tabs to html list

[Anonymous Monk]

converts 4-space (or tab) indented text into a html list of lists

#!/usr/bin/perl --

use strict;
use warnings;

use Data::Dumper;
use HTML::TreeBuilder;

Main(@ARGV);
exit(0);

sub Main {
    my $list = Blahblah( \*DATA );
    {
        my $tree = HTML::TreeBuilder->new();
        $tree->eof;

#~         $tree->look_down(qw!_tag body!)->push_content( @$list );
        $tree->look_down(qw!_tag body!)->push_content($list);
        print $tree->as_HTML( '><&' => "\t" ), "\n";    # tab indent
        $tree->delete;
        undef $tree;
    }

#~    warn Dumper($list), "\n";
} ## end sub Main

sub Blahblah {
    my $fh       = shift;
    my $root_ref = ['ul'];

#~     my $root_ref = [ ];
    my $curr_ref = $root_ref;
    my @prev_ref;
    my $curr = 0;
    my $last = 0;

    while ( readline $fh ) {
        next if !/\w/;    # skip blank lines
        chomp;
        $last = $curr;
        $curr =
          /^((?:    |\t)+)/
          ? length($1) / 4
          : 0;

        if ( $curr > $last ) {
            my $new = ['ul'];
            push @{
                ref $curr_ref->[-1]
                ? $curr_ref->[-1]
                : $curr_ref
              },
              $new;
            push @prev_ref, $curr_ref;
            $curr_ref = $new;
        } elsif ( $curr < $last ) {
            while ( $last != $curr ) {
                $last--;
                $curr_ref = pop @prev_ref;
            }
        } ## end elsif ( $curr < $last )

        s/^\s+//;
        push @$curr_ref, [ 'li', $_ ];
    } ## end while ( readline $fh )
    return $root_ref;
} ## end sub Blahblah

__DATA__
a
    1
    2
    3
b
    1
    2
        a
        b
            1
            2
                a
                b
                c
            3
        c
    3
c
d


    7
        7a
        7b
            7b1
            7b2
                7b2a
                7b2b
                7b2c
            7b3
        7c
    8

output

• a
  • 1
  • 2
  • 3
• b
  • 1
  • 2
    • a
    • b
      • 1
      • 2
        • a
        • b
        • c
      • 3
    • c
  • 3
• c
• d
  • 7
    • 7a
    • 7b
      • 7b1
      • 7b2
        • 7b2a
        • 7b2b
        • 7b2c
      • 7b3
    • 7c
  • 8

not the first time i wrote something like this, hopefully I won't reinvent again