If you want to convert Latin-1 to UTF-8 it is another question. Take a look on Unicode::String.

--
Ilya Martynov (http://martynov.org/)

sub latin1_to_utf8 {
return( pack( 'U*', unpack("C*", @_[0] ) ) );
}

sub utf8_to_latin1 {
return( pack("C*", unpack("U*",@_[0]) ) );
}
[download]

Please s/ascii/latin1/. Conversion from ASCII to UTF8 is nonsense. It is NOOP.

--
Ilya Martynov (http://martynov.org/)

I was needing something cheap that would work on both 5.6.x and 5.8.x versions. (I'm trying to keep the code the same while moving back and forth between systems) Tried your utf8_to_latin1() and ran into problems on 5.8.x. The below seems to work on both versions. Thanks for the code.

sub utf8_to_latin1 {
# return( pack("U0C*", unpack(  "U*",@_[0]) ) );
  return( pack(  "C*", unpack("U0U*",@_[0]) ) );
}
[download]


Updated: (see sig) So I go and feed the original (commented out) line into the big program and it fails. It worked in the test program on both 5.6 and 5.8. Wander all over and find the 5.8 perluniintro where they explicitly say


$native_string  = pack("C*", unpack("U*", $Unicode_string));
[download]

just like the original poster. But that didn't work alike on both Perl versions. Played around some more and hit upon the other variant above. This now works in both test and 'real' programs on both Perl versions. (sigh)

--
I'm a pessimist about probabilities; I'm an optimist about possibilities. - - Lewis Mumford

Why convert from Latin-1 to UTF-8 at all? As I understand your spellchecker uses UTF-8. All you have to do is ask browser to use same charset. Issue header Content-Type: text/html; charset=UTF-8 and you are done. Form input will be in UTF-8 and you can output UTF-8 too.

--
Ilya Martynov (http://martynov.org/)

1. see what it does
2. understand your request ;)

use strict ;
use CGI qw/:standard/;
use CGI::Carp qw(fatalsToBrowser);
use HTML::Template;
use Data::Dumper;
use Unicode::String;
use lemma;


my $q = new CGI;
my $template = "bbc/lemma/wsillafu.tmpl";
my $tmpl = new HTML::Template( filename =>$template,associate => $q );
+ 
# Consider using an array to make getting and setting easier.
# You could then use array slices to set them ;)
my $errorchecker = 0;
my $memoryoverloadchecker = 0;
my $memorychecker = 0;
my $dictionarychecker = 0;
my $otherchecker = 0;
my $memoryfinder = 0; 


print $q->header();
   
if ($q->param()) {
               
  my ( $spellingcheck, $error) = lookup($q->param("brawddeg")); # Is t
+his pronounced brahtheg?   
                    
  $tmpl->param( spellingcheck => $spellingcheck );

  ($errorchecker,$memoryfinder) = (1,1) if $error == 1;
  ($memoryoverloadchecker,$memoryfinder) = (1,1) if $error == -1;
  ($memorychecker, $memoryfinder) = (1,1) if $error == -2;
  ($dictionarychecker, $memoryfinder) = (1,1) if $error == -3;
  ($otherchecker,$memoryfinder) = (1,1) if $error == -4;   
}

print $tmpl->output;
     
if( $dictionarychecker ) {
  print "<br><br>Nid yw\'r geiriadur ar-lein ar gael.  E-bostiwch meis
+tr y wefan am gymorth \n";
}
     
if( $otherchecker ){
   print "<br><br>Mae gwall yn y gwirydd.  E-bostwich meistr y wefan a
+m gymorth \n";
}
     
if( $memoryoverloadchecker ){
   print "<br><br>Roedd mwy na 10 gwall yn y testun.  Dim ond y 10 gwa
+ll cyntaf sydd wedi ei cywiro.<br>
   Cywirwch rhain cyn cario ymlaen i gywiro.\n";
}
     
     
if( $memorychecker ) {
   print "<br><br>Mae'r gwirydd wedi rhedeg allan o gof.  E-bostiwch m
+eistr y wefan am gymorth.\n";
}

if( $errorchecker ) {
   print "<br><br><font color=\"red\"><b>Mae rhai gwallau yng nghorff 
+y testun.</b></font><br>
         Dewiswch air â gynhigir yn y blwch tynnu i lawr i bob gair sy
+dd wedi ei sillafu'n anghywir.
         <br>Yna gwagswch y botwm gwirio isod i gael eich testun wedi 
+ei wirio.<br>\n";
   print "<form name=\"ArgraffyddTestun\" method=\"POST\" Action=\"wsi
+llafudangos.pl\">\n";
   print "<input type=\"Submit\" name=\"Gwirio\" value=\"Gwirio\"></fo
+rm>\n";
}
      
unless( $memoryfinder ) {
   print "<br><br><b>Nid oedd gwall yn y frawddeg</b>\n";
}
     
sub lookup { # Don't  think you want a prototype as you had it ;)
    $_ = shift; # the @_ is redundant, it's implicit.
    my ($errors, $str) = lemma::GetCysillSpellingErrors($_, 10);
    my @fields = split ("(<awgrymiadau>[^<]*</awgrymiadau>)", $str);
    my $error = @fields > 1;
    my @spellingcheck = ();
    
    return (\@spellingcheck, 'odd amount of values') unless @fields % 
+2;
    
    # Remove the last two items
    my $last_field = pop(@fields);
    pop(@fields);

    my ($key,$value);
    while ( ($key,$value,@fields) = @fields ) {     
           $value =~ s/<[^>]*>//g;

           my @SuggestionList = split /,/, $value; # Consider Text::CS
+V here ;)
    
           # Saves a bit of space.
           my @suggestions = map{suggestion=> $_}@SuggestionList;

       # Create an anonymous hash
       push @spellingcheck, {text=> $key, thesuggestions=>\@suggestion
+s};
    }

    return ([ @spellingcheck,{text=> $last_field} ], $errors) ;    
}
[download]

However I can't see anything here that refers to Latin or UTF-8.. where is that glitch occuring?

