More encoding mapping magic.

[p5sagit/p5-mst-13.2.git] / lib / open.pm

diff --git a/lib/open.pm b/lib/open.pm
index c90181b..c6978bb 100644 (file)
--- a/lib/open.pm
+++ b/lib/open.pm
@@ -2,15 +2,48 @@ package open;
 use Carp;
 $open::hint_bits = 0x20000;
 
-# layers array and hash mainly manipulated by C code in perlio.c
-use vars qw(%layers @layers);
+our $VERSION = '1.01';
 
-# Populate hash in non-PerlIO case
-%layers = (crlf => 1, raw => 0) unless (@layers);
+my $locale_encoding;
 
-# warn join(',',keys %layers);
+sub in_locale { $^H & $locale::hint_bits }
 
-our $VERSION = '1.00';
+sub _get_locale_encoding {
+    unless (defined $locale_encoding) {
+       eval { use I18N::Langinfo qw(langinfo CODESET) };
+       unless ($@) {
+           $locale_encoding = langinfo(CODESET);
+       }
+       my $country_language;
+        if (not $locale_encoding && in_locale()) {
+           if ($ENV{LC_ALL} =~ /^([^.]+)\.([^.]+)$/) {
+               ($country_language, $locale_encoding) = ($1, $2);
+           } elsif ($ENV{LANG} =~ /^([^.]+)\.([^.]+)$/) {
+               ($country_language, $locale_encoding) = ($1, $2);
+           }
+       } else {
+           # Could do heuristics based on the country and language
+           # parts of LC_ALL and LANG (the parts before the dot (if any)),
+           # since we have Locale::Country and Locale::Language available.
+           # TODO: get a database of Language -> Encoding mappings
+           # (the Estonian database would be excellent!)
+           # --jhi
+       }
+       if (defined $locale_encoding &&
+           $locale_encoding eq 'euc' &&
+           defined $country_language) {
+           if ($country_language =~ /^ja_JP|japan(?:ese)$/i) {
+               $locale_encoding = 'eucjp';
+           } elsif ($country_language =~ /^ko_KR|korea(?:n)$/i) {
+               $locale_encoding = 'euckr';
+           } elsif ($country_language =~ /^zh_TW|taiwan(?:ese)$/i) {
+               $locale_encoding = 'euctw';
+           }
+           croak "Locale encoding 'euc' too ambiguous"
+               if $locale_encoding eq 'euc';
+       }
+    }
+}
 
 sub import {
     my ($class,@args) = @_;
@@ -25,7 +58,19 @@ sub import {
        my @val;
        foreach my $layer (split(/\s+/,$discp)) {
             $layer =~ s/^://;
-           unless(exists $layers{$layer}) {
+           if ($layer eq 'locale') {
+               use Encode;
+               _get_locale_encoding()
+                   unless defined $locale_encoding;
+               croak "Cannot figure out an encoding to use"
+                   unless defined $locale_encoding;
+               if ($locale_encoding =~ /^utf-?8$/i) {
+                   $layer = "utf8";
+               } else {
+                   $layer = "encoding($locale_encoding)";
+               }
+           }
+           unless(PerlIO::Layer::->find($layer)) {
                carp("Unknown discipline layer '$layer'");
            }
            push(@val,":$layer");
@@ -90,33 +135,14 @@ everywhere if PerlIO is enabled.
 
 =head1 IMPLEMENTATION DETAILS
 
-There are two package variables C<%layers> and C<@layers> which are
-mainly manipulated by C code in F<perlio.c>, but are visible to the
-nosy:
-
-  print "Have ",join(',',keys %open::layers),"\n";
-  print "Using ",join(',',@open::layers),"\n";
-
-The C<%open::layers> hash is a record of the available "layers" that
-may be pushed onto a C<PerlIO> stream. The values of the hash are Perl
-objects, of class C<PerlIO::Layer> which are created by the C code in
-F<perlio.c>.  As yet there is nothing useful you can do with the
-objects at the perl level.
-
-The C<@open::layers> array is the current set of layers and their
-arguments.  The array consists of layer => argument pairs and I<must>
-always have even number of entries and the even entries I<must> be
-C<PerlIO::Layer> objects or Perl will "die" when it attempts to open a
-filehandle. In most cases the odd entry will be C<undef>, but in the
-case of (say) ":encoding(iso-8859-1)" it will be 'iso-8859-1'. These
-argument entries are currently restricted to being strings.
-
-When a new C<PerlIO> stream is opened, the C code looks at the array
-to determine the default layers to be pushed. So with care it is
-possible to manipulate the default layer "stack":
-
-    splice(@PerlIO::layers,-2,2);
-    push(@PerlIO::layers,$PerlIO::layers{'stdio'} => undef);
+There is a class method in C<PerlIO::Layer> C<find> which is implemented as XS code.
+It is called by C<import> to validate the layers:
+
+   PerlIO::Layer::->find("perlio")
+
+The return value (if defined) is a Perl object, of class C<PerlIO::Layer> which is
+created by the C code in F<perlio.c>.  As yet there is nothing useful you can do with the
+object at the perl level.
 
 =head1 SEE ALSO