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

Monks I have a regex question I would like to ask, probably a simple answer for you gurus out there. I have an hash value returned which can be one of two strings. 

Value A
 
CDATA[Starting SITUATION1 <311446961,2559575967> for K01.K01CSYSTE5.
 
and 
 
Value B
 
CDATA[Starting Enterprise situation SITUATION1 <311439572,1066403668> 
+for K01.K01CSYSTE5.

[download]

Now, I need to match certain parts of the string and pass them off as parameters, so for example, if value A is returned, I need to match the following values from the pattern below 

CDATA[Starting SITUATION1 <311446961,2559575967> for K01.K01CSYSTE5.
 
Starting (Param 1)
SITUATION1 (Param 2)
K01.K01CSYSTE5 (Param 3)

[download]

So I coded the following which works for value A 


if ($elementHash->{MSGTEXT} =~ /CDATA\[(\S+)\s+(\S+)\s+.+(on|for)\s+(\
+S+)\./)
{
 
    $status = $1 (Which as the value "Starting")
    $sitname = $2 (Which as the value "SITUATION1")
    $type = $4 (Which as the value "K01.K01CSYSTE5")
}

[download]

But I need to alter the regex pattern so I would also match param $1, $2 and $4 for value B

CDATA[Starting Enterprise situation SITUATION1 <311439572,1066403668> for K01.K01CSYSTE5.

Notice the extra bit of text "Enterprise situation" being return in value B.

So what I need is a universal pattern which will match 

 
Starting as $1 
SITUATION1 as $2
K01.K01CSYSTE5 as $4

[download]

for the possibility of either VALUE A or VALUE B string being returned Hope this makes sense

Replies are listed 'Best First'.
Re: RegEx pattern
by cdarke (Prior) on Aug 30, 2011 at 09:00 UTC
    Why not just make the 'Enterprise situation' text optional?
    if ($elementHash->{MSGTEXT} =~ /CDATA\[(\S+)\s+(?:Enterprise situation
+\s+)?(\S+)\s+.+(on|for)\s+(\S+)\./)

    [download]
    Notice that the group (?:Enterprise situation\s+) is a non-capturing parentheses group, which means that $1,$2,$4 are unaffected by the new group. You might also do the same with (?:on|for) which would mean that $4 becomes $3. This avoids the overhead of capturing text you don't need and, more importantly, means you can concentrate on the groups you are actually interested in.
[reply]
[d/l]
[select]

      Exactly what I needed, thank you

[reply]
Re: RegEx pattern
by duyet (Friar) on Aug 30, 2011 at 08:00 UTC
    You can also write a small program to do the job: 
    my $elementHash = { MSGTEXT1 => 'CDATA[Starting SITUATION1 <311446961,
+2559575967> for K01.K01CSYSTE5',
                    MSGTEXT2 => 'CDATA[Starting Enterprise situation S
+ITUATION1 <311439572,1066403668> for K01.K01CSYSTE5.',
                  };  

foreach my $key ( keys %{ $elementHash } ) 
{
    my @tmp = parse_str( $elementHash->{$key} );
    print Dumper( \@tmp );
}

exit;


sub parse_str
{
    my $str = shift;
    my @items = split ' ', $str;
    my @tmp = split '\[', $items[0];
    my $size = $#items;
    my $param2 = ( $size > 4 ) ? $items[3] : $items[1];
    ( $tmp[1], $param2, $items[ $size ] );
}

    [download]
    Dumped data: 
    $VAR1 = [
          'Starting',
          'SITUATION1',
          'K01.K01CSYSTE5'
        ];
$VAR1 = [
          'Starting',
          'SITUATION1',
          'K01.K01CSYSTE5.'
        ];

    [download]
[reply]
[d/l]
[select]

Back to Seekers of Perl Wisdom