3 # this version of locales processes not only all required .pl files
4 # but also all parse_html_templated files.
12 unshift(@INC, $FindBin::Bin . '/../modules/override'); # Use our own versions of various modules (e.g. YAML).
13 push (@INC, $FindBin::Bin . '/..');
20 use File::Slurp qw(slurp);
24 use List::MoreUtils qw(apply);
25 use List::Util qw(first);
28 use SL::System::Process;
31 $OUTPUT_AUTOFLUSH = 1;
41 my $basedir = "../..";
42 my $locales_dir = ".";
43 my $bindir = "$basedir/bin/mozilla";
44 my @progdirs = ( "$basedir/SL" );
45 my @menufiles = glob("${basedir}/menus/*/*");
46 my @javascript_dirs = ($basedir .'/js', $basedir .'/templates/webpages');
47 my $javascript_output_dir = $basedir .'/js';
48 my $submitsearch = qr/type\s*=\s*[\"\']?submit/i;
53 my %ignore_unused_templates = (
54 map { $_ => 1 } qw(ct/testpage.html oe/periodic_invoices_email.txt part/testpage.html t/render.html t/render.js task_server/failure_notification_email.txt
55 failed_background_jobs_report/email.txt)
58 my (%referenced_html_files, %locale, %htmllocale, %alllocales, %cached, %submit, %jslocale);
59 my ($ALL_HEADER, $MISSING_HEADER, $LOST_HEADER);
64 my ($top_dir_name) = @_;
71 tie my %dir_h, 'IO::Dir', $dir_name;
73 push @files, grep { -f } map { "${dir_name}/${_}" } keys %dir_h;
74 my @sub_dirs = grep { -d } map { "${dir_name}/${_}" } grep { ! m/^\.\.?$/ } keys %dir_h;
76 $finder->($_) for @sub_dirs;
79 $finder->($top_dir_name);
85 # overwrite existing entries with the ones from 'missing'
86 $self->{texts}->{$_} = $missing->{$_} for grep { $missing->{$_} } keys %alllocales;
88 # try to set missing entries from lost ones
89 my %lost_by_text = map { ($_->{text} => $_->{translation}) } @lost;
90 $self->{texts}->{$_} = $lost_by_text{$_} for grep { !$self->{texts}{$_} } keys %alllocales;
93 my @bindir_files = find_files($bindir);
94 my @progfiles = map { m:^(.+)/([^/]+)$:; [ $2, $1 ] } grep { /\.pl$/ && !/_custom/ } @bindir_files;
95 my @customfiles = grep /_custom/, @bindir_files;
97 push @progfiles, map { m:^(.+)/([^/]+)$:; [ $2, $1 ] } grep { /\.pm$/ } map { find_files($_) } @progdirs;
102 foreach my $sub_dir ("Pg-upgrade2", "Pg-upgrade2-auth") {
103 my $dir = "$basedir/sql/$sub_dir";
104 tie %dir_h, 'IO::Dir', $dir;
105 push @dbplfiles, map { [ $_, $dir ] } grep { /\.pl$/ } keys %dir_h;
108 # slurp the translations in
109 if (-f "$locales_dir/all") {
110 require "$locales_dir/all";
112 # load custom translation (more_texts)
113 for my $file (glob("${locales_dir}/more/*")) {
114 if (open my $in, "<", "$file") {
119 $self->{more_texts_temp}{$_} = $self->{more_texts}{$_} for keys %{ $self->{more_texts} };
122 $self->{more_texts} = delete $self->{more_texts_temp};
124 if (-f "$locales_dir/missing") {
125 require "$locales_dir/missing" ;
126 unlink "$locales_dir/missing";
128 if (-f "$locales_dir/lost") {
129 require "$locales_dir/lost";
130 unlink "$locales_dir/lost";
133 my %old_texts = %{ $self->{texts} || {} };
135 handle_file(@{ $_ }) for @progfiles;
136 handle_file(@{ $_ }) for @dbplfiles;
137 scanmenu($_) for @menufiles;
