//质量之星
select b.user_name as userName,
c.expert_id as expertId,
b.user_photo as userHeadIcon,
amout as codeSum,
rn as rank
from user_info b,(
select expert_id,amout,row_number() over(order by amout desc) as rn
from(
 select expert_id, count(expert_id) amout
  from code_infos a
  group by a.expert_id
 )
 ) c
 where b.user_id = c.expert_id and c.rn <= 50;

//

with 
sql0 as (SELECT USER_NAME, USER_ID FROM USER_INFO),
sql1 as (SELECT APPROVE_FLOW_ID ,ASSET_TYPE ,ASSET_ID, APPLI_OPERATOR ,CREATED_DT ,APPLI_OPINION FROM  ASSET_APPROVE_LOGS WHERE APPROVE_FLOW_ID = :approveFlowId AND PROCESS_ID = :processId),
sql2 as (SELECT PROJECT_ID, PROJECT_LABEL, DOMAIN_CODE ,PROJECT_NAME ,PROJECT_DESC ,RELEASE_TIME ,STATUS  FROM MODEL_PROJECT_INFO WHERE APPROVE_FLOW_ID = :approveFlowId),
sql3 as (SELECT DOMAIN_NAME, DOMAIN_CODE FROM ASSET_DOMAIN_INFOS)   

select 
   sql0.USER_NAME              AS userName, 
   sql1.APPROVE_FLOW_ID        AS approveFlowId,
   DECODE(sql1.ASSET_TYPE, 'A00001', '模型', 'A00002', '专题', 'A00003', '指标', 'A00004', '标签', 'A00005', '代码')
                                         AS assetType,
   sql1.ASSET_ID               AS assetId,
   sql1.CREATED_DT             AS appliTime, 
   sql1.APPLI_OPINION          AS appliOpinion,  
   sql2.PROJECT_NAME           AS assetName,    
   sql2.PROJECT_DESC           AS assetDesc,   
   sql2.RELEASE_TIME           AS releaseTime,   
   DECODE(sql2.STATUS, '0', '草稿', '1', '待审核', '2', '已发布', '3', '已下线', '4', '被驳回', '5', '被退回', '6', '已下发', '7', '已删除')  
                                         AS assetStatus,
   sql3.DOMAIN_NAME            AS domainName,
   sql2.PROJECT_LABEL          AS assetLabel            
from sql0  
join sql1 on sql0.USER_ID = sql1.APPLI_OPERATOR  
join sql2 on sql1.ASSET_ID = sql2.PROJECT_ID  
join sql3 on sql2.DOMAIN_CODE = sql3.DOMAIN_CODE  

//marge into的用法

//start  with的用法