Clean up older Locale::Maketext internals to no longer do crazy string
authorTodd Rinaldo <toddr@cpanel.net>
Thu, 15 Jul 2010 18:24:07 +0000 (14:24 -0400)
committerJesse Vincent <jesse@bestpractical.com>
Thu, 15 Jul 2010 18:25:14 +0000 (14:25 -0400)
eval tricks.

dist/Locale-Maketext/lib/Locale/Maketext.pm
dist/Locale-Maketext/lib/Locale/Maketext/Guts.pm
dist/Locale-Maketext/lib/Locale/Maketext/GutsLoader.pm

index 7a10ffb..929a70e 100644 (file)
@@ -10,6 +10,21 @@ use I18N::LangTags 0.30 ();
 BEGIN { unless(defined &DEBUG) { *DEBUG = sub () {0} } }
 # define the constant 'DEBUG' at compile-time
 
+# turn on utf8 if we have it (this is what GutsLoader.pm used to do essentially )
+#    use if (exists $INC{'utf8.pm'} || eval 'use utf8'), 'utf8';
+BEGIN {
+
+    # if we have it || we can load it
+    if ( exists $INC{'utf8.pm'} || eval { local $SIG{'__DIE__'};require utf8; } ) {
+        utf8->import();
+        DEBUG and warn " utf8 on for _compile()\n";
+    }
+    else {
+        DEBUG and warn " utf8 not available for _compile() ($INC{'utf8.pm'})\n$@\n";
+    }
+}
+
+
 $VERSION = '1.15';
 @ISA = ();
 
@@ -404,10 +419,6 @@ sub _add_supers {
 #
 ###########################################################################
 
-use Locale::Maketext::GutsLoader;
-
-###########################################################################
-
 my %tried = ();
 # memoization of whether we've used this module, or found it unusable.
 
@@ -470,4 +481,312 @@ sub _lex_refs {  # report the lexicon references for this handle's class
 
 sub clear_isa_scan { %isa_scan = (); return; } # end on a note of simplicity!
 
