Shell Programming and Scripting

What do you want to do with the very long SQL statements? Do you want to indent them on second/third lines?

BTW: This is what perl was designed for:

Code:

#!/usr/bin/perl
format =
@######  @<<<<<<<<  @<<<<<<<<<<<<<<<  ^<<<<<<<<<<<<<<<<<<<<<<<<<<<
$pid, $status, $datetime, $sql
~~       ^<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<
$sql                                                                                           .
while (<>) {
  ($pid,$status,$datetime,$sql) = split(" ",$_,4);
  write;
}

pipe the output of awk into this program and save as your report. It will look real pretty

If you need more help with formatting, do man perlform

---------- Post updated at 06:45 PM ---------- Previous update was at 06:38 PM ----------

Jack, here's a break-down of the awk command:

Code:

awk 'NR==FNR{A[$1]=$2;next} A[$1]{$5=A[$1];print}'

In awk, NR is the total input lines seen, while FNR is the number of input lines seen in the current file. So this essentially means: if we're processing the first file, "do this" (the code in the first set of braces {...}).

That code sets an associate array to the value of the 2nd column (in the first file, remember), where the index is the 1st column -- which is common in both files.

Now do a "next" which means do not process any more code for the current line. This ensures the rest of the awk script is not executed for the first file.

So the first "pattern/program" applies to the first file -- and only the first -- while the second "pattern/program" applies to the second (and subsequent) file(s).

The second pattern/program looks at each line (in the second file) and if the first column is found in the array A, and if the value is not null or not blank, it runs the portion between the braces.

The code in the braces simply replaces the 5th field of that line with the contents of what was seen in line indexed by the first column in the first file.

thanks otheus, very well explained. one question though. "and if the value is not null or not blank"

which bit is doing the above validation?

The A[$1] before the code block: A[$1]{$5=A[$1];print}

nulls and blanks are considered false and won't run the code, anything else is true and will run the code block.

Thank You guys for the clear and detailed explanation.

man perlform has great options for formatting. I will try to use those options.