Using regexp_replace

Question

Hello, I could use some assistance in using regexp_replace for the following results:
I have string: OUTLET_ID, RETAILER_ID which is stored in variable v_pk
I want to use regexp_replace to convert to v_pk to *'OUTLET_ID' || 'RETAILER_ID'*
Any ideas?
Thank you

Answer 1

JesusLuvR wrote:
Hello, I could use some assistance in using regexp_replace for the following results:
I have string: OUTLET_ID, RETAILER_ID which is stored in variable v_pk
I want to use regexp_replace to convert to v_pk to *'OUTLET_ID' || 'RETAILER_ID'*'OUTLET_ID' || 'RETAILER_ID' is the same as 'OUTLET_IDRETAILER_ID'.
Are you saying you just want to get rig of the comma? You could do that with REGEXP_REPLACE, but it's more efficient to use plain old REPLACE:
REPLACE ( v_pk
Any ideas?If you wanted to change that one string into two separate strings
(1) 'OUTLET_ID' and
(2) 'RETAILER_ID'
then you could do something like this:
REGEXP_SUBSTR ( v_pk
           , '[^,]+'
           , 1
           , n
           )to get the n-th comma-delimited item in v_pk (or NULL, if there are fewer than n-1 commas in v_pk).
This assumes that there is something, if only a single space, between commas. For example
WITH    test_data   AS
     SELECT  ',BAR' AS v_pk  FROM dual
SELECT  v_pk
,     REGEXP_SUBSTR (V_PK, '[^,]+', 1, 1)     AS part_1
,     REGEXP_SUBSTR (V_PK, '[^,]+', 1, 2)     AS part_2
FROM     test_data;produces:
V_PK     PART_1     PART_2
,BAR     BARIf you want something that recognizes part_1 as being NULL, and part_2='BAR', then it's more complicated (at least in versions earlier than Oracle 11.2):
LTRIM ( REGEXP_SUBSTR ( ',' || v_pk
                      , ',[^,]*'
                      , 1
                      , n
      )