+#--------------------------------------------------------------------------
+
+sub _compile {
+    # This big scary routine compiles an entry.
+    # It returns either a coderef if there's brackety bits in this, or
+    #  otherwise a ref to a scalar.
+
+    my $target = ref($_[0]) || $_[0];
+
+    my(@code);
+    my(@c) = (''); # "chunks" -- scratch.
+    my $call_count = 0;
+    my $big_pile = '';
+    {
+        my $in_group = 0; # start out outside a group
+        my($m, @params); # scratch
+
+        while($_[1] =~  # Iterate over chunks.
+            m/\G(
+                [^\~\[\]]+  # non-~[] stuff
+                |
+                ~.       # ~[, ~], ~~, ~other
+                |
+                \[          # [ presumably opening a group
+                |
+                \]          # ] presumably closing a group
+                |
+                ~           # terminal ~ ?
+                |
+                $
+            )/xgs
+        ) {
+            DEBUG>2 and warn qq{  "$1"\n};
+
+            if($1 eq '[' or $1 eq '') {       # "[" or end
+                # Whether this is "[" or end, force processing of any
+                #  preceding literal.
+                if($in_group) {
+                    if($1 eq '') {
+                        $target->_die_pointing($_[1], 'Unterminated bracket group');
+                    }
+                    else {
+                        $target->_die_pointing($_[1], 'You can\'t nest bracket groups');
+                    }
+                }
+                else {
+                    if ($1 eq '') {
+                        DEBUG>2 and warn "   [end-string]\n";
+                    }
+                    else {
+                        $in_group = 1;
+                    }
+                    die "How come \@c is empty?? in <$_[1]>" unless @c; # sanity
+                    if(length $c[-1]) {
+                        # Now actually processing the preceding literal
+                        $big_pile .= $c[-1];
+                        if($USE_LITERALS and (
+                                (ord('A') == 65)
+                                ? $c[-1] !~ m/[^\x20-\x7E]/s
+                                # ASCII very safe chars
+                                : $c[-1] !~ m/[^ !"\#\$%&'()*+,\-.\/0-9:;<=>?\@A-Z[\\\]^_`a-z{|}~\x07]/s
+                                # EBCDIC very safe chars
+                            )) {
+                            # normal case -- all very safe chars
+                            $c[-1] =~ s/'/\\'/g;
+                            push @code, q{ '} . $c[-1] . "',\n";
+                            $c[-1] = ''; # reuse this slot
+                        }
+                        else {
+                            push @code, ' $c[' . $#c . "],\n";
+                            push @c, ''; # new chunk
+                        }
+                    }
+                    # else just ignore the empty string.
+                }
+
+            }
+            elsif($1 eq ']') {  # "]"
+                # close group -- go back in-band
+                if($in_group) {
+                    $in_group = 0;
+
+                    DEBUG>2 and warn "   --Closing group [$c[-1]]\n";
+
+                    # And now process the group...
+
+                    if(!length($c[-1]) or $c[-1] =~ m/^\s+$/s) {
+                        DEBUG>2 and warn "   -- (Ignoring)\n";
+                        $c[-1] = ''; # reset out chink
+                        next;
+                    }
+
+                    #$c[-1] =~ s/^\s+//s;
+                    #$c[-1] =~ s/\s+$//s;
+                    ($m,@params) = split(/,/, $c[-1], -1);  # was /\s*,\s*/
+
+                    # A bit of a hack -- we've turned "~,"'s into DELs, so turn
+                    #  'em into real commas here.
+                    if (ord('A') == 65) { # ASCII, etc
+                        foreach($m, @params) { tr/\x7F/,/ }
+                    }
+                    else {              # EBCDIC (1047, 0037, POSIX-BC)
+                        # Thanks to Peter Prymmer for the EBCDIC handling
+                        foreach($m, @params) { tr/\x07/,/ }
+                    }
+
+                    # Special-case handling of some method names:
+                    if($m eq '_*' or $m =~ m/^_(-?\d+)$/s) {
+                        # Treat [_1,...] as [,_1,...], etc.
+                        unshift @params, $m;
+                        $m = '';
+                    }
+                    elsif($m eq '*') {
+                        $m = 'quant'; # "*" for "times": "4 cars" is 4 times "cars"
+                    }
+                    elsif($m eq '#') {
+                        $m = 'numf';  # "#" for "number": [#,_1] for "the number _1"
+                    }
+
+                    # Most common case: a simple, legal-looking method name
+                    if($m eq '') {
+                        # 0-length method name means to just interpolate:
+                        push @code, ' (';
+                    }
+                    elsif($m =~ /^\w+(?:\:\:\w+)*$/s
+                            and $m !~ m/(?:^|\:)\d/s
+                        # exclude starting a (sub)package or symbol with a digit
+                    ) {
+                        # Yes, it even supports the demented (and undocumented?)
+                        #  $obj->Foo::bar(...) syntax.
+                        $target->_die_pointing(
+                            $_[1], q{Can't use "SUPER::" in a bracket-group method},
+                            2 + length($c[-1])
+                        )
+                        if $m =~ m/^SUPER::/s;
+                        # Because for SUPER:: to work, we'd have to compile this into
+                        #  the right package, and that seems just not worth the bother,
+                        #  unless someone convinces me otherwise.
+
+                        push @code, ' $_[0]->' . $m . '(';
+                    }
+                    else {
+                        # TODO: implement something?  or just too icky to consider?
+                        $target->_die_pointing(
+                            $_[1],
+                            "Can't use \"$m\" as a method name in bracket group",
+                            2 + length($c[-1])
+                        );
+                    }
+
+                    pop @c; # we don't need that chunk anymore
+                    ++$call_count;
+
+                    foreach my $p (@params) {
+                        if($p eq '_*') {
+                            # Meaning: all parameters except $_[0]
+                            $code[-1] .= ' @_[1 .. $#_], ';
+                            # and yes, that does the right thing for all @_ < 3
+                        }
+                        elsif($p =~ m/^_(-?\d+)$/s) {
+                            # _3 meaning $_[3]
+                            $code[-1] .= '$_[' . (0 + $1) . '], ';
+                        }
+                        elsif($USE_LITERALS and (
+                                (ord('A') == 65)
+                                ? $p !~ m/[^\x20-\x7E]/s
+                                # ASCII very safe chars
+                                : $p !~ m/[^ !"\#\$%&'()*+,\-.\/0-9:;<=>?\@A-Z[\\\]^_`a-z{|}~\x07]/s
+                                # EBCDIC very safe chars
+                            )) {
+                            # Normal case: a literal containing only safe characters
+                            $p =~ s/'/\\'/g;
+                            $code[-1] .= q{'} . $p . q{', };
+                        }
+                        else {
+                            # Stow it on the chunk-stack, and just refer to that.
+                            push @c, $p;
+                            push @code, ' $c[' . $#c . '], ';
+                        }
+                    }
+                    $code[-1] .= "),\n";
+
+                    push @c, '';
+                }
+                else {
+                    $target->_die_pointing($_[1], q{Unbalanced ']'});
+                }
+
+            }
+            elsif(substr($1,0,1) ne '~') {
+                # it's stuff not containing "~" or "[" or "]"
+                # i.e., a literal blob
+                $c[-1] .= $1;
+
+            }
+            elsif($1 eq '~~') { # "~~"
+                $c[-1] .= '~';
+
+            }
+            elsif($1 eq '~[') { # "~["
+                $c[-1] .= '[';
+
+            }
+            elsif($1 eq '~]') { # "~]"
+                $c[-1] .= ']';
+
+            }
+            elsif($1 eq '~,') { # "~,"
+                if($in_group) {
+                    # This is a hack, based on the assumption that no-one will actually
+                    # want a DEL inside a bracket group.  Let's hope that's it's true.
+                    if (ord('A') == 65) { # ASCII etc
+                        $c[-1] .= "\x7F";
+                    }
+                    else {              # EBCDIC (cp 1047, 0037, POSIX-BC)
+                        $c[-1] .= "\x07";
+                    }
+                }
+                else {
+                    $c[-1] .= '~,';
+                }
+
+            }
+            elsif($1 eq '~') { # possible only at string-end, it seems.
+                $c[-1] .= '~';
+
+            }
+            else {
+                # It's a "~X" where X is not a special character.
+                # Consider it a literal ~ and X.
+                $c[-1] .= $1;
+            }
+        }
+    }
+
+    if($call_count) {
+        undef $big_pile; # Well, nevermind that.
+    }
+    else {
+        # It's all literals!  Ahwell, that can happen.
+        # So don't bother with the eval.  Return a SCALAR reference.
+        return \$big_pile;
+    }
+
+    die q{Last chunk isn't null??} if @c and length $c[-1]; # sanity
+    DEBUG and warn scalar(@c), " chunks under closure\n";
+    if(@code == 0) { # not possible?
+        DEBUG and warn "Empty code\n";
+        return \'';
+    }
+    elsif(@code > 1) { # most cases, presumably!
+        unshift @code, "join '',\n";
+    }
+    unshift @code, "use strict; sub {\n";
+    push @code, "}\n";
+
+    DEBUG and warn @code;
+    my $sub = eval(join '', @code);
+    die "$@ while evalling" . join('', @code) if $@; # Should be impossible.
+    return $sub;
+}
+
+#--------------------------------------------------------------------------
+
+sub _die_pointing {
+    # This is used by _compile to throw a fatal error
+    my $target = shift; # class name
+    # ...leaving $_[0] the error-causing text, and $_[1] the error message
+
+    my $i = index($_[0], "\n");
+
+    my $pointy;
+    my $pos = pos($_[0]) - (defined($_[2]) ? $_[2] : 0) - 1;
+    if($pos < 1) {
+        $pointy = "^=== near there\n";
+    }
+    else { # we need to space over
+        my $first_tab = index($_[0], "\t");
+        if($pos > 2 and ( -1 == $first_tab  or  $first_tab > pos($_[0]))) {
+            # No tabs, or the first tab is harmlessly after where we will point to,
+            # AND we're far enough from the margin that we can draw a proper arrow.
+            $pointy = ('=' x $pos) . "^ near there\n";
+        }
+        else {
+            # tabs screw everything up!
+            $pointy = substr($_[0],0,$pos);
+            $pointy =~ tr/\t //cd;
+            # make everything into whitespace, but preseving tabs
+            $pointy .= "^=== near there\n";
+        }
+    }
+
+    my $errmsg = "$_[1], in\:\n$_[0]";
+
+    if($i == -1) {
+        # No newline.
+        $errmsg .= "\n" . $pointy;
+    }
+    elsif($i == (length($_[0]) - 1)  ) {
+        # Already has a newline at end.
+        $errmsg .= $pointy;
+    }
+    else {
+        # don't bother with the pointy bit, I guess.
+    }
+    Carp::croak( "$errmsg via $target, as used" );
+}
+
 1;
index 9af292c..f4d1cd1 100644 (file)
 package Locale::Maketext::Guts;
 
-$VERSION = '1.13';
+use Locale::Maketext;
 
-BEGIN {
-    # Just so we're nice and define SOMETHING in "our" package.
-    *zorp = sub { return scalar @_ } unless defined &zorp;
-}
+=head1 NAME
 
-package Locale::Maketext;
-use strict;
-use vars qw($USE_LITERALS $GUTSPATH);
+Locale::Maketext::Guts - Deprecated module to load Locale::Maketext utf8 code
 
-BEGIN {
-    $GUTSPATH = __FILE__;
-    *DEBUG = sub () {0} unless defined &DEBUG;
-}
+=head1 SYNOPSIS
 
-use utf8;
+  # Do this instead please
+  use Locale::Maketext
 
-sub _compile {
-    # This big scary routine compiles an entry.
-    # It returns either a coderef if there's brackety bits in this, or
-    #  otherwise a ref to a scalar.
+=head1 DESCRIPTION
 
-    my $target = ref($_[0]) || $_[0];
+Previously majic was done to load Locale::Maketext when utf8 was unavailable. The subs this module provided were merged back into Locale::Maketext
 
-    my(@code);
-    my(@c) = (''); # "chunks" -- scratch.
-    my $call_count = 0;
-    my $big_pile = '';
-    {
-        my $in_group = 0; # start out outside a group
-        my($m, @params); # scratch
-
-        while($_[1] =~  # Iterate over chunks.
-            m/\G(
-                [^\~\[\]]+  # non-~[] stuff
-                |
-                ~.       # ~[, ~], ~~, ~other
-                |
-                \[          # [ presumably opening a group
-                |
-                \]          # ] presumably closing a group
-                |
-                ~           # terminal ~ ?
-                |
-                $
-            )/xgs
-        ) {
-            DEBUG>2 and print qq{  "$1"\n};
-
-            if($1 eq '[' or $1 eq '') {       # "[" or end
-                # Whether this is "[" or end, force processing of any
-                #  preceding literal.
-                if($in_group) {
-                    if($1 eq '') {
-                        $target->_die_pointing($_[1], 'Unterminated bracket group');
-                    }
-                    else {
-                        $target->_die_pointing($_[1], 'You can\'t nest bracket groups');
-                    }
-                }
-                else {
-                    if ($1 eq '') {
-                        DEBUG>2 and print "   [end-string]\n";
-                    }
-                    else {
-                        $in_group = 1;
-                    }
-                    die "How come \@c is empty?? in <$_[1]>" unless @c; # sanity
-                    if(length $c[-1]) {
-                        # Now actually processing the preceding literal
-                        $big_pile .= $c[-1];
-                        if($USE_LITERALS and (
-                                (ord('A') == 65)
-                                ? $c[-1] !~ m/[^\x20-\x7E]/s
-                                # ASCII very safe chars
-                                : $c[-1] !~ m/[^ !"\#\$%&'()*+,\-.\/0-9:;<=>?\@A-Z[\\\]^_`a-z{|}~\x07]/s
-                                # EBCDIC very safe chars
-                            )) {
-                            # normal case -- all very safe chars
-                            $c[-1] =~ s/'/\\'/g;
-                            push @code, q{ '} . $c[-1] . "',\n";
-                            $c[-1] = ''; # reuse this slot
-                        }
-                        else {
-                            push @code, ' $c[' . $#c . "],\n";
-                            push @c, ''; # new chunk
-                        }
-                    }
-                    # else just ignore the empty string.
-                }
-
-            }
-            elsif($1 eq ']') {  # "]"
-                # close group -- go back in-band
-                if($in_group) {
-                    $in_group = 0;
-
-                    DEBUG>2 and print "   --Closing group [$c[-1]]\n";
-
-                    # And now process the group...
-
-                    if(!length($c[-1]) or $c[-1] =~ m/^\s+$/s) {
-                        DEBUG > 2 and print "   -- (Ignoring)\n";
-                        $c[-1] = ''; # reset out chink
-                        next;
-                    }
-
-                    #$c[-1] =~ s/^\s+//s;
-                    #$c[-1] =~ s/\s+$//s;
-                    ($m,@params) = split(/,/, $c[-1], -1);  # was /\s*,\s*/
-
-                    # A bit of a hack -- we've turned "~,"'s into DELs, so turn
-                    #  'em into real commas here.
-                    if (ord('A') == 65) { # ASCII, etc
-                        foreach($m, @params) { tr/\x7F/,/ }
-                    }
-                    else {              # EBCDIC (1047, 0037, POSIX-BC)
-                        # Thanks to Peter Prymmer for the EBCDIC handling
-                        foreach($m, @params) { tr/\x07/,/ }
-                    }
-
-                    # Special-case handling of some method names:
-                    if($m eq '_*' or $m =~ m/^_(-?\d+)$/s) {
-                        # Treat [_1,...] as [,_1,...], etc.
-                        unshift @params, $m;
-                        $m = '';
-                    }
-                    elsif($m eq '*') {
-                        $m = 'quant'; # "*" for "times": "4 cars" is 4 times "cars"
-                    }
-                    elsif($m eq '#') {
-                        $m = 'numf';  # "#" for "number": [#,_1] for "the number _1"
-                    }
-
-                    # Most common case: a simple, legal-looking method name
-                    if($m eq '') {
-                        # 0-length method name means to just interpolate:
-                        push @code, ' (';
-                    }
-                    elsif($m =~ /^\w+(?:\:\:\w+)*$/s
-                            and $m !~ m/(?:^|\:)\d/s
-                        # exclude starting a (sub)package or symbol with a digit
-                    ) {
-                        # Yes, it even supports the demented (and undocumented?)
-                        #  $obj->Foo::bar(...) syntax.
-                        $target->_die_pointing(
-                            $_[1], q{Can't use "SUPER::" in a bracket-group method},
-                            2 + length($c[-1])
-                        )
-                        if $m =~ m/^SUPER::/s;
-                        # Because for SUPER:: to work, we'd have to compile this into
-                        #  the right package, and that seems just not worth the bother,
-                        #  unless someone convinces me otherwise.
-
-                        push @code, ' $_[0]->' . $m . '(';
-                    }
-                    else {
-                        # TODO: implement something?  or just too icky to consider?
-                        $target->_die_pointing(
-                            $_[1],
-                            "Can't use \"$m\" as a method name in bracket group",
-                            2 + length($c[-1])
-                        );
-                    }
-
-                    pop @c; # we don't need that chunk anymore
-                    ++$call_count;
-
-                    foreach my $p (@params) {
-                        if($p eq '_*') {
-                            # Meaning: all parameters except $_[0]
-                            $code[-1] .= ' @_[1 .. $#_], ';
-                            # and yes, that does the right thing for all @_ < 3
-                        }
-                        elsif($p =~ m/^_(-?\d+)$/s) {
-                            # _3 meaning $_[3]
-                            $code[-1] .= '$_[' . (0 + $1) . '], ';
-                        }
-                        elsif($USE_LITERALS and (
-                                (ord('A') == 65)
-                                ? $p !~ m/[^\x20-\x7E]/s
-                                # ASCII very safe chars
-                                : $p !~ m/[^ !"\#\$%&'()*+,\-.\/0-9:;<=>?\@A-Z[\\\]^_`a-z{|}~\x07]/s
-                                # EBCDIC very safe chars
-                            )) {
-                            # Normal case: a literal containing only safe characters
-                            $p =~ s/'/\\'/g;
-                            $code[-1] .= q{'} . $p . q{', };
-                        }
-                        else {
