neo1491 has asked for the wisdom of the Perl Monks concerning the following question:

sending a command to my storage server and parsing the response. i'm looking for any line in the FH that have the word SATA and has N/A written twice. Here's my RE that looks for a single N/A && SATA... any ideas? if ( $_ =~ /\bsata\b/i && /\bN\/A\b*\bN\/A\b/i )

Replies are listed 'Best First'.
Re: Looking for N/A twice
by FunkyMonk (Bishop) on Oct 21, 2007 at 18:42 UTC
    What about 
    $_ = "gjhgjhgjg sata kjhkhkjhkjhkj N/A khkjhkhkhkj N/A";

print "yeah"
    if m{\bsata\b}i && m{\bN/A\b.*\bN/A\b}i

    [download]

[reply]
[d/l]
[select]
Re: Looking for N/A twice
by polettix (Vicar) on Oct 21, 2007 at 18:53 UTC
    You don't need to explicitly use $_ in a pattern match, and you can use a single regexp (together with FunkyMonk's advice):
    if (m{ \bsata\b .*? \bN/A\b .*? \bN/A\b }ix) { ... }

    [download]
    I added the x modifier just to separate the different parts a bit and make it a little more readable (see perlre for more details).

    Flavio
    perl -ple'$_=reverse' <<<ti.xittelop@oivalf

    Io ho capito... ma tu che hai detto?
[reply]
[d/l]
[select]
      You can use a single regex regardless of order if you use look-ahead assertions.

      $ perl -le '
> $_ = "gjhgjhgjg sata kjhkhkjhkjhkj N/A khkjhkhkhkj N/A";
> print q{ok} if m{(?ix) (?=.*\bn/a\b.+\bn/a\b) (?=.*\bsata\b)};'
ok
$

      [download]

      Cheers,

      JohnGG

[reply]
[d/l]
        This struck me like a thunder: 
        m{(?ix) (?=.*\bn/a\b.+\bn/a\b) (?=.*\bsata\b)}

        [download]
        I thought that the span of application of (?ix) was the stuff inside the two parentheses (just like a contracted form of (?ix:)), but your regex clearly contradicts this. So I turned to the manpage and... it's there, nicely documented! I love this language.

        Flavio
        perl -ple'$_=reverse' <<<ti.xittelop@oivalf

        Io ho capito... ma tu che hai detto?
[reply]
[d/l]
[select]
      The OP didn't state the order that the three phrases will appear. Your regexp won't match N/A sata N/A, for example.

      Only neo1491 can say if this is what s?he wants.

[reply]
[d/l]
[select]
        You're right, after re-reading it, I have to admit that I was fooled... by your example :) Good catch anyway, thanks.

        Flavio
        perl -ple'$_=reverse' <<<ti.xittelop@oivalf

        Io ho capito... ma tu che hai detto?
[reply]
Re: Looking for N/A twice
by ambrus (Abbot) on Oct 21, 2007 at 20:27 UTC

    I recommend matching the same regexp twice: if (/\bsata\b/i && /\bN\/A\b/ig && /\bN\/A\b/ig)

[reply]
[d/l]
      Magic works :) 
      use re 'debug';
$_ = "one sata two N/A three N/A";


