3 # Open \______ \ ____ ____ | | _\_ |__ _______ ___
4 # Source | _// _ \_/ ___\| |/ /| __ \ / _ \ \/ /
5 # Jukebox | | ( <_> ) \___| < | \_\ ( <_> > < <
6 # Firmware |____|_ /\____/ \___ >__|_ \|___ /\____/__/\_ \
10 # Copyright (C) 2006 - 2008 by Daniel Stenberg
13 # See apps/language.c (TODO: Use common include for both)
14 # Cookie and binary version for the binary lang file
15 my $LANGUAGE_COOKIE = 0x1a;
16 my $LANGUAGE_VERSION = 0x06;
17 my $LANGUAGE_FLAG_RTL = 0x01;
20 my $SUBHEADER_SIZE = 6;
22 # A note for future users and readers: The original v1 language system allowed
23 # the build to create and use a different language than english built-in. We
24 # removed that feature from our build-system, but the build scripts still had
25 # the ability. But, starting now, this ability is no longer provided since I
26 # figured it was boring and unnecessary to write support for now since we
27 # don't use it anymore.
32 Usage: genlang [options] <langv2 file>
35 Make the tool create a [prefix].c and [prefix].h file.
38 Make the tool create a binary language (.lng) file named [outfile].
39 The use of this option requires that you also use -e, -t and -i.
42 Update language file. Given the translated file and the most recent english
43 file, you\'ll get an updated version sent to stdout. Suitable action to do
44 when you intend to update a translation.
46 -e=<english lang file>
47 Point out the english (original source) file, to use that as master
48 language template. Used in combination with -b, -u or -s.
51 Sort the Update language file in the same order as the strings in the
55 Specify which target you want the translations/phrases for. Required when
58 The target can in fact be specified as numerous different strings,
59 separated with colons. This will make genlang to use all the specified
60 strings when searching for a matching phrase.
63 The target id number, needed for -b.
66 Voice mode output. Outputs all id: and voice: lines for the given target!
69 Enables verbose (debug) output.
77 # 1) scan the english file, keep the whole <phrase> for each phrase.
78 # 2) read the translated file, for each end of phrase, compare:
79 # A) all source strings, if there's any change there should be a comment about
83 # 3) output the phrase with the comments from above
84 # 4) check which phrases that the translated version didn't have, and spit out
85 # the english version of those
96 my $check = ($binary?
1:0) + ($prefix?
1:0) + ($update?
1:0) + ($voiceout?
1:0) + ($sortfile?
1:0);
99 print STDERR
"Please use only one of -p, -u, -o, -b and -s\n";
103 print STDERR
"Please use at least one of -p, -u, -o, -b and -s\n";
108 if(($binary || $update || $voiceout || $sortfile) && !$english) {
109 print STDERR
"Please use -e too when you use -b, -o, -u or -s\n";
114 if($binary && !$target_id) {
115 print STDERR
"Please specify a target id number (with -i)!\n";
120 if(!$target && !$update && !$sortfile) {
121 print STDERR
"Please specify a target (with -t)!\n";
124 my @target_parts = split ':', $target;
128 my %id; # string to num hash
129 my @idnum; # num to string array
131 my %allphrases; # For sorting - an array of the <phrase> elements
132 my %source; # id string to source phrase hash
133 my %dest; # id string to dest phrase hash
134 my %voice; # id string to voice phrase hash
139 my $input = $ARGV[0];
153 my ($full, $n, $v)=@_;
159 my ($full, $n, $v)=@_;
165 my ($full, $n, $v)=@_;
170 my ($debug, $strref, $full, $n, $v)=@_;
172 my @all= split(" *, *", $n);
178 $regexp = qr/^$test\z/;
179 # print "TEST ($debug) $target for $test\n";
180 for my $part (@target_parts) {
181 if($part =~ $regexp) {
183 # print "MATCH: $test => $v\n";
193 parsetarget
("src", \
$src, @_);
198 parsetarget
("dest", \
$dest, @_);
203 parsetarget
("voice", \
$voice, @_);
209 # For the cases where the english file needs to be scanned/read, we do
210 # it before we read the translated file. For -b it isn't necessary, but for
211 # -u it is convenient.
