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

I have tab delimited text file,which i split with tabs store it in array and display.One of the column data has a tab in between,so instead of taking the column data as a whole it splits that also,here is my data 
data    name    aliasname,    alsoknown    class    type

[download]

output 

data
name
aliasname,
alsoknown
class
type

[download]
what i want is 
data
name
aliasname, alsoknown
class
type

[download]

here is the code that i have tried 

#!/usr/bin/perl
use strict;
my $var;

open(FH,"sample.txt")or die("can't open file:$!");
while($var=<FH>)
{
    my @vareach=split("\t",$var);
    for my $each(@vareach)
    {
        print "$each\n";
    }
}

[download]

any suggestions on how to split these kinds of data?

Replies are listed 'Best First'.
Re: The best way to split tab delimited file
by gmargo (Hermit) on Nov 16, 2009 at 15:35 UTC

    Perfect opportunity for a negative lookbehind. 

    my @vareach=split(/(?<!,)\t/,$var);

    [download]
[reply]
[d/l]
      Below is the data,where the regex is not working.. 
      my $var='474627    asidase ta sidase ala,"lpha-D- ctoside gtohydrolase
+","razyme","arazyme      (enz Corp)","Melie","lagal","idase bta",    
+   rug    00103';

      [download]
[reply]
[d/l]

        There are no tab characters in that line. Presumably due to a cut/paste issue. Can you try again to get the tabs in there? And also show the output you expect vs the output you are getting?

[reply]
Re: The best way to split tab delimited file
by BioLion (Curate) on Nov 16, 2009 at 15:27 UTC

    You should check out Text::Delimited for parsing delimited text!

    It allows you use column headers, extract columns, extract it into various datastructures etc... taking away a lot of the pain!

    For getting all the columsn back as an array ref (columns in orginal order) see the read() method and it's __DATA__ key, or you can ask for a particular column.

    Updated: Last para.

    Just a something something...
[reply]
[d/l]
[select]
Re: The best way to split tab delimited file
by Utilitarian (Vicar) on Nov 16, 2009 at 15:33 UTC
    if the comma is a fixed aspect of this file, then 
    @fields = split (/[^,]\t/,$record)

    [download]
    would probably do what you need, but BioLion's suggestion above might make your life easier further down the road, as more "little quirks" emerge in the data
[reply]
[d/l]

      By placing the comma in a negated character class you lose any character preceding the tab, apart from a comma, in the resultant array because it becomes part of the split term. Placing the comma in a negative zero-width look-behind, as gmargo does here, the characters are retained.

      $ perl -le '
> $txt = qq{abc\tdef,\tghi\tjkl\tmno};
> print for split m{[^,]\t}, $txt;
> print q{-} x 20;
> print for split m{(?<!,)\t}, $txt;'
ab
def,    gh
jk
mno
--------------------
abc
def,    ghi
jkl
mno
$

      [download]

      I hope this is of interest.

      Cheers,

      JohnGG

[reply]
[d/l]
Re: The best way to split tab delimited file
by keszler (Priest) on Nov 16, 2009 at 15:41 UTC
    split uses a regex, not a string. (It is accepting your tab-in-a-string as a regex, but IMHO it shouldn't.)

    Given that you can use a regex, it looks like in this particular case you want to split on tabs, unless the tab is preceeded by a comma. There are several ways to construct such a regex. A couple off the top of my head:

    • Negated character class comma preceeding tab
    • Negative lookbehind for comma

    See perlretut for more ideas

    update - too slow. Those two ideas were already mentioned.

[reply]
Re: The best way to split tab delimited file
by Anonymous Monk on Nov 17, 2009 at 23:37 UTC
[reply]

Back to Seekers of Perl Wisdom