Shell Programming and Scripting

Or if the line number might vary, but it's the NULL alone on a line immediately after the START_DT line,

Code:

sed -i '/as START_DT/,/^NULL$/s/^NULL$/'"date '1999-01-01' as end_dt /" filename

If your sed doesn't support the -i option, you will need to store the results in a temporary file and move it back on top of the original file.

Hi era,
Its not working .(NULL alone on a line immediately after the START_DT line in all my scripts)

sed '/as START_DT/,/^NULL$/s/^NULL$/'"date '1999-01-01' as end_dt /" c.txt

insert into PRD_TYPE
(
Period_Cd,
Period_Desc,
Ind1,
IND2,
Start_Dt,
End_Dt
)
select
Period_Cd,
Period_Desc,
NULL,
NULL,
date '1900-01-01' as START_DT,
NULL
from PRD_TYPE_2;

Thanks,
MR

Then I guess the NULL is not really alone on the line; do you have spaces on either side? Fixing that is a trivial exercise. Or use the regex suggested earlier by learnbash, i.e. \<NULL\> instead of ^NULL$

Hi Era,

Sorry ,Its working (spaces infornt Null and removed ^ before null).Thanks for your help.How can do it same thing in perl



sed '/as START_DT/,/NULL$/s/NULL$/'"date '1999-01-01' as end_dt /" c.txt

Thanks,
MR

In Perl, the line range operator is different -- actually there are two different operators to choose from, ... or .. (I leave it to you to figure out from the documentation what their differences are).

Code:

perl -pe 'if (/as START_DT/ .. /NULL$/) { s/NULL$/'"date '1999-01-01' as end_dt / }" c.txt

Or in Perl, you can do matching across line boundaries with -0777:

Code:

perl -0777 -pe 's/(as START_DT,\s*\n\s*)NULL\n/$1'"date '1999-01-01' as end_dt\n/" c.txt

There is also a s2p script which ships with the Perl distribution which automatically translates sed scripts to Perl scripts (although the end result is not always particularly elegant, and sometimes even wrong, I hear).

Hi Era,

Thanks for your great help .

Thanks,
MR

Hi cfajohnson ,

I am checking the following sequence in my file,there are 3 tables last 3 columns having similar sequnce but
when calling the below script its not suppose to give any table name but its giving two table name.Please help


Standardized_Ind
New_Cd
Table_Name



FILE=$1
awk ' /^CREATE SET TABLE/ { tablename = $4 }
(/^Table_Name/ && last != "Standardized_Ind") ||
(/^New_Cd/ && last != "Table_Name" )
{ printf "Error: %s\n", tablename } { last = $1 } ' "$FILE" | uniq


calling unix script


#ab.sh b.txt

Error: XX
Error: XX2


b.txt

CREATE SET TABLE XX ,NO FALLBACK ,
NO BEFORE JOURNAL,
NO AFTER JOURNAL,
CHECKSUM = DEFAULT
(
Categ_Cd VARCHAR(10) CHARACTER ,
Categ_Desc VARCHAR(250) ,
Standardized_Ind CHAR(1) ,
New_Cd VARCHAR(20) ,
Table_Name VARCHAR(100) ) ;

CREATE SET TABLE XX1 ,NO FALLBACK ,
NO BEFORE JOURNAL,
NO AFTER JOURNAL,
CHECKSUM = DEFAULT
(
Categ_Cd1 VARCHAR(10) ,
Categ_Desc1 VARCHAR(250) ,
Standardized_Ind CHAR(1) ,
New_Cd VARCHAR(20) ,
Table_Name VARCHAR(100) ) ;

CREATE SET TABLE XX2 ,NO FALLBACK ,
NO BEFORE JOURNAL,
NO AFTER JOURNAL,
CHECKSUM = DEFAULT
(
Financial_Asset_Categ_Cd VARCHAR(10) ,
Financial_Asset_Categ_Desc VARCHAR(250) ,
Standardized_Ind CHAR(1) ,
New_Cd VARCHAR(20) ,
Table_Name VARCHAR(100) );

Thanks,
MR