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'
147 # Generate "all" without translations in more_texts.
148 # But keep the ones which are in both old_texts (texts) and more_texts,
149 # because this are ones which are overwritten in more_texts for custom usage.
151 $to_keep{$_} = 1 for grep { !!$self->{more_texts}{$_} } keys %old_texts;
152 my @new_all = grep { $to_keep{$_} || !$self->{more_texts}{$_} } sort keys %alllocales;
155 file => "$locales_dir/all",
156 header => $ALL_HEADER,
157 data_name => '$self->{texts}',
158 data_sub => sub { _print_line($_, $self->{texts}{$_}, @_) for @new_all },
161 open(my $js_file, '>:encoding(utf8)', $javascript_output_dir .'/locale/'. $locale .'.js') || die;
162 print $js_file 'namespace("kivi").setupLocale({';
164 for my $key (sort(keys(%jslocale))) {
165 my $trans = $self->{more_texts}{$key} // $self->{texts}{$key};
166 print $js_file ((!$first_entry ? ',' : '') ."\n". _double_quote($key) .':'. _double_quote($trans));
169 print $js_file ("\n");
170 print $js_file ('});'."\n");
173 foreach my $text (keys %$missing) {
174 if ($locale{$text} || $htmllocale{$text}) {
175 unless ($self->{texts}{$text}) {
176 $self->{texts}{$text} = $missing->{$text};
182 # calc and generate missing
183 # don't add missing ones if we have a translation in more_texts
184 my @new_missing = grep { !$self->{more_texts}{$_} && !$self->{texts}{$_} } sort keys %alllocales;
188 my %existing_lc = map { (lc $_ => $_) } grep { $self->{texts}->{$_} } keys %{ $self->{texts} };
189 foreach my $entry (@new_missing) {
190 my $other = $existing_lc{lc $entry};
191 print "W: No entry for '${entry}' exists, but there is one with different case: '${other}'\n" if $other;
196 for my $string (@new_missing) {
197 print "new string '$string' in files:\n";
200 apply { s{^(?:\.\./)+}{} }
201 grep { $cached{$_}{all}{$string} }
207 file => "$locales_dir/missing",
208 header => $MISSING_HEADER,
209 data_name => '$missing',
210 data_sub => sub { _print_line($_, '', @_) for @new_missing },
214 # calc and generate lost
215 while (my ($text, $translation) = each %old_texts) {
216 next if ($alllocales{$text});
217 push @lost, { 'text' => $text, 'translation' => $translation };
221 splice @lost, 0, (scalar @lost - 50) if (scalar @lost > 50);
223 file => "$locales_dir/lost",
224 header => $LOST_HEADER,
226 data_name => '@lost',
228 _print_line($_->{text}, $_->{translation}, @_, template => " { 'text' => %s, 'translation' => %s },\n") for @lost;
233 my $trlanguage = slurp("$locales_dir/LANGUAGE");
236 search_unused_htmlfiles() if $opt_c;
238 my $count = scalar keys %alllocales;
239 my $notext = scalar @new_missing;
240 my $per = sprintf("%.1f", ($count - $notext) / $count * 100);
241 print "\n$trlanguage - ${per}%";
242 print " - $notext/$count missing" if $notext;
251 # These are all the texts to build the translations files.
252 # The file has the form of 'english text' => 'foreign text',
253 # you can add the translation in this file or in the 'missing' file
254 # run locales.pl from this directory to rebuild the translation files
256 $MISSING_HEADER = <<EOL;
257 # add the missing texts and run locales.pl to rebuild
259 $LOST_HEADER = <<EOL;
260 # The last 50 text strings, that have been removed.
261 # This file has been auto-generated by locales.pl. Please don't edit!
