Shell Programming and Scripting

{ - for any input line.
line=""; - reset our output variable 'line' to empty.
for (i=1;i<=NF;i++) { - loop around the fields.
if ($i) { - if this field is not empty.
line = line ? line OFS $i : $i - if the 'line' variable is already set to something, append OFS (comma) and the current field to it. Otherwise just set it to the current field (this is so we don't get an extra OFS at the start of the line).
}
}
print line - print the string we just constructed.
} FS=, OFS=, file - set field separator & output field separators to ',', read input file 'file'.

---------- Post updated at 21:41 ---------- Previous update was at 21:16 ----------

If you want to restrict it by columns for certain first fields, you could do something like:

Code:

$ cat x.awk
$1 ~ /^A$/ {
   line="";
   for (i=1;i<=NF;i++) {
      if ($i || (! ((i >=6) && (i <= 29)))) {
         line = line ? line OFS $i : $i
      }
   }
   print line
}
$1 ~ /^B1$/ {
   line="";
   for (i=1;i<=NF;i++) {
      if ($i || (! ( ((i>=2) && (i<=5)) || ((i>=17) && (i<=29))))) {
         line = line ? line OFS $i : $i
      }
   }
   print line
}

$ awk -f x.awk FS=, OFS=, file
A,10,20,30,40,50,60
B1,,12,24,36,48,54,58,68,,,rak,,,,,
B1,,12,24,36,48,54,58,68,,,rak,
B1,,12,24,36,48,54,58,68,,,rak,
B1,,12,24,36,48,54,58,68,,,rak,

and similar for B2 and C.

Hi Carol, Thanks a lot, but how do i remove the ',' at end of the line, and is there
any other way to do this.

Can some one please help ??????

There are trailing commas because that's what you asked for - there are more than 29 fields in those lines. If you want to remove more then just change the field numbers the script is checking for.