213 my @idnum = ((0)); # start with a true number
214 my @vidnum = ((0x8000)); # first voice id
215 open(ENG
, "<$english") || die "Error: can't open $english";
222 my $numusers = 1; # core is already in the users map
226 # get rid of DOS newlines
229 if($_ =~ /^ *\<phrase\>/) {
230 # this is the start of a phrase
232 elsif($_ =~ /^ *\<\/phrase\
>/) {
234 # if id is something, when we count and store this phrase
236 # voice-only entries get a difference range
237 if($id =~ /^VOICE_/) {
238 # Assign an ID number to this entry
239 $idmap[$user]{$id}=$vidnum[$user];
243 # Assign an ID number to this entry
244 $idmap[$user]{$id}=$idnum[$user];
246 # print STDERR "DEST: bumped idnum to $idnum[$user]\n";
249 # this is the end of a phrase, add it to the english hash
250 $english{$id}=join("", @phrase);
256 # gather everything related to this phrase
258 if($_ =~ /^ *\<dest\>/i) {
262 elsif($withindest && ($_ =~ /^ *\<\/dest\
>/i
)) {
265 if($update || ($deststr && ($deststr !~ /^none\z/i))) {
266 # we unconditionally always use all IDs when the "update"
269 # print "DEST: use this id $id\n";
272 # print "skip $maybeid for $name\n";
275 elsif($withindest && ($_ =~ / *([^:]+): *(.*)/)) {
276 my ($name, $val)=($1, $2);
277 $dest=""; # in case it is left untouched for when the
278 # model name isn't "our"
279 dest
($_, $name, $val);
282 # Store the current dest string. If this target matches
283 # multiple strings, it will get updated several times.
289 if($_ =~ /^ *id: ([^ \t\n]+)/i) {
291 $sortorder{$maybeid}=$numphrases++;
293 if($_ =~ /^ *user: ([^ \t\n]+)/i) {
295 if(!(defined $user)) {
304 # a function that compares the english phrase with the translated one.
305 # compare source strings and desc
307 # Then output the updated version!
309 my ($idstr, $engref, $locref)=@_;
311 my ($esource, $lsource);
314 for my $l (@
$engref) {
319 if($l =~ /^ *desc: (.*)/) {
322 elsif($l =~ / *\<source\>/i) {
326 if($l =~ / *\<\/source\
>/i
) {
337 for my $l (@
$locref) {
338 if($l =~ /^ *desc: (.*)/) {
340 if(trim
($edesc) ne trim
($ldesc)) {
341 $l = "### The 'desc' field differs from the english!\n### the previously used desc is commented below:\n### desc: $ldesc\n desc: $edesc\n";
345 elsif($l =~ / *\<source\>/i) {
350 if($l =~ / *\<\/source\
>/i
) {
353 if(trim
($esource) ne trim
($lsource)) {
354 print "### The <source> section differs from the english!\n",
355 "### the previously used one is commented below:\n";
356 for(split("\n", $lsource)) {
364 undef @show; # start over
381 my @idcount; # counter for lang ID numbers
382 my @voiceid; # counter for voice-only ID numbers
386 push @voiceid, 0x8001;
390 # Now start the scanning of the selected language string
393 open(LANG
, "<$input") || die "Error: couldn't read language file named $input\n";
402 # get rid of DOS newlines
405 if($_ =~ /^( *\#|[ \t\n\r]*\z)/) {
406 # comment or empty line - output it if it's part of the header
407 if ($header and ($update || $sortfile)) {
420 # this is an XML-lookalike tag
421 if (/^(<|[^\"<]+<)([^>]*)>/) {
423 # print "P: $part\n";
426 # this was a closing tag
428 if($part eq "/phrase") {
431 my $idstr = $phrase{'id'};
434 if($binary && !$english{$idstr}) {
435 # $idstr doesn't exist for english, skip it\n";
437 elsif($dest =~ /^none\z/i) {
438 # "none" as dest (without quotes) means that this entire
439 # phrase is to be ignored
442 $allphrases{$idstr}=join('',@phrase);
445 # we don't do the fully detailed analysis when we "update"
446 # since we don't do it for a particular target etc
448 # allow the keyword 'deprecated' to be used on dest and
449 # voice strings to mark that as deprecated. It will then
450 # be replaced with "".