268 my ($opt_no_c, $ignore_for_compatiblity);
271 'check-files' => \$ignore_for_compatiblity,
272 'no-check-files' => \$opt_no_c,
273 'verbose' => \$opt_v,
274 'filenames' => \$opt_f,
288 pod2usage(-exitstatus => 0, -verbose => 2);
293 my $arg = shift @ARGV;
295 foreach my $dir ("../locale/$arg", "locale/$arg", "../$arg", $arg) {
296 next unless -d $dir && -f "$dir/all" && -f "$dir/LANGUAGE";
305 print "The locale directory '$arg' could not be found.\n";
309 } elsif (!-f 'all' || !-f 'LANGUAGE') {
310 print "locales.pl was not called from a locale/* subdirectory,\n"
311 . "and no locale directory name was given.\n";
315 $locale ||= (grep { $_ } split m:/:, getcwd())[-1];
320 my ($file, $dir) = @_;
321 print "\n$file" if $opt_v;
325 &scanfile("$dir/$file");
327 # scan custom_{module}.pl or {login}_{module}.pl files
328 foreach my $customfile (@customfiles) {
329 if ($customfile =~ /_$file/) {
330 if (-f "$dir/$customfile") {
331 &scanfile("$dir/$customfile");
339 sub extract_text_between_parenthesis {
340 my ($fh, $line) = @_;
341 my ($inside_string, $pos, $text, $quote_next) = (undef, 0, "", 0);
344 if (length($line) <= $pos) {
346 return ($text, "") unless ($line);
350 my $cur_char = substr($line, $pos, 1);
352 if (!$inside_string) {
353 if ((length($line) >= ($pos + 3)) && (substr($line, $pos, 2)) eq "qq") {
354 $inside_string = substr($line, $pos + 2, 1);
357 } elsif ((length($line) >= ($pos + 2)) &&
358 (substr($line, $pos, 1) eq "q")) {
359 $inside_string = substr($line, $pos + 1, 1);
362 } elsif (($cur_char eq '"') || ($cur_char eq '\'')) {
363 $inside_string = $cur_char;
365 } elsif (($cur_char eq ")") || ($cur_char eq ',')) {
366 return ($text, substr($line, $pos + 1));
371 $text .= '\\' unless $cur_char eq "'";
375 } elsif ($cur_char eq '\\') {
378 } elsif ($cur_char eq $inside_string) {
379 undef($inside_string);
392 my $dont_include_subs = shift;
393 my $scanned_files = shift;
398 $scanned_files = {} unless ($scanned_files);
399 return if ($scanned_files->{$file});
400 $scanned_files->{$file} = 1;
402 if (!defined $cached{$file}) {
404 return unless (-f "$file");
406 my $fh = new FileHandle;
407 open $fh, '<:encoding(utf8)', $file or die "$! : $file";
409 my ($is_submit, $line_no, $sub_line_no) = (0, 0, 0);
412 last if /^\s*__END__/;
416 # is this another file
417 if (/require\s+\W.*\.pl/) {
419 $newfile =~ s/require\s+\W//;
420 $newfile =~ s|bin/mozilla||;
421 $cached{$file}{scan}{"$bindir/$newfile"} = 1;
422 } elsif (/use\s+SL::([\w:]*)/) {
425 $cached{$file}{scannosubs}{"../../SL/${module}.pm"} = 1;
428 # Some calls to render() are split over multiple lines. Deal
430 while (/(?:parse_html_template2?|render)\s*\( *$/) {
435 # is this a template call?
436 if (/(?:parse_html_template2?|render)\s*\(\s*[\"\']([\w\/]+)\s*[\"\']/) {
437 my $new_file_base = "$basedir/templates/webpages/$1.";
438 if (/parse_html_template2/) {
439 print "E: " . strip_base($file) . " is still using 'parse_html_template2' for " . strip_base("${new_file_base}html") . ".\n";
443 foreach my $ext (qw(html js json)) {
444 my $new_file = "${new_file_base}${ext}";
446 $cached{$file}{scanh}{$new_file} = 1;
452 if ($opt_c && !$found_one) {
453 print "W: missing HTML template: " . strip_base($new_file_base) . "{html,json,js} (referenced from " . strip_base($file) . ")\n";
462 my ($null, $country) = split(/,/);
463 $country =~ s/^ +[\"\']//;
464 $country =~ s/[\"\'].*//;
470 # is it a submit button before $locale->
471 if (/$submitsearch/) {
473 if ($` !~ /locale->text/) {
475 $sub_line_no = $line_no;
480 if (/ (?: locale->text | \b t8 ) \b .*? \(/x) {
487 ($string, $_) = extract_text_between_parenthesis($fh, $postmatch);
