Shell Programming and Scripting

Hi,
Need awk help to group and print lines to format the output as shown below

INPUT FORMAT
set echo on
set heading on
set spool on
/* SCHEMA1 */ CREATE TABLE T1;
/* SCHEMA1 */ CREATE TABLE T2;
/* SCHEMA1 */ CREATE TABLE T3;
/* SCHEMA1 */ CREATE TABLE T4;
/* SCHEMA1 */ CREATE TABLE T5;
/* SCHEMA2 */ CREATE TABLE T1;
/* SCHEMA2 */ CREATE TABLE T2;
/* SCHEMA2 */ CREATE TABLE T3;
/* SCHEMA3 */ CREATE TABLE T1;
/* SCHEMA3 */ CREATE TABLE T2;
/* SCHEMA3 */ CREATE TABLE T3;
/* SCHEMA4 */ CREATE TABLE T1;
set feedback on
set spool off

DESIRED OUTPUT
set echo on
set heading on
set spool on
connect SCHEMA1
/* SCHEMA1 */ CREATE TABLE T1;
/* SCHEMA1 */ CREATE TABLE T2;
/* SCHEMA1 */ CREATE TABLE T3;
/* SCHEMA1 */ CREATE TABLE T4;
/* SCHEMA1 */ CREATE TABLE T5;
connect SCHEMA2
/* SCHEMA2 */ CREATE TABLE T1;
/* SCHEMA2 */ CREATE TABLE T2;
/* SCHEMA2 */ CREATE TABLE T3;
connect SCHEMA3
/* SCHEMA3 */ CREATE TABLE T1;
/* SCHEMA3 */ CREATE TABLE T2;
/* SCHEMA3 */ CREATE TABLE T3;
connect SCHEMA4
/* SCHEMA4 */ CREATE TABLE T1;
set feedback on
set spool off
I am looking forward to add the connect lines where all the SQL lines start with /* SCHEMAn */ values as a consistent pattern

i was able to sort the distict values of the pattern using the below awk

nawk '/^\/\* /{dup[$2]++;if (dup[$2]>=1){name[$2]=$2}};END{for(i in name)print name[i]} ' inputfile

Output
-------
SCHEMA1
SCHEMA2
SCHEMA3
SCHEMA4

but i am not sure how to place these lines in before the patterns and skip till the next pattern is found.

Thanks in advance,
Rajan

This should work for you:

Code:

awk '/SCHEMA/ && a != $2{a=$2;print "connect "a}1' file

Worked perfect.. Thanks a lot....

nawk '{
if(index($2,"SCHEMA")!=0 && temp!=$2)
{
print "connect",$2
temp=$2
print $0
}
else
print
}' file

Hi Danmero! i got a doubt here what does '1' do in the below code, what all i can use in place of 1. I am using solaris 10, and the above is not working in my system.
"awk '/SCHEMA/ && a != $2{a=$2;print "connect "a}1' file"

On Solaris use /usr/xpg4/bin/awk