Regexp_substr question

Question

Oracle 10g and Oracle 11g
with mike_test as
(select
'Dont want this Line
Name want this line
Name want this line
Dont want this line
Name want this line' xx
from dual)
select regexp_substr(xx,'Name(.*)') from mike_test
I'm looking to return
Name want this line
Name want this line
Name want this line
but can only get one occurrence. I keep trying to use the matching mode 'm' to search through all lines but I'm not having success.
thanks
Mike

Answer 1

This should also work. As long as the "Name" doesn't include a special char used to do some regexp_conversions. I used the char "§".
with mike_test as
(select
'Dont want this Line
Name want this line
Naem not want this line !
Dont want this line
Name want this line' xx
from dual)
select regexp_substr(xx,'^Name.*$',1,1,'m') first_result
   ,regexp_count(xx, '^Name.*$', 1,'m') number_of_lines
   ,regexp_replace(
       regexp_replace(
           regexp_replace(xx,'(Name)','§',1,0,'m')
           ,'^[^§](.*)$','',1,0,'m')
       ,'§','Name',1,0,'m') all_lines
from mike_test
ALL_LINES
Name want this line
Name want this line"The problem with regexp_substr is that it will deliver only one occurence of the string.
Regexp_replace can eliminate all occurences. However you can't easily say something like: Everthing that is not starting with "Name".
Therefore I first replace the "Name" Part by a single character § and then later used the [^§] not replacement to eliminate the not needed lines. Then replace the substitution character § back to the wanted "Name".
Edited by: Sven W. on Dec 3, 2012 10:56 PM