490 # if there is no $ in the string record it
491 unless (($string =~ /\$\D.*/) || ("" eq $string)) {
493 # this guarantees one instance of string
494 $cached{$file}{locale}{$string} = 1;
496 # this one is for all the locales
497 $cached{$file}{all}{$string} = 1;
499 # is it a submit button before $locale->
501 $cached{$file}{submit}{$string} = 1;
504 } elsif ($postmatch =~ />/) {
508 # exit loop if there are no more locales on this line
509 ($rc) = ($postmatch =~ /locale->text | \b t8/x);
511 if ( ($postmatch =~ />/)
512 || (!$found && ($sub_line_no != $line_no) && />/)) {
522 $alllocales{$_} = 1 for keys %{$cached{$file}{all}};
523 $locale{$_} = 1 for keys %{$cached{$file}{locale}};
524 $submit{$_} = 1 for keys %{$cached{$file}{submit}};
526 scanfile($_, 0, $scanned_files) for keys %{$cached{$file}{scan}};
527 scanfile($_, 1, $scanned_files) for keys %{$cached{$file}{scannosubs}};
528 scanhtmlfile($_) for keys %{$cached{$file}{scanh}};
530 $referenced_html_files{$_} = 1 for keys %{$cached{$file}{scanh}};
536 my $menu = SL::YAML::LoadFile($file);
538 for my $node (@$menu) {
539 # possible for override files
540 next unless exists $node->{name};
542 $locale{$node->{name}} = 1;
543 $alllocales{$node->{name}} = 1;
544 $cached{$file}{all}{$node->{name}} = 1;
549 # we only need to do this for auth atm, because only auth scripts can include new rights, which are translateable
552 my $dbu = SL::DBUpgrade2->new(auth => $auth, path => SL::System::Process->exe_dir . '/sql/Pg-upgrade2-auth');
554 for my $upgrade ($dbu->sort_dbupdate_controls) {
555 for my $string (@{ $upgrade->{locales} || [] }) {
556 $locale{$string} = 1;
557 $alllocales{$string} = 1;
558 $cached{$upgrade->{tag}}{all}{$string} = 1;
563 sub unescape_template_string {
574 return if defined $cached{$file};
576 my %plugins = ( 'loaded' => { }, 'needed' => { } );
578 if (!open(IN, '<:encoding(utf8)', $file)) {
579 print "E: template file '$file' not found\n";
586 while (my $line = <IN>) {
589 while ($line =~ m/\[\%[^\w]*use[^\w]+(\w+)[^\w]*?\%\]/gi) {
590 $plugins{loaded}->{$1} = 1;
593 while ($line =~ m/\[\%[^\w]*(\w+)\.\w+\(/g) {
595 $plugins{needed}->{$plugin} = 1 if (first { $_ eq $plugin } qw(HTML LxERP JavaScript JSON L P));
598 $plugins{needed}->{T8} = 1 if $line =~ m/\[\%.*\|.*\$T8/;
600 while ($line =~ m/(?: # Start von Variante 1: LxERP.t8('...'); ohne darumliegende [% ... %]-Tags
601 (LxERP\.t8)\( # LxERP.t8( ::Parameter $1::
602 ([\'\"]) # Anfang des zu übersetzenden Strings ::Parameter $2::
603 (.*?) # Der zu übersetzende String ::Parameter $3::
604 (?<!\\)\2 # Ende des zu übersetzenden Strings
605 | # Start von Variante 2: [% '...' | $T8 %]
606 \[\% # Template-Start-Tag
607 [\-~#]? # Whitespace-Unterdrückung
608 \s* # Optional beliebig viele Whitespace
609 ([\'\"]) # Anfang des zu übersetzenden Strings ::Parameter $4::
610 (.*?) # Der zu übersetzende String ::Parameter $5::
611 (?<!\\)\4 # Ende des zu übersetzenden Strings
612 \s*\|\s* # Pipe-Zeichen mit optionalen Whitespace davor und danach
613 (\$T8) # Filteraufruf ::Parameter $6::
614 .*? # Optionale Argumente für den Filter
616 [\-~#]? # Whitespace-Unterdrückung
617 \%\] # Template-Ende-Tag
620 my $module = $1 || $6;
621 my $string = $3 || $5;
622 print "Found filter >>>$string<<<\n" if $debug;
623 substr $line, $LAST_MATCH_START[1], $LAST_MATCH_END[0] - $LAST_MATCH_START[0], '';
625 $string = unescape_template_string($string);
626 $cached{$file}{all}{$string} = 1;
627 $cached{$file}{html}{$string} = 1;
628 $cached{$file}{submit}{$string} = 1 if $PREMATCH =~ /$submitsearch/;
629 $plugins{needed}->{T8} = 1 if $module eq '$T8';
630 $plugins{needed}->{LxERP} = 1 if $module eq 'LxERP.t8';