140 for my $file_name (grep { /\.(?:js|html)$/i } map({find_files($_)} @javascript_dirs)) {
141 scan_javascript_file($file_name);
144 # merge entries to translate with entries from files 'missing' and 'lost'
149 file => "$locales_dir/all",
150 header => $ALL_HEADER,
151 data_name => '$self->{texts}',
152 data_sub => sub { _print_line($_, $self->{texts}{$_}, @_) for sort keys %alllocales },
155 open(my $js_file, '>:encoding(utf8)', $javascript_output_dir .'/locale/'. $locale .'.js') || die;
156 print $js_file 'namespace("kivi").setupLocale({';
158 for my $key (sort(keys(%jslocale))) {
159 print $js_file ((!$first_entry ? ',' : '') ."\n". _double_quote($key) .':'. _double_quote($self->{texts}{$key}));
162 print $js_file ("\n");
163 print $js_file ('});'."\n");
166 foreach my $text (keys %$missing) {
167 if ($locale{$text} || $htmllocale{$text}) {
168 unless ($self->{texts}{$text}) {
169 $self->{texts}{$text} = $missing->{$text};
175 # calc and generate missing
176 # don't add missing ones if we have a translation in more_texts
177 my @new_missing = grep { !$self->{more_texts}{$_} && !$self->{texts}{$_} } sort keys %alllocales;
181 my %existing_lc = map { (lc $_ => $_) } grep { $self->{texts}->{$_} } keys %{ $self->{texts} };
182 foreach my $entry (@new_missing) {
183 my $other = $existing_lc{lc $entry};
184 print "W: No entry for '${entry}' exists, but there is one with different case: '${other}'\n" if $other;
189 for my $string (@new_missing) {
190 print "new string '$string' in files:\n";
193 apply { s{^(?:\.\./)+}{} }
194 grep { $cached{$_}{all}{$string} }
200 file => "$locales_dir/missing",
201 header => $MISSING_HEADER,
202 data_name => '$missing',
203 data_sub => sub { _print_line($_, '', @_) for @new_missing },
207 # calc and generate lost
208 while (my ($text, $translation) = each %old_texts) {
209 next if ($alllocales{$text});
210 push @lost, { 'text' => $text, 'translation' => $translation };
214 splice @lost, 0, (scalar @lost - 50) if (scalar @lost > 50);
216 file => "$locales_dir/lost",
217 header => $LOST_HEADER,
219 data_name => '@lost',
221 _print_line($_->{text}, $_->{translation}, @_, template => " { 'text' => %s, 'translation' => %s },\n") for @lost;
226 my $trlanguage = slurp("$locales_dir/LANGUAGE");
229 search_unused_htmlfiles() if $opt_c;
231 my $count = scalar keys %alllocales;
232 my $notext = scalar @new_missing;
233 my $per = sprintf("%.1f", ($count - $notext) / $count * 100);
234 print "\n$trlanguage - ${per}%";
235 print " - $notext/$count missing" if $notext;
244 # These are all the texts to build the translations files.
245 # The file has the form of 'english text' => 'foreign text',
246 # you can add the translation in this file or in the 'missing' file
247 # run locales.pl from this directory to rebuild the translation files
249 $MISSING_HEADER = <<EOL;
250 # add the missing texts and run locales.pl to rebuild
252 $LOST_HEADER = <<EOL;
253 # The last 50 text strings, that have been removed.
254 # This file has been auto-generated by locales.pl. Please don't edit!
