MON-rr') from dual union all
       select 'GBP', 'USD',  1.32, to_date('20-DEC-15', 'dd-MON-rr') from dual union all
       select 'GBP', 'USD',  1.30, to_date('01-JAN-16', 'dd-MON-rr') from dual union all
       select 'GBP', 'USD',  1.35, to_date('10-JAN-16', 'dd-MON-rr') from dual
     ),
     prep ( dt, amt, src_curr, x_rate ) as (
       select  sales_date, sales_amount, currency, null from  sales
       union all
       select  effective_date, null, source_currency, exchange_rate 
         from  exch_rate
         where target_currency = 'USD'
     ),
     with_x_rates ( dt, amt, src_curr, x_rate ) as (
       select dt, amt, src_curr,
              last_value (x_rate ignore nulls) 
                   over (partition by src_curr order by dt, x_rate) as x_rate
       from   prep
     )
select   dt as sales_date, amt as sales_amount, src_curr as currency,
         x_rate as exchange_rate, 
         amt * x_rate as sales_amount_in_usd
from     with_x_rates
where    amt is not null
order by sales_date, currency   --  if needed
;

输出

SALES_DATE SALES_AMOUNT CURRENCY EXCHANGE_RATE SALES_AMOUNT_IN_USD
---------- ------------ -------- ------------- -------------------
01-JAN-16           100 GBP              1.300             130.000
01-JAN-16           500 INR              0.014               7.000
02-JAN-16           150 GBP              1.300             195.000
02-JAN-16          1000 INR              0.015              15.000
03-JAN-16          1500 INR              0.015              22.500

5 rows selected.

注意:查询输出的每一列均具有其正确的数据类型(日期,数字等)。 格式化全部在SQL * Plus中完成。 我们不想格式化SQL查询中的数据(将其转换为字符串),因为这可能不是最终产品; 其输出可能会被进一步处理消耗。

这应该很容易理解

with ex as
(
select e.source_currency, e.exchange_rate, e.effective_start_date,
isnull(lead(dateadd(day, -1,cast(effective_start_date as date))) over(partition by source_currency order by effective_start_date ),cast('9999-12-31' as date)) eff_end_date
from exchange_rate e
) 
select s.sales_date, sum(round((s.sales_amount * e.exchange_rate),2)) amt
from ex e
 join sales_amount s
on e.source_currency = s.currency and
s.sales_date between e.effective_start_date and e.eff_end_date
group by s.sales_date
order by s.sales_date

暂无
暂无

声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.

 
粤ICP备18138465号  © 2020-2024 STACKOOM.COM