Обсуждение: Oracle to Postgres migration
Hi,
I have below query used in Oracle but while migrating to Postgres this code is not supported in Postgres.
Could you please tell me how to resolve this?
SELECT p.id_po, p.line_number, replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C, regexp_replace(p.protocol_number,'([[:cntrl:]])','', 'g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS protocol_number
,replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C, regexp_replace(p.protocol_status,'([[:cntrl:]])','', 'g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS protocol_status
,replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C, regexp_replace(p.protocol_approver,'([[:cntrl:]])','', 'g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS protocol_approver
,max(p.protocol_date) AS protocol_date
,replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C, regexp_replace(p.protocol_nota,'([[:cntrl:]])','', 'g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS protocol_nota
,sum(coalesce(p.protocol_value,0)) protocol_value
FROM podl_extended_protocol p
where upper(p.protocol_status) not in ('REJEITADO','ELIMINADO')
group by p.id_po, p.line_number
,replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C, regexp_replace(p.protocol_status,'([[:cntrl:]])','', 'g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS protocol_status
,replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C, regexp_replace(p.protocol_approver,'([[:cntrl:]])','', 'g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS protocol_approver
,max(p.protocol_date) AS protocol_date
,replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C, regexp_replace(p.protocol_nota,'([[:cntrl:]])','', 'g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS protocol_nota
,sum(coalesce(p.protocol_value,0)) protocol_value
FROM podl_extended_protocol p
where upper(p.protocol_status) not in ('REJEITADO','ELIMINADO')
group by p.id_po, p.line_number
Thanks,
Bimal
Hi Bimal, On Fri, Dec 22, 2023 at 10:28 AM bimal maity <bimal.af2020@gmail.com> wrote: > I have below query used in Oracle but while migrating to Postgres this code is not supported in Postgres. I didn't try your query, but I guess it complains about RTRIM, because it should accept text as an argument. If it is a case, you can try to dig in direction of something like this: > SELECT p.id_po, p.line_number, replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C, regexp_replace(p.protocol_number,'([[:cntrl:]])','','g'))ORDER BY 1)::text (with explicit type casting) best regards, Ilya -- Ilya Kosmodemiansky CEO, Founder Data Egret GmbH Your remote PostgreSQL DBA team T.: +49 6821 919 3297 ik@dataegret.com
bimal maity schrieb am 20.12.2023 um 04:35:
> Hi,
>
> I have below query used in Oracle but while migrating to Postgres this code is not supported in Postgres.
> Could you please tell me how to resolve this?
>
> SELECT p.id_po, p.line_number, replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C,
regexp_replace(p.protocol_number,'([[:cntrl:]])','','g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS
protocol_number
> ,replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C,
regexp_replace(p.protocol_status,'([[:cntrl:]])','','g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS
protocol_status
> ,replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C,
regexp_replace(p.protocol_approver,'([[:cntrl:]])','','g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS
protocol_approver
> ,max(p.protocol_date) AS protocol_date
> ,replace(replace(replace(RTRIM(XMLAGG(XMLELEMENT(name C,
regexp_replace(p.protocol_nota,'([[:cntrl:]])','','g'))ORDER BY 1), ',' ) ,'</C><C>','/'),'<C>',''),'</C>','') AS
protocol_nota
> ,sum(coalesce(p.protocol_value,0)) protocol_value
> FROM podl_extended_protocol p
> where upper(p.protocol_status) not in ('REJEITADO','ELIMINADO')
> group by p.id_po, p.line_number
What exactly does it do? I have often seen the hack using xmlagg/xmlelement/regexp_replace to do some kind of poor
man'sunnest/string_agg.
If you tell us, what exactly the goal is, I am confident there is a better solution in Postgres.