Question

试图优化我的代码而不是使用临时的，对SQL Server来说是新的代码，我编写的代码为我提供了我所需的信息

Create table FI (
tipodoc numeric,
nmdoc varchar(20),
fno numeric,
rdata datetime,
fnoft numeric,
design varchar(60);

Insert into Fi (tipodoc ,
nmdoc,
fno,
rdata ,
fnoft ,
design)
values(1,'FT','18',2017-31-12,null,null),
(2,'NC',1,2018-31-12,18,'FT 18 18-12-2017'),
(1,'FT',1,2018-01-01,null,null),
(2,'NC',1,2018-31-12,1,'FT 1 01-01-2018'), 
(1,'FT',2,2018-31-12,null,null),
(2,'NC',1,2019-31-12,5,'FT 2 31-12-2018')

select
Cod_final = convert(int,fi.fnoft+fi.tipodoc+fi.rdata),
fi.tipodoc,
fi.nmdoc,
fi.fno,
fi.rdata, 
fi.fnoft, 
ltrim(rtrim(CONVERT(varchar(30),replace(right(design,10 ),'.','-')))) as Ftdata
into tempFI
from fi
where fi.nmdoc Like '%Credito' and design like '%factura%';

select
tempFI.Cod_final,
fi.nmdoc,
fi.fno,
convert(date,fi.rdata) as DataDoc,
Fi.fnoft,
tempFI.rdata, 
tempfi.Ftdata
from fi
inner join tempfi on fi.fnoft=tempFI.fnoft
where convert(int,fi.fnoft+fi.tipodoc+fi.rdata) = tempFI.Cod_final  
and fi.tipodoc =3 and tempfi.ftdata like '%-%-%'
order by tempFI.Cod_final

go
DROP TABLE IF EXISTS tempfi

原始数据 https://docs.google.com/spreadsheets/d/1VCd8ZQVt6ztLNNhUb-FRj1WIAGW8zYZP3AZbx3GaGTs/edit?usp=sharing

Data Output

我假装和做过的简短说明

Tempfi.cod_final-自从出现问题以来，tipodoc多年来一直在重复自己相同的fnoft和tipodoc（这与nmdoc关联，有一个与文档类型关联的数字）都是int，所以我转换了rdata，这是创建文档时的datatime，这是我的新数字，因此我可以正确识别。

fno和fnoft-cod_final 42726是fno 8，源自fnoft 1268，来自不同文档类型tipodoc的fno中的数字相同

tempfi.rdata-我正在测试以获取创建文档时tempfi.codfinal日期时间的正确值，但我惨败

tempfi.FTdata。是一个包含fnoft数量和创建日期的字符串，因此我的tempfi.rdata应该与tempfi.ftdata

相同

感谢您的帮助，对不起，如果这有点混乱

更新 因此，在阅读了有关CTE的内容后，我已经完成了此操作

with cte_fi2 as
(
    select
        Cod_final = convert(int,fi.fnoft+fi.tipodoc+fi.rdata),
        fi.tipodoc,
        fi.nmdoc,
        fi.fno,
        fi.rdata,
        fi.eteliquido        
    from fi
    where fi.tipodoc = 1
),

cte_fi as 
( 
    select
        Cod_final = convert(int,fi.fnoft+fi.tipodoc+fi.rdata),
        fi.tipodoc,
        fi.nmdoc,
        fi.fno,
        fi.rdata,
        fi.fnoft, 
        ltrim(rtrim(CONVERT(varchar(30),replace(right(design,10 ),'.','-')))) as Ftdata
    from fi
    where fi.tipodoc=3 and design like '%factura%'
)



select
    fi.nmdoc,
    fi.fno as NºNC,
    convert(date,fi.rdata) as DataNC,
    Fi.fnoft as NºFT,
    cte_fi2.rdata as DataFTcte2,
    cte_fi.ftdata,

from fi
    inner join cte_fi on fi.fno=cte_fi.fno 
    inner join cte_fi2 on fi.fno = cte_fi2.fno
where  fi.tipodoc =3 and cte_fi.ftdata like '%-%-%'
order by cte_fi.Cod_final

我对ct2_fi2.rdata仍然有问题，它应该返回与cte_fi.tdata相同的值，ct2_fi2.rdata是fi.tipodoc = 1的日期创建猜我在内部联接中做错了什么。我还不明白

 Nmdoc     FNo   fi.rdata   Fnoft  cte_fi2.rdata   cte_fi.ftdata
   NC       3     2013-02-08  0     08-02-2013     2016-01-04

输出

使用替代离开而不是临时表将数据联接到自身

0 个答案: