#!/usr/bin/perl -w

##################################################
##
## External speller binary for gnuheter.com
##
## Takes a string and finds all missspelled
## words and wraps them in a span tag with
## suggestions in the title attrib and
## style class set to "speling".
##
##################################################

use strict;

use Lingua::Ispell qw( :all );
use Getopt::Long;

$Lingua::Ispell::path = "/usr/bin/aspell";

my ($dictionary, $html, $file, $string);
my $ret = GetOptions ('d=s', \$dictionary,
		      "html+", \$html,
		      "file=s", \$file,
		      "s=s", \$string
		      );

unless (defined $dictionary) {
    $dictionary = "english";
}

unless (defined $string) {
  if (defined $file || -s $file) {
      open(FILE, "$file");
      $string = join('',<FILE>);
      close(FILE);
  } else {
      print "WHA ($string)?\n";
      exit;
  }
}


my $html_result .= spell($string, $dictionary);
print $html_result;


sub spell {
    my $string = shift() || "";
    my $dictionary = shift() || "english";
    my $result = "Huh? No input.<br/>\n";
    my $result_string = $string;

    Lingua::Ispell::use_dictionary($dictionary);
#    Lingua::Ispell::allow_compounds(1);

    $string =~ s/[\n\r]/ /g;
    $string =~ s/<[^>]+>//g;
    ## warn ($string);
    my %checked;
    foreach my $r ( spellcheck( $string ) ) {
	next if (defined $checked{$r->{term}});
	$checked{$r->{term}} = 1;
	if ( $r->{'type'} eq 'ok' ) {
	    # as in the case of 'hello'
	    # $result .= "'$r->{'term'}' was found in the dictionary.\n";
	}
	elsif ( $r->{'type'} eq 'root' ) {
	    # as in the case of 'hacking'
	    $result .= "'$r->{'term'}' can be formed from root '$r->{'root'}'\n";
	}
	elsif ( $r->{'type'} eq 'miss' ) {
	    # as in the case of 'perl'
	    my $desc = "<span class='speling' title='" . "Förslag: " . join(', ', @{$$r{'misses'}}) . "'>" . $r->{term} . "</span>";
	    $result_string =~ s/$r->{term}/$desc/g;
	    # $result .= "'$r->{'term'}' was not found in the dictionary;\n";
	    # $result .= "Near misses: $r->{'misses'}\n";
	}
	elsif ( $r->{'type'} eq 'guess' ) {
	    # as in the case of 'salmoning'
	    $result .= "'$r->{'term'}' was not found in the dictionary;\n";
	    $result .= "Root/affix Guesses: $r->{'guesses'}\n";
	}
	elsif ( $r->{'type'} eq 'compound' ) {
	    # as in the case of 'fruithammer'
	    $result .= "'$r->{'term'}' is a valid compound word.\n";
	}
	elsif ( $r->{'type'} eq 'none' ) {
	    # as in the case of 'shrdlu'
	    $result .= "No match for term '$r->{'term'}'\n";
	}
	# and numbers are skipped entirely, as in the case of 42.
    }

    return "$result_string";

}