-                            # Stow it on the chunk-stack, and just refer to that.
-                            push @c, $p;
-                            push @code, ' $c[' . $#c . '], ';
-                        }
-                    }
-                    $code[-1] .= "),\n";
-
-                    push @c, '';
-                }
-                else {
-                    $target->_die_pointing($_[1], q{Unbalanced ']'});
-                }
-
-            }
-            elsif(substr($1,0,1) ne '~') {
-                # it's stuff not containing "~" or "[" or "]"
-                # i.e., a literal blob
-                $c[-1] .= $1;
-
-            }
-            elsif($1 eq '~~') { # "~~"
-                $c[-1] .= '~';
-
-            }
-            elsif($1 eq '~[') { # "~["
-                $c[-1] .= '[';
-
-            }
-            elsif($1 eq '~]') { # "~]"
-                $c[-1] .= ']';
-
-            }
-            elsif($1 eq '~,') { # "~,"
-                if($in_group) {
-                    # This is a hack, based on the assumption that no-one will actually
-                    # want a DEL inside a bracket group.  Let's hope that's it's true.
-                    if (ord('A') == 65) { # ASCII etc
-                        $c[-1] .= "\x7F";
-                    }
-                    else {              # EBCDIC (cp 1047, 0037, POSIX-BC)
-                        $c[-1] .= "\x07";
-                    }
-                }
-                else {
-                    $c[-1] .= '~,';
-                }
-
-            }
-            elsif($1 eq '~') { # possible only at string-end, it seems.
-                $c[-1] .= '~';
-
-            }
-            else {
-                # It's a "~X" where X is not a special character.
-                # Consider it a literal ~ and X.
-                $c[-1] .= $1;
-            }
-        }
-    }
-
-    if($call_count) {
-        undef $big_pile; # Well, nevermind that.
-    }
-    else {
-        # It's all literals!  Ahwell, that can happen.
-        # So don't bother with the eval.  Return a SCALAR reference.
-        return \$big_pile;
-    }
-
-    die q{Last chunk isn't null??} if @c and length $c[-1]; # sanity
-    DEBUG and warn scalar(@c), " chunks under closure\n";
-    if(@code == 0) { # not possible?
-        DEBUG and warn "Empty code\n";
-        return \'';
-    }
-    elsif(@code > 1) { # most cases, presumably!
-        unshift @code, "join '',\n";
-    }
-    unshift @code, "use strict; sub {\n";
-    push @code, "}\n";
-
-    DEBUG and warn @code;
-    my $sub = eval(join '', @code);
-    die "$@ while evalling" . join('', @code) if $@; # Should be impossible.
-    return $sub;
-}
-
-# - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
-
-sub _die_pointing {
-    # This is used by _compile to throw a fatal error
-    my $target = shift; # class name
-    # ...leaving $_[0] the error-causing text, and $_[1] the error message
-
-    my $i = index($_[0], "\n");
-
-    my $pointy;
-    my $pos = pos($_[0]) - (defined($_[2]) ? $_[2] : 0) - 1;
-    if($pos < 1) {
-        $pointy = "^=== near there\n";
-    }
-    else { # we need to space over
-        my $first_tab = index($_[0], "\t");
-        if($pos > 2 and ( -1 == $first_tab  or  $first_tab > pos($_[0]))) {
-            # No tabs, or the first tab is harmlessly after where we will point to,
-            # AND we're far enough from the margin that we can draw a proper arrow.
-            $pointy = ('=' x $pos) . "^ near there\n";
-        }
-        else {
-            # tabs screw everything up!
-            $pointy = substr($_[0],0,$pos);
-            $pointy =~ tr/\t //cd;
-            # make everything into whitespace, but preseving tabs
-            $pointy .= "^=== near there\n";
-        }
-    }
-
-    my $errmsg = "$_[1], in\:\n$_[0]";
-
-    if($i == -1) {
-        # No newline.
-        $errmsg .= "\n" . $pointy;
-    }
-    elsif($i == (length($_[0]) - 1)  ) {
-        # Already has a newline at end.
-        $errmsg .= $pointy;
-    }
-    else {
-        # don't bother with the pointy bit, I guess.
-    }
-    Carp::croak( "$errmsg via $target, as used" );
-}
+=cut
 
