mjhjmtu

I would like to order the 5th column by highest value and select the 4th if it matches the 5th column.

cat TABLE_LIST

C1 C2 C3 C4 C5 C6
3 No ENCRYPTION /opt/oracle/oradata/ORCLCDB/encryption.dbf 8 0 
2 No RETENTION /opt/oracle/oradata/ORCLCDB/retention.dbf 5 0 
4 No ORACLE /opt/oracle/oradata/ORCLCDB/oracle.dbf 2 0 
1 No USERS /opt/oracle/oradata/ORCLCDB/users01.dbf 3 0`

It should look like this.

for file in $C5; do
select $C4 from from dual;
done

select /opt/oracle/oradata/ORCLCDB/users01.dbf from dual;
select /opt/oracle/oradata/ORCLCDB/retention.dbf from dual;
select /opt/oracle/oradata/ORCLCDB/oracle.dbf from dual;
select /opt/oracle/oradata/ORCLCDB/encryption.dbf from dual;`

What I have tried so far is working but it read the output according to the value of C1, but I want to read according to C5, highest to lowest.

for (( x=1; x <= $FILE_COUNT; x++)) ; do
 FILE_NAME=cat $TABLE_LIST|sort -rk5 |awk -vx="$x" '$1 == x print $4 $3 == x print $1'
 FILE_SIZE=cat $TABLE_LIST|sort -rk5 |awk -vx="$x" '$1 == x print $5 $3 == x print $1'
done

Thanks!

According to comments you want to list the 4th column in the order dictated by the 5th column (in reverse numerical order).

$ sed '1d' <TABLE_LIST | sort -k5,5nr | awk ' print $4 '
/opt/oracle/oradata/ORCLCDB/encryption.dbf
/opt/oracle/oradata/ORCLCDB/retention.dbf
/opt/oracle/oradata/ORCLCDB/users01.dbf
/opt/oracle/oradata/ORCLCDB/oracle.dbf

The sed command strips the header from the file and the sort sorts the remaining lines on the 5th column in decreasing numerical order. The final awk extracts the 4th column.

This would work as long as none of the columns contained a whitespace character within the column itself.

Would you want to insert those select and from dual; strings, then you could modify the awk part of the pipeline:

$ sed '1d' <TABLE_LIST | sort -k5,5nr | awk ' printf("select %s from dual;n", $4) '
select /opt/oracle/oradata/ORCLCDB/encryption.dbf from dual;
select /opt/oracle/oradata/ORCLCDB/retention.dbf from dual;
select /opt/oracle/oradata/ORCLCDB/users01.dbf from dual;
select /opt/oracle/oradata/ORCLCDB/oracle.dbf from dual;

Using GNU Awk > 4.0

gawk '
 FNR>1 a[$5] = $4 
 END 
 PROCINFO["sorted_in"] = "@ind_num_desc"; 
 for (i in a) print "select " a[i] " from dual;"
 
' TABLE_LIST 
select /opt/oracle/oradata/ORCLCDB/encryption.dbf from dual;
select /opt/oracle/oradata/ORCLCDB/retention.dbf from dual;
select /opt/oracle/oradata/ORCLCDB/users01.dbf from dual;
select /opt/oracle/oradata/ORCLCDB/oracle.dbf from dual;