Merging upstream version 0.73.

Signed-off-by: Daniel Baumann <daniel.baumann@progress-linux.org>

5 files changed, 106 insertions, 42 deletions

diff --git a/lib/Locale/Po4a/Common.pm b/lib/Locale/Po4a/Common.pm
index 1b09adc..9a9f069 100644
--- a/lib/Locale/Po4a/Common.pm
+++ b/lib/Locale/Po4a/Common.pm
@@ -224,7 +224,9 @@ BEGIN {
     if ( eval { require Locale::gettext } ) {
         import Locale::gettext;
         require POSIX;
-        POSIX::setlocale( &POSIX::LC_MESSAGES, '' );
+
+        # This cannot be done on Windows
+        POSIX::setlocale( &POSIX::LC_MESSAGES, '' ) unless $^O eq 'MSWin32';
     } else {
         eval '
            sub bindtextdomain($$) { }
diff --git a/lib/Locale/Po4a/Po.pm b/lib/Locale/Po4a/Po.pm
index d87a145..e5d0253 100644
--- a/lib/Locale/Po4a/Po.pm
+++ b/lib/Locale/Po4a/Po.pm
@@ -541,6 +541,9 @@ sub write {
           or croak wrap_mod( "po4a::po", dgettext( "po4a", "Cannot write to %s: %s" ), $filename, $! );
     }
 
+    # Some old perl versions qwak when the encoding is only set to utf. We need to first reset it to raw before setting utf8 again. Not sure why it's so.
+    binmode( $fh, ':raw' );
+    binmode( $fh, ':utf8' );
     print $fh "" . format_comment( $self->{header_comment}, "" )
       if length( $self->{header_comment} );
 
@@ -1537,7 +1540,6 @@ sub canonize {
     return $text;
 }
 
-
 # Wraps the string. We don't use Text::Wrap since it mangles whitespace at the
 # end of the split line.
 #
@@ -1549,7 +1551,7 @@ sub canonize {
 #  - The extra length allowed for the first line. Default: -10 (which means it
 #    will be wrapped 10 characters shorter).
 sub wrap {
-    my $text        = shift;
+    my $text = shift;
     return "0" if ( $text eq '0' );
     my $col         = shift || 76;
     my $first_shift = shift || -10;
diff --git a/lib/Locale/Po4a/Pod.pm b/lib/Locale/Po4a/Pod.pm
index 4b3dc04..1933ae3 100644
--- a/lib/Locale/Po4a/Pod.pm
+++ b/lib/Locale/Po4a/Pod.pm
@@ -27,6 +27,8 @@ use vars qw(@ISA);
 
 use Carp qw(croak confess);
 
+use Locale::Po4a::Common;
+
 sub initialize { }
 
 sub translate {
@@ -84,16 +86,24 @@ sub command {
         $charset =~ s/^\s*(.*?)\s*$/$1/s;
 
         my $master_charset = $self->get_in_charset;
-        croak wrap_mod(
-            "po4a::pod",
-            dgettext(
-                "po4a",
-                "The file %s declares %s as encoding, but you provided %s as master charset. Please change either setting."
-            ),
-            $self->{DOCPOD}{refname},
-            $charset,
-            $master_charset,
-        ) if ( length( $master_charset // '' ) > 0 && uc($charset) ne uc($master_charset) );
+
+        # in POD at least, there is no difference between utf8 and UTF-8. The major POD parsers handle "both encodings" in the exact same way.
+        # Despite https://perldoc.perl.org/Encode#UTF-8-vs.-utf8-vs.-UTF8
+        $master_charset = 'UTF-8' if $master_charset // '' =~ /utf-?8/i;
+        $charset        = 'UTF-8' if $charset              =~ /utf-?8/i;
+
+        if ( length( $master_charset // '' ) > 0 && uc($charset) ne uc($master_charset) ) {
+            croak wrap_mod(
+                "po4a::pod",
+                dgettext(
+                    "po4a",
+                    "The file %s declares %s as encoding, but you provided %s as master charset. Please change either setting."
+                ),
+                $self->{DOCPOD}{refname},
+                $charset,
+                $master_charset,
+            );
+        }
 