 1;
-
index daa9840..421d538 100644 (file)
@@ -1,49 +1,22 @@
 package Locale::Maketext::GutsLoader;
 
-$VERSION = '1.13';
+use Locale::Maketext;
 
-use strict;
 sub zorp { return scalar @_ }
 
-BEGIN {
-    $Locale::Maketext::GutsLoader::GUTSPATH = __FILE__;
-    *Locale::Maketext::DEBUG = sub () {0}
-    unless defined &Locale::Maketext::DEBUG;
-}
-
-#
-# This whole drama is so that we can load the utf8'd code
-# in Locale::Maketext::Guts, but if that fails, snip the
-# utf8 and then try THAT.
-#
-
-$Locale::Maketext::GUTSPATH = '';
-Locale::Maketext::DEBUG and warn "Requiring Locale::Maketext::Guts...\n";
-eval 'require Locale::Maketext::Guts';
-
-if ($@) {
-    my $path = $Locale::Maketext::GUTSPATH;
-
-    die "Can't load Locale::Maketext::Guts\nAborting" unless $path;
-
-    die "No readable file $Locale::Maketext::GutsLoader::GUTSPATH\nAborting"
-    unless -e $path and -f _ and -r _;
-
-    open(IN, $path) or die "Can't read-open $path\nAborting";
-
-    my $source;
-    { local $/;  $source = <IN>; }
-    close(IN);
-    unless( $source =~ s/\b(use utf8)/# $1/ ) {
-        Locale::Maketext::DEBUG and
-        print "I didn't see 'use utf8' in $path\n";
-    }
-    eval $source;
-    die "Can't compile $path\n...The error I got was:\n$@\nAborting" if $@;
-    Locale::Maketext::DEBUG and warn "Non-utf8'd Locale::Maketext::Guts fine\n";
-}
-else {
-    Locale::Maketext::DEBUG and warn "Loaded Locale::Maketext::Guts fine\n";
-}
+=head1 NAME
+
+Locale::Maketext::GutsLoader - Deprecated module to load Locale::Maketext utf8 code
+
+=head1 SYNOPSIS
+
+  # Do this instead please
+  use Locale::Maketext
+
+=head1 DESCRIPTION
+
+Previously majic was done to load Locale::Maketext when utf8 was unavailable. The subs this module provided were merged back into Locale::Maketext
+
+=cut
 
 1;