wesnoth/utils/wml2po.pl
2004-08-15 21:21:27 +00:00

255 lines
6.6 KiB
Perl
Executable File

#! /usr/bin/perl -w
# WARNING:
# DO NOT commit po-files convverted with this script yet. There is
# still some work to do !
# USAGE:
# ./utils/wml2po.pl data/translations/XXXXXXX.cfg po/XX.po
# TODO:
# - maybe try to detect when a single english string was translated in
# different ways, instead of just taking the 1st version ?
use strict;
## process wml files
require "utils/wmltrans.pm";
# get some id->english from english.cfg
our ($wmltransfile, $pofile) = @ARGV;
our %english = readwml ('data/translations/english.cfg');
our %revenglish;
sub set($$) {
my ($id, $str) = @_;
if (defined $english{$id}) {
print STDERR "WARNING: duplicate def for $id: $str\n";
return;
}
$english{$id}=$str;
# print STDERR "$id=$str\n";
}
# get ids from unit files
our @wmlfiles = glob ("data/units/*.cfg");
foreach my $wmlfile (@wmlfiles) {
open (WML, $wmlfile) or die "cannot open $wmlfile";
my ($id, $str);
my $readingattack = 0;
while (<WML>) {
if (m/id\s*=\s*(.*)/) {
$id = $str = $1;
$id =~ s/\s+//g;
set($id,$str);
# } elsif (m,\[/.*\],) {
# $id = undef;
} elsif (m/\[attack\]/) {
$readingattack = 1;
} elsif (m/ability\s*=\s*(.*)/) {
set('ability_' . $1, $1);
} elsif (m/unit_description\s*=\s*(?:_\s*)\"(.*)\"\s*$/) {
# single-line
if (defined $id) {
set($id . '_description',$1);
} else {
print STDERR "No id for unit_description $1\n";
}
} elsif ($readingattack) {
if (m/name=(.*)/) {
set('weapon_name_' . $1, $1);
} elsif (m/type=(.*)/) {
set('weapon_type_' . $1, $1);
} elsif (m/special=(.*)/) {
set('weapon_special_' . $1, $1);
}
}
}
close WML;
}
our %suffix = (
'difficulty_descriptions' => '_difficulties',
'cannot_use_message' => '_cannot_use_message',
'objectives' => '_objectives',
);
# get ids from other wml files
@wmlfiles = (qw(data/game.cfg data/help.cfg data/items.cfg data/multiplayer.cfg
data/schedules.cfg),
glob ("data/scenarios/*/*.cfg"));
foreach my $wmlfile (@wmlfiles) {
open (WML, $wmlfile) or die "cannot open $wmlfile";
print STDERR " Processing $wmlfile\n";
my $id = 'no_id';
my ($str, $key, $value,$strtype);
while (<WML>) {
# print STDERR $_;
next if m/^\s*\#/ and !defined $key;
if (m/id\s*=\s*"(.*)"/) {
$id = $1;
} elsif (m/id\s*=\s*(.*)/) {
$id = $1;
# print STDERR "--> $id\n";
# } elsif (m,\[/.*\],) {
# $id = undef;
} elsif (m/(difficulty\_descriptions|cannot\_use\_message|objectives)\s*=\s*(?:_\s*)?\"(.*)\"\s*$/) {
# single-line
die "nested key" if defined $key;
if (defined $id) {
set($id . $suffix{$1},$2);
} else {
print STDERR "No id for $1 $2\n";
}
} elsif (m/(difficulty\_descriptions|cannot\_use\_message|objectives)\s*=\s*(?:_\s*)?\"(.*)\s*$/) {
# start of multi-line
die "nested key" if defined $key;
$strtype=1;
$key = $id . $suffix{$1};
$value = $2 . "\n";
} elsif (m/(?:title|message|name|story)\s*=\s*(?:_\s*)\"(.*)\"\s*$/) {
# single-line
set($id,$1);
} elsif (m/(title|message|name|story)\s*=\s*(?:_\s*)\"(.*)\s*$/) {
# start of multi-line
die "nested key" if defined $key;
$strtype = 2;
$key = $1;
$value = $2 . "\n";
