asr_seq.pl

#!/usr/bin/perl

# parse iqtree .state output
# perl asr_seq.pl foo.state output_file

use strict;
use warnings;
use feature 'say';

my $input_file = $ARGV[0]
   or die "usage: $0 <input file> <output file>\n";
my $output_file = $ARGV[1]
   or die "usage: $0 <input file> <output file>\n";

$input_file =~ /\.state$/ or die "not a .state file?";

my %node_positions = ();
my %nodes = ();

open my $ifh, '<', $input_file or die "Cannot open $input_file: $!";
open my $ofh, '>', $output_file or die "Cannot open $output_file: $!";

while (my $line = <$ifh>) {
  chomp $line;
  my @cols = split /\s+/, $line;
  if ($line =~ /^Node\s+Site/) {
    for my $i (0 .. $#cols) {
      my $col = $cols[$i];
      $node_positions{"$1"} = $i if $col =~ /p_(.)/;
    }
  } elsif ($line =~ /^Node(\d+)/) {
    my $node = $1;
    my $site = $cols[1];
    my $state = $cols[2];
    # my $prob = $cols[$node_positions{$state}];
    $nodes{$node + 0} .= $state;
    die "Assumption of increasing sites violated, contact me" if $site != length($nodes{$node});
  }
}

for my $node (sort {$a <=> $b} keys %nodes) {
  say $ofh $node, ": ", $nodes{$node};
}