po4a-translate

#! /usr/bin/env perl
eval 'exec perl -S $0 ${1+"$@"}'
  if $running_under_some_shell;

# po4a-translate -- translate doc files using a message catalog(ie, PO file)
#
# Copyright 2002-2023 by SPI, inc.
#
# This program is free software; you can redistribute it and/or modify it
# under the terms of GPL v2.0 or later (see COPYING).

=encoding UTF-8

=head1 NAME

po4a-translate - convert a PO file back to documentation format

=head1 SYNOPSIS

B<po4a-translate> B<-f> I<fmt> B<-m> I<master.doc> B<-p> I<XX.po> B<-l> I<XX.doc>

(I<XX.doc> is the output, all others are inputs)

=head1 DESCRIPTION

The po4a (PO for anything) project goal is to ease translations (and more
interestingly, the maintenance of translations) using gettext tools on
areas where they were not expected like documentation.

The B<po4a-translate> script is in charge of converting the translation
(which was done in a PO file) back into the documentation format. The
provided PO file should be the translation of the POT file which was
produced by L<po4a-gettextize(1)>.

=head1 OPTIONS

=over 4

=item B<-f>, B<--format>

Format of the documentation you want to handle. Use the B<--help-format>
option to see the list of available formats.

=item B<-a>, B<--addendum>

Add a file to the resulting file (to put translator's name or a section
"About this translation", for example). The first line of the file to insert
should be a PO4A header indicating where it should be added (see section
B<HOWTO add extra text to translations> in L<po4a(7)>).

=item B<-A>, B<--addendum-charset>

Charset of the addenda. Note that all the addenda should be in the same
charset.

=item B<-m>, B<--master>

File containing the master document to translate.

=item B<-M>, B<--master-charset>

Charset of the file containing the document to translate.

=item B<-l>, B<--localized>

File where the localized (translated) document should be written.

=item B<-L>, B<--localized-charset>

Charset of the file containing the localized document.

=item B<-p>, B<--po>

File from which the message catalog should be read.

=item B<-o>, B<--option>

Extra option(s) to pass to the format plugin. See the documentation of each
plugin for more information about the valid options and their meanings. For 
example, you could pass '-o tablecells' to the AsciiDoc parser, while the 
text parser would accept '-o tabs=split'.

=item B<-k>, B<--keep>

Minimal threshold for translation percentage to keep (i.e. write) the
resulting file (default: 80). I.e. by default, files have to be translated
at least at 80% to get written.

=item B<-w>, B<--width>

Column at which we should wrap the resulting file if the format supports it (default: 76).
If the provided value is 0 or negative, then the resulting file will not be wraped at all.
This is useful with some formats that could break when rewraped.

=item B<-h>, B<--help>

Show a short help message.

=item B<--help-format>

List the documentation formats understood by po4a.

=item B<-V>, B<--version>

Display the version of the script and exit.

=item B<-v>, B<--verbose>

Increase the verbosity of the program.

=item B<-d>, B<--debug>

Output some debugging information.

=item B<--porefs> I<type>[,B<wrap>|B<nowrap>]

Specify the reference format. Argument I<type> can be one of B<never>
to not produce any reference, B<file> to only specify the file
without the line number, B<counter> to replace line number by an
increasing counter, and B<full> to include complete references (default: full).

Argument can be followed by a comma and either B<wrap> or B<nowrap> keyword.
References are written by default on a single line.  The B<wrap> option wraps
references on several lines, to mimic B<gettext> tools (B<xgettext> and
B<msgmerge>).  This option will become the default in a future release, because
it is more sensible.  The B<nowrap> option is available so that users who want
to keep the old behavior can do so.

=back

=head1 Adding content (beside translations) to generated files

To add some extra content to the generated document beside what you
translated (like the name of the translator, or an "About this translation"
section), you should use the B<--addendum> option.

The first line of the addendum must be a header indicating where to put
it in the document (it can be before or after a given part of the
document).  The rest of the file will be added verbatim to the resulting
file without further processing.

Note that if po4a-translate fails to add one of the given files, it discards
the whole translation (because the missing file could be the one indicating
the author, what would prevent the users to contact him to report bugs in
the translation).

The header has a pretty rigid syntax. For more information on how to use
this feature and how it works, please refer to the L<po4a(7)> man page.

=head1 SEE ALSO

L<po4a-gettextize(1)>,
L<po4a-normalize(1)>,
L<po4a-updatepo(1)>,
L<po4a(7)>

=head1 AUTHORS

 Denis Barbier <barbier@linuxfr.org>
 Nicolas François <nicolas.francois@centraliens.net>
 Martin Quinson (mquinson#debian.org)

=head1 COPYRIGHT AND LICENSE

Copyright 2002-2023 by SPI, inc.

This program is free software; you may redistribute it and/or modify it
under the terms of GPL v2.0 or later (see the COPYING file).

