Perl: search index of element using regex

Question

I have a file that contains several columns separated with \t, and I want to get the position (index) of some columns by using regular expressions to specify the name of header of columns. Here is my code, but I can't search using regular expressions:

#!"C:\xampp\perl\bin\perl.exe"
my %dic;
my %index;
while(<>) {
    my @array2 = split(/\t/, $_);
    @index{@array2} = (0..$#array2);

    my $column13= $index{"name13"};// Here I want to search  using regex
    my $column17= $index{"name17"};// Here I want to search  using regex
    my $column21= $index{"name21"};// Here I want to search  using regex
    my $column34= $index{"name32"};// Here I want to search  using regex
    my $column43= $index{"name43"};// Here I want to search  using regex

    print $array2[$column13]$.",".$array2[$column17].",".$array2[$column21].
          ",".$array2[$column34].",".$array2[$column43]."\n"; 
}

For example the value of $columns13 should be 12 (position 12) and:

 $column17 = 16
 $column21 = 20
 $column34 = 33
 $column43 = 42

My input is a file that contains several columns separated with \t:

name1   name2   name3...    name85
1   2   3       4   ....     765
6   5   9       67  ....      8768
87  787 767     7687 ......   8768

My output should contains only the colums that have been searched:

name13  name17  name21...   name43
    876 76  87      4  .... 87687
   787  987 9       67  ...  87686
    53  765 767     7687 .... 8686

Perhaps you should give an example of what you are trying to do, with input and expected output. — TLP
– TLP, Commented Dec 25, 2014 at 15:38
Your print statement print $array2[$column13].",".... can be written much simpler with an array slice and join: print join(",", @array2[$column13, $column17, ...]), "\n"; — TLP
– TLP, Commented Dec 25, 2014 at 15:41
I want only search specified name using regex in headr of columns if match then it should give my the position of this column — Imii Iik
– Imii Iik, Commented Dec 25, 2014 at 15:50

Borodin · Accepted Answer · 2014-12-25 16:22:09Z

1

You're specification is rather sloppy, but I think this will do as you ask. It takes the first non-blank line from the input as the header line, and creates a list of corresponding indices in @indices. The corresponding columns from each subsequent are printed to STDOUT.

use strict;
use warnings;

my @selection = qw(
    name1
    name3
    name85
);

my @indices;

while (<>) {
  next unless /\S/;
  chomp;
  my @fields = split /\t/;

  unless (@indices) {
    @indices = grep {
      my $i = $_;
      grep { $fields[$i] =~ /$_/ } @selection;
    } 0 .. $#fields;
  }

  print join("\t", @fields[@indices]), "\n";
}

answered Dec 25, 2014 at 16:22

Borodin

127k9 gold badges72 silver badges146 bronze badges

Sign up to request clarification or add additional context in comments.

14 Comments

Imii Iik Over a year ago

Thanks for your response, but I don't have an array that contains the header, the header is the first line in the file who I give as argument

Imii Iik Over a year ago

And what is @selection please?

Borodin Over a year ago

I can't help you if you won't answer where your list of wanted column headers comes from.

Imii Iik Over a year ago

I have an input file, the first line is considered as headers, after headers comes others lines that contains values, and I execute my code it works i have any other inputs, I want to search the position of columns by specifyin the name of a header

Borodin Over a year ago

Okay, so if you have no other inputs then there is no way to specify the names of the headers.

|

Collectives™ on Stack Overflow

Perl: search index of element using regex

1 Answer 1

14 Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

14 Comments

Your Answer

Sign up or log in

Post as a guest

Related