261 my ($opt_no_c, $ignore_for_compatiblity);
264 'check-files' => \$ignore_for_compatiblity,
265 'no-check-files' => \$opt_no_c,
266 'verbose' => \$opt_v,
267 'filenames' => \$opt_f,
281 pod2usage(-exitstatus => 0, -verbose => 2);
286 my $arg = shift @ARGV;
288 foreach my $dir ("../locale/$arg", "locale/$arg", "../$arg", $arg) {
289 next unless -d $dir && -f "$dir/all" && -f "$dir/LANGUAGE";
298 print "The locale directory '$arg' could not be found.\n";
302 } elsif (!-f 'all' || !-f 'LANGUAGE') {
303 print "locales.pl was not called from a locale/* subdirectory,\n"
304 . "and no locale directory name was given.\n";
308 $locale ||= (grep { $_ } split m:/:, getcwd())[-1];
313 my ($file, $dir) = @_;
314 print "\n$file" if $opt_v;
318 &scanfile("$dir/$file");
320 # scan custom_{module}.pl or {login}_{module}.pl files
321 foreach my $customfile (@customfiles) {
322 if ($customfile =~ /_$file/) {
323 if (-f "$dir/$customfile") {
324 &scanfile("$dir/$customfile");
332 sub extract_text_between_parenthesis {
333 my ($fh, $line) = @_;
334 my ($inside_string, $pos, $text, $quote_next) = (undef, 0, "", 0);
337 if (length($line) <= $pos) {
339 return ($text, "") unless ($line);
343 my $cur_char = substr($line, $pos, 1);
345 if (!$inside_string) {
346 if ((length($line) >= ($pos + 3)) && (substr($line, $pos, 2)) eq "qq") {
347 $inside_string = substr($line, $pos + 2, 1);
350 } elsif ((length($line) >= ($pos + 2)) &&
351 (substr($line, $pos, 1) eq "q")) {
352 $inside_string = substr($line, $pos + 1, 1);
355 } elsif (($cur_char eq '"') || ($cur_char eq '\'')) {
356 $inside_string = $cur_char;
358 } elsif (($cur_char eq ")") || ($cur_char eq ',')) {
359 return ($text, substr($line, $pos + 1));
364 $text .= '\\' unless $cur_char eq "'";
368 } elsif ($cur_char eq '\\') {
371 } elsif ($cur_char eq $inside_string) {
372 undef($inside_string);
385 my $dont_include_subs = shift;
386 my $scanned_files = shift;
391 $scanned_files = {} unless ($scanned_files);
392 return if ($scanned_files->{$file});
393 $scanned_files->{$file} = 1;
395 if (!defined $cached{$file}) {
397 return unless (-f "$file");
399 my $fh = new FileHandle;
400 open $fh, '<:encoding(utf8)', $file or die "$! : $file";
402 my ($is_submit, $line_no, $sub_line_no) = (0, 0, 0);
405 last if /^\s*__END__/;
409 # is this another file
410 if (/require\s+\W.*\.pl/) {
412 $newfile =~ s/require\s+\W//;
413 $newfile =~ s|bin/mozilla||;
414 $cached{$file}{scan}{"$bindir/$newfile"} = 1;
415 } elsif (/use\s+SL::([\w:]*)/) {
418 $cached{$file}{scannosubs}{"../../SL/${module}.pm"} = 1;
421 # Some calls to render() are split over multiple lines. Deal
423 while (/(?:parse_html_template2?|render)\s*\( *$/) {
428 # is this a template call?
429 if (/(?:parse_html_template2?|render)\s*\(\s*[\"\']([\w\/]+)\s*[\"\']/) {
430 my $new_file_base = "$basedir/templates/webpages/$1.";
431 if (/parse_html_template2/) {
432 print "E: " . strip_base($file) . " is still using 'parse_html_template2' for " . strip_base("${new_file_base}html") . ".\n";
436 foreach my $ext (qw(html js json)) {
437 my $new_file = "${new_file_base}${ext}";
439 $cached{$file}{scanh}{$new_file} = 1;
445 if ($opt_c && !$found_one) {
446 print "W: missing HTML template: " . strip_base($new_file_base) . "{html,json,js} (referenced from " . strip_base($file) . ")\n";
455 my ($null, $country) = split(/,/);
456 $country =~ s/^ +[\"\']//;
457 $country =~ s/[\"\'].*//;
463 # is it a submit button before $locale->
464 if (/$submitsearch/) {
466 if ($` !~ /locale->text/) {
468 $sub_line_no = $line_no;
473 if (/ (?: locale->text | \b t8 ) \b .*? \(/x) {
480 ($string, $_) = extract_text_between_parenthesis($fh, $postmatch);
