Read apps list from PSI-Probe HTML output and print as plain text

psiprobe-htmltoplain.pl

#!/usr/bin/perl

use strict;
use warnings;

use XML::LibXML;

die "Usage: $0 file.html\n" if @ARGV != 1;

my $dom = XML::LibXML->load_html(
        location        => $ARGV[0],
        recover         => 1,
        suppress_errors => 1,
);

my $xpath = '//div/table//tr';
my @nodes = $dom->findnodes($xpath)->to_literal_list;
shift @nodes;  #..strip header

foreach (@nodes) {
        my @a = split/[\r\n]+/;
        my @b = map { /^\s*(.+)\s*$/ ? $1 : $_ } @a;
        my @c = grep !/^\s*$/, @b;
        my $d = join(' -- ', @c);
        print $d, "\n";
}