oracle 正则表达式 替换 ‘’：

select/*+parallel(8)*/

t.nameas报表名称,

t4.umidas上报人UM,

t4.created_dateas上报时间,

t4.updated_date,

t4.state,

t4.eoa_session_id,

t3.base_path,

t6.deptid_descr,

dbms_lob.substr(

REGEXP_REPLACE(replace(substr(t4.query_condition,

instr(t4.query_condition,'collspanpoint=')+14,

instr(t4.query_condition,

'}',

instr(t4.query_condition,

'collspanpoint='))-

instr(t4.query_condition,'collspanpoint=')-13),

'''',

''),

'[[:alpha:]]+\:',

'')

)

--dbms_lob.substr(substr(t4.query_condition,instr(t4.query_condition,'collspanpoint=')+14,instr(t4.query_condition,'}',instr(t4.query_condition,'collspanpoint='))-instr(t4.query_condition,'collspanpoint=')-13))

fromparp_report_infot,

parp_report_modulet3,

parp_report_eoa_session_detailt4,

(selecta.login_username,a.dept_code

From(select/*+parallel(8)*/

l.login_username,

l.dept_code,

row_number()over(partitionbyl.login_usernameorderbyl.created_datedesc)asordernum

Fromparp_user_login_logl

wherel.dept_codeisnotnull)a

wherea.ordernum=1)t5,

parp_paic_dept_infot6

wheret.id_report_info=t4.id_report_info(+)

andt.id_report_module=t3.id_report_module(+)

andt4.umid=t5.login_username

andt5.dept_code=t6.paic_unique_deptid

and(t3.base_pathlike'寿险数据采集%'ort3.base_pathlike'金管家数据采集%'or

t3.base_pathlike'行销数据采集%')

andto_char(t4.created_date,'yyyy-MM-dd')>='2017-07-01'

andto_char(t4.created_date,'yyyy-MM-dd')<='2018-06-30'

andt4.state='Y'

orderbyt4.created_datedesc