die if (/\bsata\b/i && /\bN\/A\b/ig && /\bN\/A\b/ig) ;
__END__


Compiling REx `\bsata\b'
size 5 Got 44 bytes for offset annotations.
first at 1
   1: BOUND(2)
   2: EXACTF <sata>(4)
   4: BOUND(5)
   5: END(0)
stclass `BOUND' minlen 4
Offsets: [5]
        1[2] 3[4] 0[0] 7[2] 9[0]
Compiling REx `\bN/A\b'
size 5 Got 44 bytes for offset annotations.
first at 1
   1: BOUND(2)
   2: EXACTF <N/A>(4)
   4: BOUND(5)
   5: END(0)
stclass `BOUND' minlen 3
Offsets: [5]
        1[2] 3[3] 0[0] 6[2] 8[0]
Compiling REx `\bN/A\b'
size 5 Got 44 bytes for offset annotations.
first at 1
   1: BOUND(2)
   2: EXACTF <N/A>(4)
   4: BOUND(5)
   5: END(0)
stclass `BOUND' minlen 3
Offsets: [5]
        1[2] 3[3] 0[0] 6[2] 8[0]
Matching REx `\bsata\b' against `one sata two N/A three N/A'
Matching stclass `BOUND' against `one sata two N/A three '
  Setting an EVAL scope, savestack=3
   0 <> <one sata two>    |  1:  BOUND
   0 <> <one sata two>    |  2:  EXACTF <sata>
                            failed...
  Setting an EVAL scope, savestack=3
   3 <one> < sata two>    |  1:  BOUND
   3 <one> < sata two>    |  2:  EXACTF <sata>
                            failed...
  Setting an EVAL scope, savestack=3
   4 <one > <sata two>    |  1:  BOUND
   4 <one > <sata two>    |  2:  EXACTF <sata>
   8 < sata> < two N/>    |  4:  BOUND
   8 < sata> < two N/>    |  5:  END
Match successful!
Matching REx `\bN/A\b' against `one sata two N/A three N/A'
Matching stclass `BOUND' against `one sata two N/A three N'
  Setting an EVAL scope, savestack=3
   0 <> <one sata two>    |  1:  BOUND
   0 <> <one sata two>    |  2:  EXACTF <N/A>
                            failed...
  Setting an EVAL scope, savestack=3
   3 <one> < sata two>    |  1:  BOUND
   3 <one> < sata two>    |  2:  EXACTF <N/A>
                            failed...
  Setting an EVAL scope, savestack=3
   4 <one > <sata two>    |  1:  BOUND
   4 <one > <sata two>    |  2:  EXACTF <N/A>
                            failed...
  Setting an EVAL scope, savestack=3
   8 < sata> < two N/>    |  1:  BOUND
   8 < sata> < two N/>    |  2:  EXACTF <N/A>
                            failed...
  Setting an EVAL scope, savestack=3
   9 <sata > <two N/A>    |  1:  BOUND
   9 <sata > <two N/A>    |  2:  EXACTF <N/A>
                            failed...
  Setting an EVAL scope, savestack=3
  12 <a two> < N/A th>    |  1:  BOUND
  12 <a two> < N/A th>    |  2:  EXACTF <N/A>
                            failed...
  Setting an EVAL scope, savestack=3
  13 < two > <N/A thr>    |  1:  BOUND
  13 < two > <N/A thr>    |  2:  EXACTF <N/A>
  16 <o N/A> < three >    |  4:  BOUND
  16 <o N/A> < three >    |  5:  END
Match successful!
Matching REx `\bN/A\b' against ` three N/A'
Matching stclass `BOUND' against ` three N'
  Setting an EVAL scope, savestack=3
  16 <o N/A> < three >    |  1:  BOUND
  16 <o N/A> < three >    |  2:  EXACTF <N/A>
                            failed...
  Setting an EVAL scope, savestack=3
  17 < N/A > <three N>    |  1:  BOUND
  17 < N/A > <three N>    |  2:  EXACTF <N/A>
                            failed...
  Setting an EVAL scope, savestack=3
  22 </A three> < N/A>    |  1:  BOUND
  22 </A three> < N/A>    |  2:  EXACTF <N/A>
                            failed...
  Setting an EVAL scope, savestack=3
  23 </A three > <N/A>    |  1:  BOUND
  23 </A three > <N/A>    |  2:  EXACTF <N/A>
  26 </A three N/A> <>    |  4:  BOUND
  26 </A three N/A> <>    |  5:  END
Match successful!
Died at same.regex.twice.pl line 5.
Freeing REx: `"\\bsata\\b"'
Freeing REx: `"\\bN/A\\b"'
Freeing REx: `"\\bN/A\\b"'

      [download]
[reply]
[d/l]
Re: Looking for N/A twice
by neo1491 (Beadle) on Oct 21, 2007 at 18:58 UTC
    sweet... this actually worked. THANKS! if ( $_ =~ /\bsata\b/i && /\bN\/A\b*\bN\/A\b/i )
[reply]
Re: Looking for N/A twice
by Cody Pendant (Prior) on Oct 22, 2007 at 04:59 UTC
    Off-topic, but I realised that you could search for two or more occurrences of a substring in a string with nested index() calls like this: 
    $str = 'N/A N/A blah blah blah SATA N/A blah blah N/A';

$substr = 'N/A';

if (index($str,$substr, index($str,$substr)+1 ) ) {
print "two or more occurrences of '$substr' in '$str'"
}

    [download]
    which is probably way faster than any RegEx method.

    Nobody says perl looks like line-noise any more
    kids today don't know what line-noise IS ...
[reply]
[d/l]

      No way. index($str,$substr, index($str,$substr)+1 practically always returns true, even if $str has $substr only once or not at all. I don't think it's way faster than regexen either.

[reply]
[d/l]

Back to Seekers of Perl Wisdom