elinks/po/perl/msgaccel-prepare

#! /usr/bin/perl
# The copyright notice and license are in the POD at the bottom.

use strict;
use warnings;
use Locale::PO qw();
use Getopt::Long qw(GetOptions :config bundling gnu_compat);
use autouse 'Pod::Usage' => qw(pod2usage);
use autouse 'File::Spec::Functions' => qw(catfile);

my $VERSION = "1.2";

sub show_version
{
    print "msgaccel-prepare $VERSION\n";
    pod2usage({-verbose => 99, -sections => "COPYRIGHT AND LICENSE",
	       -exitval => 0});
}

{
    package Contextline;
    use fields qw(lineno contexts);
    sub new {
        my($self, $lineno, $contexts) = @_;
        $self = fields::new($self) unless ref $self;
	$self->{lineno} = $lineno;
	$self->{contexts} = $contexts;
	return $self;
    }
}

my @Srcpath;
my $Accelerator_tag;

# Each key is a file name.
# Each value is a reference to an array of references to Contextline
# pseudo-hashes.  The array is in ascending order by {lineno}.
my %Srcfiles;

sub open_file_on_path ($@)
{
    my($fname, @path) = @_;
    if (@path) {
	my @warnings;
	foreach my $dir (@path) {
	    my $full_fname = catfile($dir, $fname);
	    if (open my $fh, "<", $full_fname) {
		return($fh, $full_fname);
	    } else {
		push @warnings, "$full_fname: $!\n";
	    }
	}
	# Didn't find $name anywhere on @path.
	warn $_ foreach @warnings;
	return;
    } else {			# no path
	if (open my $fh, "<", $fname) {
	    return($fh, $fname);
	} else {
	    warn "$fname: $!\n";
	    return;
	}
    }
    # not reached
}

# Scan the file $src_fname for gettext_accelerator_context directives,
# cache the result in %Srcfiles, and return it in that format.
# Cache and return [] if the file cannot be read on @Srcpath.
sub contextlines ($)
{
    my($src_fname) = @_;
    return $Srcfiles{$src_fname} if exists($Srcfiles{$src_fname});
    my @contextlines = ();

    if (my($src_fh, $src_full_fname) = open_file_on_path($src_fname, @Srcpath)) {
	my @prevctxs;
	local $_;
	while (<$src_fh>) {
	    chomp;
	    if (/^\}/ && @prevctxs) {
		push @contextlines, Contextline->new($., [@prevctxs = ()]);
	    }
	    if (my($contexts) = /\[gettext_accelerator_context\(([^()]*)\)\]/) {
		my @contexts = grep { $_ ne "" } split(/\s*,\s*/, $contexts);
		s/^\./${src_fname}:/ foreach @contexts;
		warn "$src_full_fname:$.: Previous context not closed\n"
		    if @prevctxs && @contexts;
		warn "$src_full_fname:$.: Context already closed\n"
		    if !@prevctxs && !@contexts;
		push @contextlines, Contextline->new($., [@prevctxs = @contexts]);
	    } elsif (/gettext_accelerator_context/) {
		warn "$src_full_fname:$.: Suspicious non-directive: $_\n";
	    }
	}
	warn "$src_full_fname:$.: Last context not closed\n" if @prevctxs;
    } # if opened ok

    return $Srcfiles{$src_fname} = \@contextlines;
}

sub contexts ($$)
{
    my($srcfile, $lineno) = @_;
    # Could use a binary search here.
    my $contextlines = contextlines($srcfile);
    my @contexts = ();
    foreach my Contextline $contextline (@{$contextlines}) {
	return @contexts if $contextline->{lineno} > $lineno;
	@contexts = @{$contextline->{contexts}};
    }
    return ();
}

sub gather_accelerator_contexts ($$)
{
    my($pos, $po_fname) = @_;
    foreach my $po (@$pos) {
	my $automatic = $po->automatic();
	$automatic =~ s/^\[gettext_accelerator_context\(.*(?:\n|\z)//mg
	    if defined($automatic);

	if ($po->msgid() =~ /\Q$Accelerator_tag/s) {
	    my @po_contexts = ();
	    foreach my $ref (split(' ', $po->reference())) {
		my @parts = split(/\:/, $ref);
		warn "weird reference: $ref\n", next unless @parts == 2;
		my @ref_contexts = contexts($parts[0], $parts[1]);
		if (@ref_contexts) {
		    push @po_contexts, grep { $_ ne "IGNORE" } @ref_contexts;
		} else {
		    warn "$ref: No accelerator context for msgid " . $po->msgid() . "\n";
		}
	    }
	    if (@po_contexts) {
		# sort and uniquify
		@po_contexts = sort keys %{{map { $_ => 1 } @po_contexts}};
		$automatic .= "\n" if defined($automatic) and $automatic ne "";
		$automatic .= "accelerator_context(" . join(", ", @po_contexts) . ")";
	    }
	}
	$po->automatic($automatic);
    }
}

GetOptions("srcdir|source-directory|S=s" => \@Srcpath,
	   "accelerator-tag=s" => sub {
	       my($option, $value) = @_;
	       die "Cannot use multiple --accelerator-tag options\n"
		   if defined($Accelerator_tag);
	       die "--accelerator-tag requires a single-character argument\n"
		   if length($value) != 1;
	       $Accelerator_tag = $value;
	   },
	   "help" => sub { pod2usage({-verbose => 1, -exitval => 0}) },
	   "version" => \&show_version)
    or exit 2;
$Accelerator_tag = "~" unless defined $Accelerator_tag;
print(STDERR "$0: missing file operand\n"), exit 2 unless @ARGV;
print(STDERR "$0: too many operands\n"), exit 2 if @ARGV > 1;

my($po_fname) = @ARGV;
my $pos = Locale::PO->load_file_asarray($po_fname) or die "$po_fname: $!";
gather_accelerator_contexts($pos, $po_fname);
Locale::PO->save_file_fromarray($po_fname, $pos) or die "$po_fname: $!";

__END__

=head1 NAME

msgaccel-prepare - Augment a PO file with information for detecting
accelerator conflicts.