         # The =encoding line will be added by docheader
     } else {
diff --git a/lib/Locale/Po4a/TransTractor.pm b/lib/Locale/Po4a/TransTractor.pm
index 0bc5b8f..f943ba1 100644
--- a/lib/Locale/Po4a/TransTractor.pm
+++ b/lib/Locale/Po4a/TransTractor.pm
@@ -11,7 +11,7 @@ use warnings;
 
 use subs qw(makespace);
 use vars qw($VERSION @ISA @EXPORT);
-$VERSION = "0.72";
+$VERSION = "0.73";
 @ISA     = qw(DynaLoader);
 @EXPORT  = qw(new process translate
   read write readpo writepo
@@ -403,7 +403,7 @@ sub new {
         $self->{TT}{debug} = $options{'debug'};
     }
     if ( defined $options{'wrapcol'} ) {
-        if ( $options{'wrapcol'} < 0) {
+        if ( $options{'wrapcol'} < 0 ) {
             $self->{TT}{wrapcol} = 'Inf';
         } else {
             $self->{TT}{wrapcol} = $options{'wrapcol'};
@@ -421,11 +421,14 @@ sub new {
 
 =over 4
 
-=item read($$)
+=item read($$$)
 
-Add another input document data at the end of the existing array
-C<< @{$self->{TT}{doc_in}} >>. The argument is the filename to read. If a second
-argument is provided, it is the filename to use in the references.
+Add another input document data at the end of the existing array C<< @{$self->{TT}{doc_in}} >>.
+
+This function takes two mandatory arguments and an optional one.
+ * The filename to read on disk;
+ * The name to use as filename when building the reference in the PO file;
+ * The charset to use to read that file (UTF-8 by default)
 
 This array C<< @{$self->{TT}{doc_in}} >> holds this input document data as an
 array of strings with alternating meanings.
@@ -440,8 +443,8 @@ function when you're done with packing input files into the document.
 
 sub read() {
     my $self     = shift;
-    my $filename = shift or confess "Cannot write to a file without filename";
-    my $refname  = shift or confess "Cannot write to a file without refname";
+    my $filename = shift or confess "Cannot read from a file without filename";
+    my $refname  = shift or confess "Cannot read from a file without refname";
     my $charset  = shift || 'UTF-8';
     my $linenum  = 0;
 
@@ -474,8 +477,14 @@ sub read() {
     my $error = $@;
     if ( length($error) ) {
         chomp $error;
-        die wrap_msg( dgettext( "po4a", "Malformed encoding while reading from file %s with charset %s: %s" ),
-            $filename, $charset, $error );
+        die wrap_msg(
+            dgettext(
+                "po4a",
+                "Malformed encoding while reading from file %s with charset %s: %s\nIf %s is not the expected charset, you need to configure the right one with with --master-charset or other similar flags."
+            ),
+            $filename,
+            $charset, $error, $charset
+        );
     }
 
     # Croak if we need to
@@ -545,11 +554,23 @@ sub write {
             binmode STDERR, ':encoding(UTF-8)';
             my $char = chr( hex($1) );
             die wrap_msg(
-                dgettext( "po4a", "Malformed encoding while writing char '%s' to file %s with charset %s: %s" ),
-                $char, $filename, $charset, $error );
+                dgettext(
+                    "po4a",
+                    "Malformed encoding while writing char '%s' to file %s with charset %s: %s\nIf %s is not the expected charset, you need to configure the right one with with --localized-charset or other similar flags."
+                ),
+                $char,
+                $filename,
+                $charset, $error, $charset
+            );
         } else {
-            die wrap_msg( dgettext( "po4a", "Malformed encoding while writing to file %s with charset %s: %s" ),
-                $filename, $charset, $error );
+            die wrap_msg(
+                dgettext(
+                    "po4a",
+                    "Malformed encoding while writing to file %s with charset %s: %s\nIf %s is not the expected charset, you need to configure the right one with with --localized-charset or other similar flags."
+                ),
+                $filename,
+                $charset, $error, $charset
+            );
         }
     };
 