483 # if there is no $ in the string record it
484 unless (($string =~ /\$\D.*/) || ("" eq $string)) {
486 # this guarantees one instance of string
487 $cached{$file}{locale}{$string} = 1;
489 # this one is for all the locales
490 $cached{$file}{all}{$string} = 1;
492 # is it a submit button before $locale->
494 $cached{$file}{submit}{$string} = 1;
497 } elsif ($postmatch =~ />/) {
501 # exit loop if there are no more locales on this line
502 ($rc) = ($postmatch =~ /locale->text | \b t8/x);
504 if ( ($postmatch =~ />/)
505 || (!$found && ($sub_line_no != $line_no) && />/)) {
515 $alllocales{$_} = 1 for keys %{$cached{$file}{all}};
516 $locale{$_} = 1 for keys %{$cached{$file}{locale}};
517 $submit{$_} = 1 for keys %{$cached{$file}{submit}};
519 scanfile($_, 0, $scanned_files) for keys %{$cached{$file}{scan}};
520 scanfile($_, 1, $scanned_files) for keys %{$cached{$file}{scannosubs}};
521 scanhtmlfile($_) for keys %{$cached{$file}{scanh}};
523 $referenced_html_files{$_} = 1 for keys %{$cached{$file}{scanh}};
529 my $menu = SL::YAML::LoadFile($file);
531 for my $node (@$menu) {
532 # possible for override files
533 next unless exists $node->{name};
535 $locale{$node->{name}} = 1;
536 $alllocales{$node->{name}} = 1;
537 $cached{$file}{all}{$node->{name}} = 1;
542 # we only need to do this for auth atm, because only auth scripts can include new rights, which are translateable
545 my $dbu = SL::DBUpgrade2->new(auth => $auth, path => SL::System::Process->exe_dir . '/sql/Pg-upgrade2-auth');
547 for my $upgrade ($dbu->sort_dbupdate_controls) {
548 for my $string (@{ $upgrade->{locales} || [] }) {
549 $locale{$string} = 1;
550 $alllocales{$string} = 1;
551 $cached{$upgrade->{tag}}{all}{$string} = 1;
556 sub unescape_template_string {
567 return if defined $cached{$file};
569 my %plugins = ( 'loaded' => { }, 'needed' => { } );
571 if (!open(IN, '<:encoding(utf8)', $file)) {
572 print "E: template file '$file' not found\n";
579 while (my $line = <IN>) {
582 while ($line =~ m/\[\%[^\w]*use[^\w]+(\w+)[^\w]*?\%\]/gi) {
583 $plugins{loaded}->{$1} = 1;
586 while ($line =~ m/\[\%[^\w]*(\w+)\.\w+\(/g) {
588 $plugins{needed}->{$plugin} = 1 if (first { $_ eq $plugin } qw(HTML LxERP JavaScript JSON L P));
591 $plugins{needed}->{T8} = 1 if $line =~ m/\[\%.*\|.*\$T8/;
593 while ($line =~ m/(?: # Start von Variante 1: LxERP.t8('...'); ohne darumliegende [% ... %]-Tags
594 (LxERP\.t8)\( # LxERP.t8( ::Parameter $1::
595 ([\'\"]) # Anfang des zu übersetzenden Strings ::Parameter $2::
596 (.*?) # Der zu übersetzende String ::Parameter $3::
597 (?<!\\)\2 # Ende des zu übersetzenden Strings
598 | # Start von Variante 2: [% '...' | $T8 %]
599 \[\% # Template-Start-Tag
600 [\-~#]? # Whitespace-Unterdrückung
601 \s* # Optional beliebig viele Whitespace
602 ([\'\"]) # Anfang des zu übersetzenden Strings ::Parameter $4::
603 (.*?) # Der zu übersetzende String ::Parameter $5::
604 (?<!\\)\4 # Ende des zu übersetzenden Strings
605 \s*\|\s* # Pipe-Zeichen mit optionalen Whitespace davor und danach
606 (\$T8) # Filteraufruf ::Parameter $6::
607 .*? # Optionale Argumente für den Filter
609 [\-~#]? # Whitespace-Unterdrückung
610 \%\] # Template-Ende-Tag
613 my $module = $1 || $6;
614 my $string = $3 || $5;
615 print "Found filter >>>$string<<<\n" if $debug;
616 substr $line, $LAST_MATCH_START[1], $LAST_MATCH_END[0] - $LAST_MATCH_START[0], '';
618 $string = unescape_template_string($string);
619 $cached{$file}{all}{$string} = 1;
620 $cached{$file}{html}{$string} = 1;
621 $cached{$file}{submit}{$string} = 1 if $PREMATCH =~ /$submitsearch/;
622 $plugins{needed}->{T8} = 1 if $module eq '$T8';
623 $plugins{needed}->{LxERP} = 1 if $module eq 'LxERP.t8';
