Transliterate to ascii while importing text from the catalogue.

git-svn-id: svn://tug.org/texlive/trunk@15176 c570f23f-e606-0410-a88d-b1316a301751

1 files changed, 7 insertions, 4 deletions

diff --git a/Master/tlpkg/TeXLive/TeXCatalogue.pm b/Master/tlpkg/TeXLive/TeXCatalogue.pm
index a634e4d69bc..ad9d83c0ac6 100644
--- a/Master/tlpkg/TeXLive/TeXCatalogue.pm
+++ b/Master/tlpkg/TeXLive/TeXCatalogue.pm
@@ -10,7 +10,7 @@
 use XML::Parser;
 use XML::XPath;
 use XML::XPath::XMLParser;
-use Encode;
+use Text::Unidecode;
 
 package TeXLive::TeXCatalogue::Entry;
 
@@ -78,7 +78,9 @@ sub initialize {
   my $docset = $parser->find('/entry/documentation');
   foreach my $node ($docset->get_nodelist) {
     my $docfile = $parser->find('./@href',$node);
-    my $details = $parser->find('./@details',$node);
+    # see comments at end of beautify()
+    my $details = Text::Unidecode::unidecode(
+        $parser->find('./@details',$node));
     my $language = $parser->find('./@language',$node);
     $self->{'docs'}{$docfile}{'available'} = 1;
     if ($details) { $self->{'docs'}{$docfile}{'details'} = $details; }
@@ -91,8 +93,9 @@ sub beautify{
   $txt =~ s/\n/ /g;
   $txt =~ s/^[[:space:]]*//g;
   $txt =~ s/[[:space:]][[:space:]]*/ /g;
-  # turn of the utf8 flag by encoding it to utf8
-  return(Encode::encode("utf-8", $txt));
+  # transliterate to ascii: it allows the final tlpdb to be pure ascii,
+  # avoiding problems since we don't control the user's terminal encoding
+  return(Text::Unidecode::unidecode($txt));
 }
 
 sub name {