Shell Programming and Scripting

Hi All,

I need to search for a string in a file that I've opened and base a decision on the result. The logic is this:

"if the word 'Shared' appears on the first line then

do this on the whole file

else

do this on the whole file

"

The code I currently have isn't working:

Code:

           
if ($. = 1 && $_ =~ /^*Shared*$/) {

             $_ if ($awk - $F) . "\",\"";if ($. != 1) {
             $n = substr($Fld[11], 1, length($Fld[11]) - 2);
             $lines{$. - 1} = $n;

             }
           } else {

             $_ if ($awk - $F) . "\",\"";if ($. != 1) {
             $n = $Fld[10];
             $lines{$. - 1} = $n;

             }
           }

The pattern matching in this code was provided by using a2p (I know awk better than I know perl)...

Thanks in advance.

Hi.

I hope you will have a lot of fun and reward with perl -- I certainly have.

Briefly, regular expressions are not like filename expressions, so "d*" in perl (as in "... shared*") means "0 or more d's", not "d followed by anything".

I suggest you provide:

1) a small sample input that fails, and / or:

2) be more precise by "isn't working" ... cheers, drl

( edit 1: corrected explanation )

Hi drl,

Thanks for your reply - I'm certainly finding Perl more productive than using bash shell scripting but it's a little harder to learn - I'm sure I'll get there tho

The match I'm looking for is actually in the top line of the file and not the filename itself:

There are two types of input files (output from sar) because the systems I'm extracting the data from are configured slightly differently. Hence the need to work out which type of file I'm looking at...

So when I've found out whether the work 'Shared' exists in the header line I want to perform one of two possible options on the rest of the file.

Hope this makes my question clearer?

Thanks, p.

Hi All,

I think I've found the solution to my problem. Here's how I found the answer:

1. I knew how to do it in awk:

Code:

$ cat awk1
NR = 1 && $0 ~ /Shared/ { print $0 }

2. So I used a2p to convert it into Perl:
3. I extracted the only bit that I'm interested in:

Code:

if ($. = 1 && $_ =~ /Shared/)

Basically the logic is matching where line number is equal to one AND where that lines contents contains 'Shared'.

Cheers, p.

you have not tried Python yet.