--

Brother Frankus.

¤

use strict ;


use CGI qw/:standard/;
use CGI::Carp qw(fatalsToBrowser);
  
use HTML::Template;
use Data::Dumper;

use Unicode::String;


my $q = new CGI;

print $q->header();



  #my $template = "wsillafu.tmpl";
  my $template = "bbc/lemma/wsillafu.tmpl";
  my $tmpl = new HTML::Template( filename => $template, associate => $
+q );
  
  my $errorchecker = "0";my $memoryoverloadchecker = "0"; my $memorych
+ecker = "0";
  my $dictionarychecker = "0";my $otherchecker = "0";my $memoryfinder 
+= "0"; 
   
   if ($q->param()) {
               
           (my $spellingcheck, my $error) = lookup($q->param("brawddeg
+"));    
                    
           #print Dumper( $spellingcheck );
           #print "error = $error<br>";
           $tmpl->param( spellingcheck => $spellingcheck );
          if($error == 1){$errorchecker = "1"; $memoryfinder = "1";}
          if($error == -1){$memoryoverloadchecker = "1"; $memoryfinder
+ = "1";}
          if($error == -2){$memorychecker = "1"; $memoryfinder = "1";}
       if($error == -3){$dictionarychecker = "1"; $memoryfinder = "1";
+}
       if($error == -4){$otherchecker = "1"; $memoryfinder = "1";}
       
   }

     print $tmpl->output;
     
     if( $dictionarychecker eq "1" )
     {
         print "<br><br>Nid yw\'r geiriadur ar-lein ar gael.  E-bostiw
+ch meistr y wefan am gymorth \n";
     }
     
     if(  $otherchecker eq "1" )
     {
         print "<br><br>Mae gwall yn y gwirydd.  E-bostwich meistr y w
+efan am gymorth \n";
     }
     
     if( $memoryoverloadchecker eq "1" )
     {
         print "<br><br>Roedd mwy na 10 gwall yn y testun.  Dim ond y 
+10 gwall cyntaf sydd wedi ei cywiro.<br>
         Cywirwch rhain cyn cario ymlaen i gywiro.\n";
     }
     
     
     if( $memorychecker eq "1" )
     {
         print "<br><br>Mae'r gwirydd wedi rhedeg allan o gof.  E-bost
+iwch meistr y wefan am gymorth.\n";
     }
     
     if( $errorchecker eq "1" )
     {
         print "<br><br><font color=\"red\"><b>Mae rhai gwallau yng ng
+horff y testun.</b></font><br>
         Dewiswch air â gynhigir yn y blwch tynnu i lawr i bob gair sy
+dd wedi ei sillafu'n anghywir.
         <br>Yna gwagswch y botwm gwirio isod i gael eich testun wedi 
+ei wirio.<br>\n";
         print "<form name=\"ArgraffyddTestun\" method=\"POST\" Action
+=\"wsillafudangos.pl\">\n";
         print "<input type=\"Submit\" name=\"Gwirio\" value=\"Gwirio\
+"></form>\n";
     }
     
     
         
     if( $memoryfinder eq "0" ) {print "<br><br><b>Nid oedd gwall yn y
+ frawddeg</b>\n";}
     
  



use lemma;

sub lookup()
{

    $_ = shift(@_);
    
    #print "Paramater: $_ \n";
    
    
    #my $br = "";
    #$br->String::latin1( $_ );
    
    #print "Latin1 $br \n";
     
    
    (my $errors, my $str) = lemma::GetCysillSpellingErrors($_, 10);
    
    
    #print "$str\n";

    my @fields = split ("(<awgrymiadau>[^<]*</awgrymiadau>)", $str);
    
    #print Dumper(@fields);
    
    
    my $error = ( @fields  > 1);
    
    my @spellingcheck = ();
    
    for(my $i=0; $i<@fields; $i+=2)
    {
    
       my %pair;
       if($i+1 == @fields)
       {
           %pair = (text=> @fields[$i]);
       }
       else
       {
           
           @fields[$i+1] =~ s/<[^>]*>//g;
           my @SuggestionList = split /,/, @fields[$i+1];
     
         #print "SuggestionList = ",Dumper(@SuggestionList);
    
        my @suggestions = ();
        my $suggestion;
        foreach $suggestion (@SuggestionList)
        {
          my %hash = (suggestion=> $suggestion);
          push @suggestions, \%hash;
        }
    
    
           %pair = (text=> @fields[$i], thesuggestions=>\@suggestions)
+;
       }
       push @spellingcheck, \%pair;
    
    }
    
    #print Dumper(@spellingcheck);
    
    
    return (\@spellingcheck, $errors) ;
    
}
[download]

If I'm correct in guessing that you're asking where to put the header information to set the character set, change your line:

print $q->header();
[download]

to the following (untested):

print $q->header(-type    => 'text/html',
                 -charset => 'UTF-8'
                );
[download]

-rattus

__________
He seemed like such a nice guy to his neighbors / Kept to himself and never bothered them with favors
- Jefferson Airplane, "Assassin"