@@ -648,8 +669,14 @@ sub addendum_parse {
     } or do {
         my $error = $@ || 'Unknown failure';
         chomp $error;
-        die wrap_msg( dgettext( "po4a", "Malformed encoding while reading from file %s with charset %s: %s" ),
-            $filename, $charset, $error );
+        die wrap_msg(
+            dgettext(
+                "po4a",
+                "Malformed encoding while reading from file %s with charset %s: %s\nIf %s is not the expected charset, you need to configure the right one with with --master-charset or other similar flags."
+            ),
+            $filename,
+            $charset, $error, $charset
+        );
     };
 
     unless ( $header =~ s/PO4A-HEADER://i ) {
@@ -720,8 +747,14 @@ sub addendum_parse {
     my $error = $@;
     if ( length($error) ) {
         chomp $error;
-        die wrap_msg( dgettext( "po4a", "Malformed encoding while reading from file %s with charset %s: %s" ),
-            $filename, $charset, $error );
+        die wrap_msg(
+            dgettext(
+                "po4a",
+                "Malformed encoding while reading from file %s with charset %s: %s\nIf %s is not the expected charset, you need to configure the right one with with --master-charset or other similar flags."
+            ),
+            $filename,
+            $charset, $error, $charset
+        );
     }
     close INS;
 
@@ -995,7 +1028,7 @@ sub translate {
     # }
 
     if ( !defined $options{'wrapcol'} ) {
-        $options{'wrapcol'} = $self->{TT}{wrapcol}
+        $options{'wrapcol'} = $self->{TT}{wrapcol};
     } elsif ( $options{'wrapcol'} < 0 ) {
         $options{'wrapcol'} = $self->{TT}{wrapcol} + $options{'wrapcol'};
     }
diff --git a/lib/Locale/Po4a/Xml.pm b/lib/Locale/Po4a/Xml.pm
index 3666ac7..1c67bc7 100644
--- a/lib/Locale/Po4a/Xml.pm
+++ b/lib/Locale/Po4a/Xml.pm
@@ -899,16 +899,33 @@ sub tag_trans_xmlhead {
     my $out_charset = $self->get_out_charset;
 
     if ( defined $in_charset ) {
-        croak wrap_mod(
-            "po4a::xml",
-            dgettext(
-                "po4a",
-                "The file %s declares %s as encoding, but you provided %s as master charset. Please change either setting."
-            ),
-            $self->{'current_file'},
-            $in_charset,
-            $input_charset
-        ) if ( length( $input_charset // '' ) > 0 && uc($input_charset) ne uc($in_charset) );
+        if ( length( $input_charset // '' ) > 0 && uc($in_charset) ne uc($input_charset) ) {
+            if (   ( $in_charset eq 'UTF-8' || lc($in_charset) eq 'utf8' )
+                && ( $input_charset eq 'UTF-8' || lc($input_charset) eq 'utf8' ) )
+            {
+                croak wrap_mod(
+                    "po4a::pod",
+                    dgettext(
+                        "po4a",
+                        "The file %s declares %s as encoding, but you provided %s as master charset. Please change either setting because they really are different encoding in Perl. See https://perldoc.perl.org/Encode#UTF-8-vs.-utf8-vs.-UTF8"
+                    ),
+                    $self->{DOCPOD}{refname},
+                    $in_charset,
+                    $input_charset,
+                );
+            } else {
+                croak wrap_mod(
+                    "po4a::pod",
+                    dgettext(
+                        "po4a",
+                        "The file %s declares %s as encoding, but you provided %s as master charset. Please change either setting."
+                    ),
+                    $self->{DOCPOD}{refname},
+                    $in_charset,
+                    $input_charset,
+                );
+            }
+        }
 
         $tag =~ s/$in_charset/$out_charset/;
     } else {