} elsif (m/(\S+)\s*=\s*(?:_\s*)?\"(.*)\"\s*$/) {
# ignored single-line
} elsif (m/(\S+)\s*=\s*(?:_\s*)?\s*\"(.*)/) {
# start of ignored multi-line
$strtype = 0;
$value = $2 . "\n";
} elsif (m/(.*)\"\s*$/ and $strtype == 1) {
# end of multi-line
die "end of string without a key" unless defined $key;
$value .= $1;
set($key,$value);
$key = undef;
} elsif (m/(.*)\"\s*$/ and $strtype == 2) {
# end of multi-line
die "end of string without a key" unless defined $key;
$value .= $1;
set($id,$value);
$key = undef;
} elsif (m/(.*)\"\s*$/ and $strtype == 0) {
# end of ignored multi-line
} elsif (defined $key) {
# part of multi-line
$value .= $_;
}
}
close WML;
}
#exit; # this is for debug only :)
# reverse the hash to use the strings to find the id
foreach my $key (keys %english) {
push @{$revenglish{$english{$key}}}, $key;
}
# get translations from wml file
our %lang = readwml ($wmltransfile);
## process pofile, filling holes when we can
our $migrated = 0;
open (POFILE, $pofile) or die "cannot open $pofile";
my ($curid, $curmsg);
while (<POFILE>) {
if (m/^msgid (".*")$/) {
$curid = "$1\n"; $curmsg = undef;
} elsif (m/^msgstr (".*")$/) {
$curmsg = "$1\n";
} elsif (m/^(".*")$/) {
if (!defined $curmsg) {
$curid .= "$_";
} else {
$curmsg .= "$_";
}
} elsif (m/^$/) {
if (defined $curid and defined $curmsg) {
processentry ($curid, $curmsg);
print "\n";
} else {
print;
}
} else {
print;
}
}
# last entry does not have an empty line afterwards
if (defined $curid and defined $curmsg) {
processentry ($curid, $curmsg);
}
## keep a list of previously translated entries not migrated
my $n = (keys %lang) - 1; # initial entry does not count
if ($n == 0) {
print STDERR "\nFINAL WORD: ALL $migrated TRANSLATIONS MIGRATED !\n";
} else {
my $total = $migrated+$n;
my $percent = (100 * $n) / $total;
print STDERR "\nFINAL WARNING: the following $n/$total ($percent\%) translations were NOT migrated:\n";
foreach my $key (keys %lang) {
print STDERR $key, ' = "', $lang{$key}, "\"\n";
}
# $n = keys %english;
# print STDERR "\ALSO FINAL WARNING: the following $n english strings were NOT looked up:\n";
# foreach my $key (keys %english) {
# print STDERR $key, ' = "', $english{$key}, "\"\n";
# }
}
## Below is utility functions only - end of processing
sub processentry {
my ($curid, $curmsg) = @_;
my $output;
my $touched = 0;
# lookup
if ($curmsg eq "\"\"\n") {
my $ids = $revenglish{po2rawstring($curid)};
if (defined $ids) {
my $id = $ids->[0];
my $trans = $lang{$id};
if (defined $trans) {
$output = raw2postring($trans);
$touched = 1;
foreach my $anid (@{$ids}) { delete $lang{$anid}; };
delete $english{$id};
$migrated++;
} else {
# print STDERR "WARNING: no translation found for $id - setting to empty\n";
$output = raw2postring("");
}
} else {
# this usually denotes limitations in this script
print STDERR "WARNING: no id found (setting translation to empty) for $curid\n";
$output = raw2postring("");
}
} else {
$output = $curmsg;
}
print "#, fuzzy\n" if $touched;
print "msgid $curid";
print "msgstr $output";
$curid = undef; $curmsg = undef;
}