452 $dest =~ s/^deprecate(|d)\z/\"\"/i;
453 $voice =~ s/^deprecate(|d)\z/\"\"/i;
455 # basic syntax error alerts, if there are no quotes we
456 # will assume an empty string was intended
458 print STDERR
"$input:$line:1: warning: dest before line lacks quotes ($dest)!\n";
462 print STDERR
"$input:$line:1: warning: source before line lacks quotes ($src)!\n";
465 if($voice !~ /^\"/ and $voice !~ /^none\z/i) {
466 print STDERR
"$input:$line:1: warning: voice before line lacks quotes ($voice)!\n";
469 if($dest eq '""' && $phrase{'desc'} !~ /deprecated/i && $idstr !~ /^VOICE/) {
470 print STDERR
"$input:$line:1: warning: empty dest before line in non-deprecated phrase!\n";
473 my $userstr = trim
($phrase{'user'});
474 my $user = $users{$userstr};
475 if ($userstr eq "") {
476 print STDERR
"$input:$line:1: warning: missing user!\n";
477 $user = $users{"core"};
479 elsif(!(defined $user)) {
481 print STDERR
"$input:$line:1: warning: user was not found in $english!\n";
482 $user = keys %users; # set to an invalid user so it won't be added
485 # we found a new user, add it to the usermap
487 $users{$userstr} = $user;
491 # Use the ID name to figure out which id number range we
492 # should use for this phrase. Voice-only strings are
495 if($idstr =~ /^VOICE/) {
496 $idnum = $voiceid[$user]++;
499 $idnum = $idcount[$user]++;
502 $id{$idstr} = $idnum;
503 $idnum[$user][$idnum]=$idstr;
505 $source{$idstr}=$src;
507 $voice{$idstr}=$voice;
510 print "id: $phrase{id} ($idnum)\n";
511 print "source: $src\n";
512 print "dest: $dest\n";
513 print "voice: $voice\n";
514 print "user: $user\n";
525 my $e = $english{$idstr};
528 # compare original english with this!
529 my @eng = split("\n", $english{$idstr});
531 compare
($idstr, \
@eng, \
@phrase);
533 $english{$idstr}=""; # clear it
536 print "### $idstr: The phrase is not used. Skipped\n";
541 elsif($part eq "/options") {
542 # closing the options
543 if ($options{'rtl'}) {
544 $langoptions |= $LANGUAGE_FLAG_RTL;
546 } # end of </options>
548 # starts with a slash, this _ends_ this section
549 $m = pop @m; # get back old value, the previous level's tag
553 # This is an opening (sub) tag
555 push @m, $m; # store old value
560 if(/^ *([^:]+): *(.*)/) {
561 my ($name, $val)=($1, $2);
562 &$m($_, $name, $val);
572 "### This phrase below was not present in the translated file\n",
581 for(sort { $sortorder{$a} <=> $sortorder{$b} } keys %allphrases) {
582 print $allphrases{$_};
587 # We create a .c and .h file
589 open(HFILE_CORE
, ">$prefix/lang.h") ||
590 die "Error: couldn't create file $prefix/lang.h\n";
591 open(CFILE_CORE
, ">$prefix/lang_core.c") ||
592 die "Error: couldn't create file $prefix/lang_core.c\n";
594 # get header file name
595 $headername = "$prefix/lang.h";
596 $headername =~ s/(.*\/)*//;
598 print HFILE_CORE
<<MOO
599 /* This file was automatically generated using genlang */
601 * The str() macro/functions is how to access strings that might be
602 * translated. Use it like str(MACRO) and expect a string to be
605 #define str(x) language_strings[x]
607 /* this is the array for holding the string pointers.
