Parsing Question

[JRyanCon]

I want to parse information out of a linux command but am having a bit of trouble getting it to do what I want. The values I want to gather can be any digits caps or lowercase and can end in (rev [a-z0-9]) or not. If they end in this value I want to gather everything up to but not including the (rev 01). But if they dont end in this value I want to gather the full line. for instance if I had:

===========

blah blah 1101 blah (rev 03)
another value here blah

===========

I would want to capture:

blah blah 1101 blah
another value here blah


Any tips or ideas on how to do this?

 

[ishnid]

Something like this (untested)?:

my @values = ( "blah blah 1101 blah (rev 03)", "another value here blah" );

my @captures = map /^([\w\s]+)(?:\(rev [a-z0-9]+\))?$/i, @values;

Your description isn't totally clear on what characters are allowed before the (rev 03) so I've assumed that word or space characters are acceptable and no others.

 

[chessbot]

Wouldn't this work? (Please correct my code if necessary)

my $line = "blah blah 1101 blah (rev 03)";
my $output = $line ~= m/(rev [a-z0-9]+))$//;

--Chessbot

"Violence is the last refuge of the incompetent." -- Asimov, Foundation

 

[mlibeson]

The following code will print all output lines removing the end of those lines where there is a rev code as you specified:

open(SC, "linx_command |");
while (<SC>) {
 chomp;
 s/\s+\(\s*[rR][eE][vV]\s*[a-zA-Z0-9]+\s*\)\s*$/;
 print "$_\n";
}
close(SC);

Michael Libeson

 

[rharsh]

Hmmm.. here's kind of an ugly looking match regex that should work if you don't want to modify the original output from the command.

m/(.+?)(\(rev \w+\))?$/i;
print "$1\n";

If you're not worried about modifying the output from the command, a substitution will probably work. Give this a shot (and it's a lot cleaner.)

s/\(rev \w+\)//i;