Update Statement w/ Order By and other goodies

Question

this is a continuation of my previous post : Re: create update statement with joins/select Please see it for table structure and sample data.
Ok so this is building on my previous SQL question that was resolved by you brilliant SQLers out there. the below code will do exactly what i need it to do, but I was wanting less user interaction:
update offender_case_identifiers ociq
   set Identifier_type = 'ARN2'
where ociq.offender_book_id in
      (select offender_book_id from v_header_block
        where offender_id_display = :ppid)
    and ociq.comment_text != :parn
   AND ociq.identifier_type = 'ARN'I was only wanting the user to have to enter the ":pid" and have the script take care of the rest. I have found the best way to ensure that I'm finding the most recent "ARN" from the offender_case_identifiers (which is then going to need to be set to ":parn") is to run the following script:
select oci.identifier_type
, vhb.offender_id_display
,oci.comment_text
,vhb.offender_book_id
,oci.offender_book_id
from   offender_case_identifiers oci, v_header_block vhb
where
oci.offender_book_id = vhb.offender_book_id
and vhb.offender_id_display = :ppid
and oci.identifier_type like 'ARN'
order by comment_text descThe above code was the original starting point for my previous final SQL statement, listed first in this post, however, I have added the "ORDER BY" at the bottom of it. For additional sample data, here is the result of the above statement:
IDENTIFIER_TYPE     OFFENDER_ID_DISPLAY     COMMENT_TEXT     OFFENDER_BOOK_ID     OFFENDER_BOOK_ID_1
ARN     *0000382183*     01550228     *1,011,683*     1,011,683
ARN     *0000382183*     01550228     *1,011,683*     1,011,683
ARN     *0000382183*     01531810     *823,281*     823,281
ARN     *0000382183*     01531810     *823,281*     823,281
ARN     *0000382183*     01531810     *823,281*     823,281
ARN     *0000382183*     01531810     *823,281*     823,281
ARN     *0000382183*     01529717     *804,301*     804,301
ARN     *0000382183*     01525871     *759,982*     759,982
ARN     *0000382183*     01525871     *759,982*     759,982
ARN     *0000382183*     01516263     *688,100*     688,100
ARN     *0000382183*     01516263     *688,100*     688,100
ARN     *0000382183*     01516263     *688,100*     688,100
ARN     *0000382183*     01515789     *682,680*     682,680
ARN     *0000382183*     01510809     *624,140*     624,140
ARN     *0000382183*     01507681     *590,261*     590,261
ARN     *0000382183*     01507681     *590,261*     590,261
ARN     *0000382183*     01507681     *590,261*     590,261
ARN     *0000382183*     01484993     *454,927*     454,927
ARN     *0000382183*     01484993     *454,927*     454,927
ARN     *0000382183*     01484993     *454,927*     454,927
ARN     *0000382183*     01477592     *448,231*     448,231
ARN     *0000382183*     01477592     *448,231*     448,231
ARN     *0000382183*     01456544     *428,902*     428,902
ARN     *0000382183*     01456544     *428,902*     428,902
ARN     *0000382183*     01429778     *403,175*     403,175
ARN     *0000382183*     01427677     *401,055*     401,055
ARN     *0000382183*     01427677     *401,055*     401,055
ARN     *0000382183*     01427677     *401,055*     401,055
So, as you can see, I want the query to select from the above results, the top/1st record of "Comment_Text" and set it to the ":parn" WITHOUT prompting the user to enter it. Then I want the script to change all other "Identifier_Type" to 'ARN2' that do not share the first "Comment_Text"....so in this example, I would want all 'ARN' changed to 'ARN2' that do not have "Comment_text" = 01550228.
I hope this isnt too confusing, because it is to me. Thanks in advance for all the help.

Answer 1

Hi,
So you have a query that shows all rows, and you want to UPDATE the rows in that reuslut set that do not have the greatest value of comment_text. You can use the analytic DENSE_RANK function to number the distinct comment_texts, and then discard the rows that were assigned #1. the remained will be the ones you wqant to UPDATE.
Here's one way to do that:
UPDATE     offender_case_idetifiers
SET     identifier_type     = 'ARN2'
WHERE     (offender_book_id, comment_text)
     IN (
            WITH    got_r_num    AS
                SELECT  vhb.offender_book_id
             ,        oci.comment_text
             ,        DENSE_RANK () OVER (OREDER BY oci.comment_text  desc)
                             AS r_num
             FROM    offender_case_identifiers oci
             ,         v_header_block           vhb
             WHERE   oci.offender_book_id      = vhb.offender_book_id
             AND         vhb.offender_id_display   = :ppid
             AND        oci.identifier_type           = 'ARN'
            SELECT  offender_nook_id
            ,       comment_text
            FROM    got_r_num
            WHERE   r_num     > 1
AND     identifier_type     = 'ARN'
I hope this answers your question.
If not, post a little sample data (CREATE TABLE and INSERT statements, relevant columns only) for all tables involved, and also post the results you want from that data.
If you're asking about a DML statement, such as UPDATE, the sample data will be the contents of the table(s) before the DML, and the results will be state of the changed table(s) when everything is finished.
Explain, using specific examples, how you get those results from that data.
Always say which version of Oracle you're using (e.g., 11.2.0.2.0).
See the forum FAQ {message:id=9360002}