Shell Programming and Scripting

Hi Gurus,
I know this'll be simple task for all the geeks out here but me being a newbie is finding it hard to crack this shell.

Ok coming to the task I've a delimited file as below

==================================================
====================================================

I've to select top 'N" records based on column 4 (aaaaa"),
so the output should be as below if "N" is 2

==================================================
==================================================

The next task is if the first 3 characters of column 6 = "SQL" then it should replace column 6 with "SQL" or if column 6 contains "SELECT" then it should replace it with "SELECT" or if none of the above condtion is satisfied then it should replace column 6 with "NULL".
So the output will be somthing like below

========================================================
[========================================================

Any help is greatly appreciated.

Thanks

Code:

awk -F\" '
count[$4]++ < n {
  if ($6 ~ /^(SQL|SELECT)/)
    $6 = $6 ~ /^SQL/ ? "SQL" : "SELECT"
  else 
    $6 = "NULL"
  print 
  }' OFS=\" n=2 infile

Thank you so much radoulov you are such a genius...

Hi radoulov,
Will the above code work for this example below to display SELECT as the output

132""

Thanks

No,
would you want the script to display the above as select statement?

You could use something like this:

Code:

awk -F\" '
count[$4]++ < n {
  if ($6 ~ /^(SQL|SELECT)|select/)
    $6 = $6 ~ /^SQL/ ? "SQL" : "SELECT"
  else 
    $6 = "NULL"
  print 
  }' OFS=\" n=2 infile

Or, if you want to capture the patterns sql and select in any position in the sixth field case-insensitively with GNU awk:

Code:

awk -F\" '
count[$4]++ < n {
  if ($6 ~ /(sql|select)/)
    $6 = $6 ~ /^sql/ ? "SQL" : "SELECT"
  else 
    $6 = "NULL"
  print 
  }' OFS=\" n=2 IGNORECASE=1 infile

The IGNORECASE variable is a GNU awk extension!

Again and again you are proving that you are a genius. I really appreciate your valueble contribution.

Thanks once again...

Hi radoulov,
I've one final query,in the logic to remove records whose count is more than 'n' based on column4 if I've to include one more column say column2 will the query change as count[$2,$4]++ < n ?

Thanks,