=head1 SYNOPSIS

B<msgaccel-prepare> [I<option> ...] F<I<program>.pot>

=head1 DESCRIPTION

B<msgaccel-prepare> is part of a framework that detects conflicting
accelerator keys in Gettext PO files.  A conflict is when two items in
the same menu or two buttons in the same dialog box use the same
accelerator key.

The PO file format does not normally include any information
on which strings will be used in the same menu or dialog box.
B<msgaccel-prepare> adds this information in the form of
"accelerator_context" comments, which B<msgaccel-check>
then parses in order to detect the conflicts.

The PO file format also does not directly support definitions of
accelerator keys.  Typically, the keys are encoded in C<msgstr>
strings, by placing a tilde in front of the character that should be
used as the accelerator key.  That is also the syntax supported by
this framework and by B<msgfmt --check-accelerators> of GNU Gettext.

B<msgaccel-prepare> first reads the F<I<program>.pot> file named on
the command line.  This file must include "#:" comments that point
to the source files from which B<xgettext> extracted each C<msgid>.
B<msgaccel-prepare> then scans those source files for context
information and rewrites F<I<program>.pot> to include the
"accelerator_context" comments.  Finally, the standard tool
B<msgmerge> can be used to copy the added comments to all the
F<I<language>.po> files.

It is best to run B<msgaccel-prepare> immediately after B<xgettext>
so that the source references will be up to date.

=head2 Contexts

Whenever a source file refers to an C<msgid> that includes an
accelerator key, it must assign one or more named B<contexts> to it.
The C<msgstr>s in the PO files inherit these contexts.  If multiple
C<msgstr>s use the same accelerator (case insensitive) in the same
context, that's a conflict and can be detected automatically.

If the same C<msgid> is used in multiple places in the source code,
and those places assign different contexts to it, then all of those
contexts will apply.

The names of contexts consist of C identifiers delimited with periods.
The identifier is typically the name of a function that sets up a
dialog, or the name of an array where the items of a menu are listed.
There is a special feature for file-local identifiers (C<static> in C):
if the name begins with a period, then the period will be replaced
with the name of the source file and a colon.  The name "IGNORE" is
reserved.

If a menu is programmatically generated from multiple parts, of which
some are never used together, so that it is safe to use the same
accelerators in them, then it is necessary to define multiple contexts
for the same menu.

=head2 How to define contexts in source files

The contexts are defined with "gettext_accelerator_context" comments
in source files.  These comments delimit regions where all C<msgid>s
that seem to contain accelerators are given the same contexts.  There
must be one special comment at the top of the region; it lists the
contexts assigned to that region.  The region automatically ends at
the end of the function (found with regexp C</^\}/>), but it can also
be closed explicitly with another special comment.  The comments are
formatted like this:

  /* [gettext_accelerator_context(foo, bar, baz)]
       begins a region that uses the contexts "foo", "bar", and "baz".
       The comma is the delimiter; whitespace is optional.

     [gettext_accelerator_context()]
       ends the region.  */

B<msgaccel-prepare> removes from F<I<program>.pot> any
"gettext_accelerator_context" comments that B<xgettext --add-comments>
may have copied there.

B<msgaccel-prepare> warns if it does not find any contexts for some
use of an C<msgid> that contains the character specified with the
B<--accelerator-tag> option.  If the character does not actually
indicate an accelerator in that C<msgid> (e.g. "~" in "~/.bashrc"),
the warning can be silenced by specifying the special context
"IGNORE", which B<msgaccel-prepare> otherwise ignores.

=head1 OPTIONS

=over

=item B<-S>F<I<srcdir>>, B<--srcdir=>F<I<srcdir>>,
B<--source-directory=>F<I<srcdir>>

The directory to which the source references in "#:" lines are
relative.  Each use of this option adds one directory to the search
path.  If you do not specify this option, B<msgaccel-prepare>
implicitly searches the current directory.

=item B<--accelerator-tag=>I<character>

Specify the character that marks accelerators in C<msgid> strings.
B<msgaccel-prepare> looks up accelerator contexts for any C<msgid>
that contains this character.

Omitting the B<--accelerator-tag> option implies
B<--accelerator-tag="~">.  The option must be given to each program
separately because there is no standard way to save this information
in the PO file.

=back

=head1 ARGUMENTS

=over

=item F<I<program>.pot>

The file to augment with context information.  B<msgaccel-prepare>
first reads this file and then overwrites it.

Although this documentation keeps referring to F<I<program>.pot>,
you can also use B<msgaccel-prepare> on an already translated
F<I<language>.po>.  However, that will only work correctly if the
source references in the "#:" lines are still up to date.

=back

=head1 BUGS

B<msgaccel-prepare> assumes that source files are in the C programming
language: specifically, that a closing brace at the beginning of a
line marks the end of a function.

B<msgaccel-prepare> doesn't check whether the
"gettext_accelerator_context" comments actually are comments.

=head1 AUTHOR

Kalle Olavi Niemitalo <kon@iki.fi>

=head1 COPYRIGHT AND LICENSE

Copyright (c) 2005-2006 Kalle Olavi Niemitalo.

Permission to use, copy, modify, and distribute this software for any
purpose with or without fee is hereby granted, provided that the above
copyright notice and this permission notice appear in all copies.

THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.

=head1 SEE ALSO

L<msgaccel-check>, C<xgettext(1)>, C<msgmerge(1)>