633 while ($line =~ m/\[\% # Template-Start-Tag
634 [\-~#]* # Whitespace-Unterdrückung
635 \s* # Optional beliebig viele Whitespace
636 (?: # Die erkannten Template-Direktiven
641 \s+ # Mindestens ein Whitespace
642 [\'\"]? # Anfang des Dateinamens
643 ([^\s]+) # Beliebig viele Nicht-Whitespaces -- Dateiname
644 \.(html|js) # Endung ".html" oder ".js", ansonsten kann es der Name eines Blocks sein
646 my $new_file_name = "$basedir/templates/webpages/$1.$2";
647 $cached{$file}{scanh}{$new_file_name} = 1;
648 substr $line, $LAST_MATCH_START[1], $LAST_MATCH_END[0] - $LAST_MATCH_START[0], '';
654 foreach my $plugin (keys %{ $plugins{needed} }) {
655 next if ($plugins{loaded}->{$plugin});
656 print "E: " . strip_base($file) . " requires the Template plugin '$plugin', but is not loaded with '[\% USE $plugin \%]'.\n";
659 # copy back into global arrays
660 $alllocales{$_} = 1 for keys %{$cached{$file}{all}};
661 $locale{$_} = 1 for keys %{$cached{$file}{html}};
662 $submit{$_} = 1 for keys %{$cached{$file}{submit}};
664 scanhtmlfile($_) for keys %{$cached{$file}{scanh}};
666 $referenced_html_files{$_} = 1 for keys %{$cached{$file}{scanh}};
669 sub scan_javascript_file {
672 open(my $fh, '<:encoding(utf8)', $file) || die('can not open file: '. $file);
674 while( my $line = readline($fh) ) {
685 my $text = unescape_template_string($2);
687 $jslocale{$text} = 1;
688 $alllocales{$text} = 1;
694 sub search_unused_htmlfiles {
695 my @unscanned_dirs = ('../../templates/webpages');
697 while (scalar @unscanned_dirs) {
698 my $dir = shift @unscanned_dirs;
700 foreach my $entry (<$dir/*>) {
702 push @unscanned_dirs, $entry;
704 } elsif (!$ignore_unused_templates{strip_base($entry)} && -f $entry && !$referenced_html_files{$entry}) {
705 print "W: unused HTML template: " . strip_base($entry) . "\n";
713 my $s = "$_[0]"; # Create a copy of the string.
716 $s =~ s|templates/webpages/||;
723 $val =~ s/(\'|\\$)/\\$1/g;
724 return "'" . $val . "'";
729 $val =~ s/(\"|\\$)/\\$1/g;
730 return '"'. $val .'"';
734 my $key = _single_quote(shift);
735 my $text = _single_quote(shift);
737 my $template = $params{template} || qq| %-29s => %s,\n|;
738 my $fh = $params{fh} || croak 'need filehandle in _print_line';
740 print $fh sprintf $template, $key, $text;
746 my $file = $params{file} || croak 'need filename in generate_file';
747 my $header = $params{header};
748 my $lines = $params{data_sub};
749 my $data_name = $params{data_name};
750 my @delim = split //, ($params{delim} || '{}');
752 open my $fh, '>:encoding(utf8)', $file or die "$! : $file";
754 print $fh "#!/usr/bin/perl\n# -*- coding: utf-8; -*-\n# vim: fenc=utf-8\n\nuse utf8;\n\n";
755 print $fh $header, "\n" if $header;
756 print $fh "$data_name = $delim[0]\n" if $data_name;
760 print $fh qq|$delim[1];\n\n1;\n|;
768 locales.pl - Collect strings for translation in kivitendo
772 locales.pl [options] lang_code
775 -c, --check-files Run extended checks on HTML files (default)
776 -n, --no-check-files Do not run extended checks on HTML files
777 -f, --filenames Show the filenames where new strings where found
778 -v, --verbose Be more verbose
779 -h, --help Show this help
785 =item B<-c>, B<--check-files>
787 Run extended checks on the usage of templates. This can be used to
788 discover HTML templates that are never used as well as the usage of
789 non-existing HTML templates. This is enabled by default.
791 =item B<-n>, B<--no-check-files>
793 Do not run extended checks on the usage of templates. See
796 =item B<-v>, B<--verbose>
804 This script collects strings from Perl files, the menu files and
805 HTML templates and puts them into the file "all" for translation.