626 while ($line =~ m/\[\% # Template-Start-Tag
627 [\-~#]* # Whitespace-Unterdrückung
628 \s* # Optional beliebig viele Whitespace
629 (?: # Die erkannten Template-Direktiven
634 \s+ # Mindestens ein Whitespace
635 [\'\"]? # Anfang des Dateinamens
636 ([^\s]+) # Beliebig viele Nicht-Whitespaces -- Dateiname
637 \.(html|js) # Endung ".html" oder ".js", ansonsten kann es der Name eines Blocks sein
639 my $new_file_name = "$basedir/templates/webpages/$1.$2";
640 $cached{$file}{scanh}{$new_file_name} = 1;
641 substr $line, $LAST_MATCH_START[1], $LAST_MATCH_END[0] - $LAST_MATCH_START[0], '';
647 foreach my $plugin (keys %{ $plugins{needed} }) {
648 next if ($plugins{loaded}->{$plugin});
649 print "E: " . strip_base($file) . " requires the Template plugin '$plugin', but is not loaded with '[\% USE $plugin \%]'.\n";
652 # copy back into global arrays
653 $alllocales{$_} = 1 for keys %{$cached{$file}{all}};
654 $locale{$_} = 1 for keys %{$cached{$file}{html}};
655 $submit{$_} = 1 for keys %{$cached{$file}{submit}};
657 scanhtmlfile($_) for keys %{$cached{$file}{scanh}};
659 $referenced_html_files{$_} = 1 for keys %{$cached{$file}{scanh}};
662 sub scan_javascript_file {
665 open(my $fh, '<:encoding(utf8)', $file) || die('can not open file: '. $file);
667 while( my $line = readline($fh) ) {
678 my $text = unescape_template_string($2);
680 $jslocale{$text} = 1;
681 $alllocales{$text} = 1;
687 sub search_unused_htmlfiles {
688 my @unscanned_dirs = ('../../templates/webpages');
690 while (scalar @unscanned_dirs) {
691 my $dir = shift @unscanned_dirs;
693 foreach my $entry (<$dir/*>) {
695 push @unscanned_dirs, $entry;
697 } elsif (!$ignore_unused_templates{strip_base($entry)} && -f $entry && !$referenced_html_files{$entry}) {
698 print "W: unused HTML template: " . strip_base($entry) . "\n";
706 my $s = "$_[0]"; # Create a copy of the string.
709 $s =~ s|templates/webpages/||;
716 $val =~ s/(\'|\\$)/\\$1/g;
717 return "'" . $val . "'";
722 $val =~ s/(\"|\\$)/\\$1/g;
723 return '"'. $val .'"';
727 my $key = _single_quote(shift);
728 my $text = _single_quote(shift);
730 my $template = $params{template} || qq| %-29s => %s,\n|;
731 my $fh = $params{fh} || croak 'need filehandle in _print_line';
733 print $fh sprintf $template, $key, $text;
739 my $file = $params{file} || croak 'need filename in generate_file';
740 my $header = $params{header};
741 my $lines = $params{data_sub};
742 my $data_name = $params{data_name};
743 my @delim = split //, ($params{delim} || '{}');
745 open my $fh, '>:encoding(utf8)', $file or die "$! : $file";
747 print $fh "#!/usr/bin/perl\n# -*- coding: utf-8; -*-\n# vim: fenc=utf-8\n\nuse utf8;\n\n";
748 print $fh $header, "\n" if $header;
749 print $fh "$data_name = $delim[0]\n" if $data_name;
753 print $fh qq|$delim[1];\n\n1;\n|;
761 locales.pl - Collect strings for translation in kivitendo
765 locales.pl [options] lang_code
768 -c, --check-files Run extended checks on HTML files (default)
769 -n, --no-check-files Do not run extended checks on HTML files
770 -f, --filenames Show the filenames where new strings where found
771 -v, --verbose Be more verbose
772 -h, --help Show this help
778 =item B<-c>, B<--check-files>
780 Run extended checks on the usage of templates. This can be used to
781 discover HTML templates that are never used as well as the usage of
782 non-existing HTML templates. This is enabled by default.
784 =item B<-n>, B<--no-check-files>
786 Do not run extended checks on the usage of templates. See
789 =item B<-v>, B<--verbose>
797 This script collects strings from Perl files, the menu files and
798 HTML templates and puts them into the file "all" for translation.