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

$string = "KHI0339B__P_H_Vita_Korangi_Ind_A";
my $string_1 = "HGW6120A__S_Popalzai_Cross_A";
$string_1 =~ s/__[a-zA-Z]{1,2}_/__/g;

print ":$string_1:\n";

Output :-
:HGW6120A__gPopalzai_Cross_A:

[download]
those input I get it from a huge input file. I have only those 2 pattern types. what is the modification required to handle those both types ?

Replies are listed 'Best First'.
Re: question : regex
by moritz (Cardinal) on Oct 11, 2007 at 07:19 UTC
    If I understood your question correctly: s/__[a-zA-Z]{1,2}_(?:[a-zA-Z]_)?/__/

    Perl 6 in German

[reply]
[d/l]
Re: question : regex
by johngg (Canon) on Oct 11, 2007 at 10:20 UTC
    I'm not sure if I've understood your requirement as I can't tell where the 'g' comes from in your output. However, this should get rid of any 'X_' occurances after a '__'.

    use strict;
use warnings;

my @strings = qw{
   KHI0339B__P_H_Vita_Korangi_Ind_A
   HGW6120A__S_Popalzai_Cross_A
   };

my $rxPatt = qr{(?<=__)[A-Z]_};
foreach my $string ( @strings )
{
    print qq{Original: $string\n};
    $string =~ s{$rxPatt}{}
       while $string =~ $rxPatt;
    print qq{Modified: $string\n\n};
}

    [download]

    The output is

    Original: KHI0339B__P_H_Vita_Korangi_Ind_A
Modified: KHI0339B__Vita_Korangi_Ind_A

Original: HGW6120A__S_Popalzai_Cross_A
Modified: HGW6120A__Popalzai_Cross_A

    [download]

    I hope this is of use.

    Cheers,

    JohnGG

[reply]
[d/l]
[select]
Re: question : regex
by jesuashok (Curate) on Oct 11, 2007 at 07:11 UTC 
    s/__[a-zA-Z]{1,2}_[a-zA-Z]{0,1}/__/g

    [download]


    i m possible
[reply]
[d/l]
Re: question : regex
by Anonymous Monk on Oct 11, 2007 at 08:10 UTC
    from the examples given, my guess is that the need is for any sub-string matching  "__A_" or  "__A_A_" (where  A is any alpha character) to be replaced by  "__" (a double-underscore). (i assume the "g" in "__gPopalzi" is a typo.)

    in that case, this may be appropriate (UNTESTED): 

    $string =~ s/ __ (?: [a-zA-Z]_ ){1,2} /__/xms;

    [download]

    i also assume there is only one instance of the sub-string in a string, and so the  /g regex modifier is not needed. also, that variations on the sub-strings  "__AA_" and  "__AA_AA_" will not occur.

    note: more examples would help the humble monks to better understand the requirements.

[reply]
[d/l]
[select]
      actually, "any sub-string matching" in the first sentence should be "the first sub-string matching".
[reply]

Back to Seekers of Perl Wisdom