=cut

use 5.16.0;
use strict;
use warnings;

use Locale::Po4a::Chooser;
use Locale::Po4a::TransTractor;
use Locale::Po4a::Common qw(wrap_msg gettext);

use Pod::Usage   qw(pod2usage);
use Getopt::Long qw(GetOptions);

Locale::Po4a::Common::textdomain("po4a");

sub show_version {
    Locale::Po4a::Common::show_version("po4a-translate");
    exit 0;
}

Getopt::Long::Configure( 'no_auto_abbrev', 'no_ignore_case' );
my ( $outfile,         $width,    $threshold ) = ( '-', 76, 80 );
my ( $help,            $help_fmt, @verbose, $debug, $ignored_wrappo, @addfiles, $format, @options, $no_deprecation );
my ( $master_filename, $po_filename );
my ( $mastchar,        $locchar, $addchar );
GetOptions(
    'help|h'      => \$help,
    'help-format' => \$help_fmt,

    'master|m=s'    => \$master_filename,
    'localized|l=s' => \$outfile,
    'po|p=s'        => \$po_filename,
    'addendum|a=s'  => \@addfiles,
    'format|f=s'    => \$format,

    'master-charset|M=s'    => \$mastchar,
    'localized-charset|L=s' => \$locchar,
    'addendum-charset|A=s'  => \$addchar,

    'option|o=s' => \@options,

    # undocumented option to silence the warning about po4a-translate being deprecated
    'no-deprecation' => \$no_deprecation,

    # undocumented and ignored option that we must accept for the tests to pass it to the other scripts
    'wrap-po=s' => \$ignored_wrappo,

    'width|w=s' => \$width,
    'verbose|v' => \@verbose,
    'debug|d'   => \$debug,
    'keep|k=s'  => \$threshold,

    'version|V' => \&show_version
) or pod2usage();

$help     && pod2usage( -verbose => 1, -exitval => 0 );
$help_fmt && Locale::Po4a::Chooser::list(0);

( defined($master_filename) && length($master_filename) ) || pod2usage();
( defined($po_filename) && length($po_filename) )         || pod2usage();
-e $master_filename || die wrap_msg( gettext("File %s does not exist."), $master_filename );
-e $po_filename     || die wrap_msg( gettext("File %s does not exist."), $po_filename );

my ( @pos, @masters );
push @pos,     $po_filename;
push @masters, $master_filename;

my %options = (
    "verbose" => scalar @verbose,
    "debug"   => $debug,
    "wrapcol" => $width
);

foreach (@options) {
    if (m/^([^=]*)=(.*)$/) {
        $options{$1} = "$2";
    } else {
        $options{$_} = 1;
    }
}

unless ($no_deprecation) {
    print wrap_msg(
        gettext(
                "po4a-translate is deprecated. The unified po4a(1) program is more convenient and less error prone. "
              . "Once configured, `po4a --no-update` can be used as a drop-in replacement to `po4a-translate`."
        )
    );
}

# parser
my $doc = Locale::Po4a::Chooser::new( $format, %options );

# Prepare the document to be used as translator, but not parser
$doc->process(
    'po_in_name'       => \@pos,
    'file_in_name'     => \@masters,
    'file_in_charset'  => $mastchar,
    'file_out_charset' => $locchar,
    'addendum_charset' => $addchar,
);

my ( $percent, $hit, $queries ) = $doc->stats();
my $error = 0;

print STDERR wrap_msg( gettext("%s is %s%% translated (%s of %s strings)."),
    $master_filename, $percent, $hit, $queries )
  if ( scalar @verbose ) && ( $percent >= $threshold );

if ( $percent < $threshold ) {
    print STDERR wrap_msg( gettext("Discard the translation of %s (only %s%% translated; need %s%%)."),
        $master_filename, $percent, $threshold );
    unlink($outfile) if ( -e $outfile );
} else {
    my %discarded = ();
    my @files     = ();
    while (@addfiles) {
        my $add = shift(@addfiles);
        my $modifiers;
        $add =~ s/^([@!?]+)// and $modifiers = $1;
        next if defined( $discarded{$add} );
        if ( defined $modifiers ) {
            if ( $modifiers =~ m/!/ ) {
                $discarded{$add} = 1;
                next;
            }
            next if ( $modifiers =~ m/\?/ and not -e $add );
            if ( $modifiers =~ m/@/ ) {
                open LIST, "<", "$add" or die wrap_msg( gettext("Cannot open %s: %s"), $add, $! );
                my @new_list = ();
                while (<LIST>) {
                    chomp;
                    s/\s+$//;
                    next if length($_) == 0 or $_ =~ m/^\s*#/;
                    push( @new_list, $_ );
                }
                close LIST;
                unshift( @addfiles, @new_list );
            } else {
                push @files, $add;
            }
        } else {
            push @files, $add;
        }
    }
    for my $add (@files) {
        unless ( $doc->addendum($add) ) {
            unlink($outfile) if ( -e $outfile );
            die wrap_msg( gettext("Discard the translation of %s (addendum %s does not apply)."),
                $master_filename, $add );
        }
    }
    $doc->write( $outfile, $locchar // '' );
}

1;