SED script - modification

[joeroe3380]

Futurlet,
Thanks alot for the superb script. Though the above awk does most of the tasks required, it however doesnt change the occurance of B. to " " as required. Also i need to
replace the occurance of
'(B.In_DT,'YYDD')' to '(OT_DT,'YYDDD')In_DT'.
Please advice on modifying the awk for the above changes. Also i would deeply appreciate if you could briefly explain the above awk. I tried referring to the syntax inorder to understand the awk statement that you had written,but it looks a bit complicated.
awk '
NR==FNR {
if (skipping)
next
if ( $0 ~ /B_comb/ )
{ sub( / *B_comb.*/, "" )
skipping = 1
}
else
if ( $0 ~ /^B\.$/ )
next
print
next
}
1
' file1 file2 >OUTDAT


Thanks a lot again
Roe.

 

[futurelet]

[tt]
NR==FNR {
if (skipping)
next
sub( /\(B\.In_DT,'YYDD'\)/, "(OT_DT,'YYDDD')In_DT" )
sub( /B\./, "" )
if ( $0 ~ /B_comb/ )
{ sub( / *B_comb.*/, "" )
skipping = 1
}
print
next
}
1
[/tt]

 

[joeroe3380]

futurlet,
The modified awk worked fine for the sub( /B\./, "" ),
but the
sub( /\(B\.In_DT,'YYDD'\)/, "(OT_DT,'YYDDD')In_DT" )
doesnt still substitute for the desired one.
I have oin the first line :
select to_char(A.BASE_DT, 'YYDDD') || 'M-G' || which needs to be changed to
select to_char(CYC_DT, 'YYDDD') BASE_DT.
i modified the awk to :
awk '
NR==FNR {
if (skipping)
next
sub ( /to_char \(A\.BASE_DT, 'YYDDD'\)/ , "to_char(CYC_DT, 'YYDDD') BASE_DT" )
sub ( /A\./ , "")
if ( $0 ~ /to_char \(A\.BASE_DT, 'YYDDD'\)/ )
{ sub ( /to_char \(A\.BASE_DT, 'YYDDD'\)/ , "to_char(CYC_DT, 'YYDDD') BASE_DT" )
skipping = 1
}

if ( $0 ~ /COMBINED_RECORD/ )
{ sub ( / *COMBINED_RECORD*/, "" )
skipping = 1
}
} 1
' file1 file2 > out.sql

sed 's/to_char \(BASE_DT, 'YYDDD'\)/to_char(CYC_DT, 'YYDDD') BASE_DT/' outdat.sql

But still it doesnt change to the desired output.
Please advice in the above.
Thanks a lot.