608 It will be initialized at runtime. */
609 extern unsigned char *language_strings[];
610 /* this contains the concatenation of all strings, separated by \\0 chars */
611 extern const unsigned char core_language_builtin[];
613 /* The enum below contains all available strings */
618 print CFILE_CORE
<<MOO
619 /* This file was automatically generated using genlang, the strings come
622 #include "$headername"
624 unsigned char *language_strings[LANG_LAST_INDEX_IN_ARRAY];
625 const unsigned char core_language_builtin[] =
629 # Output the ID names for the enum in the header file
631 for $i (0 .. $idcount[$users{"core"}]-1) {
632 my $name=$idnum[$users{"core"}][$i]; # get the ID name
634 $name =~ s/\"//g; # cut off the quotes
636 printf HFILE_CORE
(" %s, /* %d */\n", $name, $i);
639 # Output separation marker for last string ID and the upcoming voice IDs
641 print HFILE_CORE
<<MOO
642 LANG_LAST_INDEX_IN_ARRAY, /* this is not a string, this is a marker */
643 /* --- below this follows voice-only strings --- */
644 VOICEONLY_DELIMITER = 0x8000,
648 # Output the ID names for the enum in the header file
649 for $i (0x8001 .. ($voiceid[$users{"core"}]-1)) {
650 my $name=$idnum[$users{"core"}][$i]; # get the ID name
652 $name =~ s/\"//g; # cut off the quotes
654 printf HFILE_CORE
(" %s, /* 0x%x */\n", $name, $i);
658 print HFILE_CORE
"\n};\n/* end of generated enum list */\n";
660 # Output the target phrases for the source file
661 for $i (0 .. $idcount[$users{"core"}]-1) {
662 my $name=$idnum[$users{"core"}][$i]; # get the ID
663 my $dest = $dest{$name}; # get the destination phrase
665 $dest =~ s
:\"$:\\0\":; # insert a \0 before the second quote
668 # this is just to be on the safe side
672 printf CFILE_CORE
(" %s\n", $dest);
675 # Output end of string chunk
676 print CFILE_CORE
<<MOO
678 /* end of generated string list */
684 } # end of the c/h file generation
686 # Creation of a binary lang file was requested
688 # We must first scan the english file to get the correct order of the id
689 # numbers used there, as that is what sets the id order for all language
690 # files. The english file is scanned before the translated file was
693 open(OUTF
, ">$binary") or die "Error: Can't create $binary";
695 printf OUTF
("%c%c%c%c", $LANGUAGE_COOKIE, $LANGUAGE_VERSION, $target_id,
696 $langoptions); # magic lang file header
698 # output the number of strings for each user
699 my $foffset = $HEADER_SIZE + $SUBHEADER_SIZE * keys(%users);
702 for $n (0 .. $idcount[$_]-1) {
703 $size += length(trim
($dest{$idnum[$_][$n]})) + 1;
705 printf OUTF
("%c%c%c%c%c%c", ($idcount[$_] >> 8), ($idcount[$_] & 0xff),
706 ($size >> 8), ($size & 0xff), ($foffset >> 8), ($foffset & 0xff));
711 # loop over the target phrases
712 for $n (0 .. $idcount[$_]-1) {
713 my $name=$idnum[$_][$n]; # get the ID
714 my $dest = $dest{$name}; # get the destination phrase
717 $dest =~ s/^\"(.*)\"\s*$/$1/g; # cut off quotes
719 # Now, make sure we get the number from the english sort order:
720 $idnum = $idmap[$_]{$name};
722 printf OUTF
("%c%c%s\x00", ($idnum>>8), ($idnum&0xff), $dest);
728 # voice output requested, display id: and voice: strings in a v1-like
733 # This loops over the strings in the translated language file order
734 my @ids = ((0 .. ($idcount[$users{"core"}]-1)));
735 push @ids, (0x8000 .. ($voiceid[$users{"core"}]-1));
742 my $name=$idnum[$users{"core"}][$i]; # get the ID
743 my $dest = $voice{$name}; # get the destination voice string
746 $dest =~ s/^\"(.*)\"\s*$/$1/g; # cut off quotes
748 # Now, make sure we get the number from the english sort order:
749 $idnum = $idmap[$users{"core"}]{$name};
754 #print "Input index $i output index $idnum\n";
757 # not used, mark it so
767 if(($o < 0) || !length($o)) {
768 print "#$i\nid: NOT_USED_$i\nvoice: \"\"\n";
772 my $name=$idnum[$users{"core"}][$o]; # get the ID
773 my $dest = $voice{$name}; # get the destination voice string
775 print "#$i ($o)\nid: $name\nvoice: $dest\n";
785 $num_str += $idcount[$_];
788 printf("%d ID strings scanned\n", $num_str);
792 printf "$_: %s\n", $head{$_};