上週處理過 oracle、sql server 資料庫相關資料，發現其實它們的 sql 查詢語句有些是不太一樣的，比如行列轉置和將查詢結果插入新錶。本人還是比較願意寫 sql 語句的，網際網路的技術日新月異，前端後端各種框架、新技術每擱幾年就得更新一遍，根本學不過來，而 sql 語句這些多卻少有變化，值得好好學習。
此篇部落格將會根據後續的工作持續更新。
一、sql server
1、查詢表資料
select * from hzwater擷取欄位的前 6 位
select substring(t.cbmonth,0,7) as cbmonths from hzwater as t原表字段 mainid、user_name、user_addr、ysxz、cbmonth、biaodi、meter_no、pianqu
首先將 cbmonth 字段擷取前 6 位改為 cbmonths 字段，然後再增加幾行，增加幾列其實是按月份增加幾個字段，如需要增加 201601、201602、201603、201604、201605、201606、201607、201608、201609、201610......201812
使用 case......when......語句實現：
select meter_no as meter_no,mainid as mainid,user_name as user_name,user_addr as user_addr,ysxz as ysxz,pianqu as pianqu,
max(case cbmonths when '201601' then cbsl else 0 end) as '201601',
max(case cbmonths when '201602' then cbsl else 0 end) as '201602',
max(case cbmonths when '201603' then cbsl else 0 end) as '201603',
max(case cbmonths when '201604' then cbsl else 0 end) as '201604',
max(case cbmonths when '201605' then cbsl else 0 end) as '201605',
max(case cbmonths when '201606' then cbsl else 0 end) as '201606',
max(case cbmonths when '201607' then cbsl else 0 end) as '201607'
from [hzwater_guotu].[dbo].middle
group by meter_no,mainid,user_name,user_addr,ysxz,pianqu
order by meter_no,user_name desc
得到的結果是：
4、將查詢結果放入建立的新錶
使用
select * into newtablename from ...select * into middle from
(select t.mainid,t.user_name,t.user_addr,t.ysxz,substring(t.cbmonth,0,7)
as cbmonths,t.cbsl,t.biaodi,t.meter_no,t.pianqu from hzcbdata as t)b
1、查詢資料
select * from temp使用 decode 語句實現：
select * from (select 用電類別,使用者名稱,用電位址,
sum (decode(dfny, '201701' , 計費電量, 0 )) "201701",
sum (decode(dfny, '201702' , 計費電量, 0 )) "201702",
sum (decode(dfny, '201703' , 計費電量, 0 )) "201703",
sum (decode(dfny, '201704' , 計費電量, 0 )) "201704",
sum (decode(dfny, '201705' , 計費電量, 0 )) "201705",
sum (decode(dfny, '201706' , 計費電量, 0 )) "201706",
sum (decode(dfny, '201707' , 計費電量, 0 )) "201707",
sum (decode(dfny, '201708' , 計費電量, 0 )) "201708",
sum (decode(dfny, '201709' , 計費電量, 0 )) "201709",
sum (decode(dfny, '201710' , 計費電量, 0 )) "201710",
sum (decode(dfny, '201711' , 計費電量, 0 )) "201711",
sum (decode(dfny, '201712' , 計費電量, 0 )) "201712",
sum (decode(dfny, '201801' , 計費電量, 0 )) "201801",
sum (decode(dfny, '201802' , 計費電量, 0 )) "201802",
sum (decode(dfny, '201803' , 計費電量, 0 )) "201803",
sum (decode(dfny, '201804' , 計費電量, 0 )) "201804",
sum (decode(dfny, '201805' , 計費電量, 0 )) "201805",
sum (decode(dfny, '201806' , 計費電量, 0 )) "201806",
sum (decode(dfny, '201807' , 計費電量, 0 )) "201807",
sum (decode(dfny, '201808' , 計費電量, 0 )) "201808",
sum (decode(dfny, '201809' , 計費電量, 0 )) "201809",
sum (decode(dfny, '201810' , 計費電量, 0 )) "201810",
sum (decode(dfny, '201811' , 計費電量, 0 )) "201811",
sum (decode(dfny, '201812' , 計費電量, 0 )) "201812"
from tempdlt20191203 group by 用電類別,使用者名稱,用電位址)t
3、將表轉置並查詢
select * from (select 使用者名稱,用電位址,sum (decode(dfny, '201701' , 計費電量, 0 )) "201701",
sum (decode(dfny, '201702' , 計費電量, 0 )) "201702",
sum (decode(dfny, '201703' , 計費電量, 0 )) "201703",
sum (decode(dfny, '201704' , 計費電量, 0 )) "201704",
sum (decode(dfny, '201705' , 計費電量, 0 )) "201705"
from tempdlt20191203 group by 使用者名稱,用電位址)t where t.使用者名稱='向陽';
create table newtablename as select * from b;除錯oracle, sql server儲存過程
關於儲存過程的除錯,知道方法以後很簡單,但在不知道的時候,為了測試乙個儲存過程的正確性,print,插入臨時表等可謂是使出了渾身解數,煩不勝煩 下面就把我工作中除錯儲存過程的方法,簡單的說明一下 除錯oracle儲存過程 環境 win2003 server oracle9i pl sql devel...
Oracle SQLServer中實現跨庫查詢
oracle sqlserver中實現跨庫查詢 一 在sqlserver中連線另乙個sqlserver庫資料 在sql中，要想在本地庫中查詢另乙個資料庫中的資料表時，可以建立乙個鏈結伺服器 exec master.dbo.sp addlinkedserver server n 別名 srvprodu...
Oracle SQLServer中實現跨庫查詢
一 在sqlserver中連線另乙個sqlserver庫資料 exec master.dbo.sp addlinkedserver server n 別名 srvproduct n 庫名 provider n sqloledb datasrc n 伺服器位址 exec master.dbo.sp a...