intltool-merge

   1 #!/usr/bin/perl -w
   2 # -*- Mode: perl; indent-tabs-mode: nil; c-basic-offset: 4  -*-
   3
   4 #
   5 #  The Intltool Message Merger
   6 #
   7 #  Copyright (C) 2000, 2003 Free Software Foundation.
   8 #  Copyright (C) 2000, 2001 Eazel, Inc
   9 #
  10 #  Intltool is free software; you can redistribute it and/or
  11 #  modify it under the terms of the GNU General Public License
  12 #  version 2 published by the Free Software Foundation.
  13 #
  14 #  Intltool is distributed in the hope that it will be useful,
  15 #  but WITHOUT ANY WARRANTY; without even the implied warranty of
  16 #  MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  17 #  General Public License for more details.
  18 #
  19 #  You should have received a copy of the GNU General Public License
  20 #  along with this program; if not, write to the Free Software
  21 #  Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  22 #
  23 #  As a special exception to the GNU General Public License, if you
  24 #  distribute this file as part of a program that contains a
  25 #  configuration script generated by Autoconf, you may include it under
  26 #  the same distribution terms that you use for the rest of that program.
  27 #
  28 #  Authors:  Maciej Stachowiak <mjs@noisehavoc.org>
  29 #            Kenneth Christiansen <kenneth@gnu.org>
  30 #            Darin Adler <darin@bentspoon.com>
  31 #
  32 #  Proper XML UTF-8'ification written by Cyrille Chepelov <chepelov@calixo.net>
  33 #
  34
  35 ## Release information
  36 my $PROGRAM = "intltool-merge";
  37 my $PACKAGE = "intltool";
  38 my $VERSION = "0.35.5";
  39
  40 ## Loaded modules
  41 use strict;
  42 use Getopt::Long;
  43 use Text::Wrap;
  44 use File::Basename;
  45
  46 my $must_end_tag      = -1;
  47 my $last_depth        = -1;
  48 my $translation_depth = -1;
  49 my @tag_stack = ();
  50 my @entered_tag = ();
  51 my @translation_strings = ();
  52 my $leading_space = "";
  53
  54 ## Scalars used by the option stuff
  55 my $HELP_ARG = 0;
  56 my $VERSION_ARG = 0;
  57 my $BA_STYLE_ARG = 0;
  58 my $XML_STYLE_ARG = 0;
  59 my $KEYS_STYLE_ARG = 0;
  60 my $DESKTOP_STYLE_ARG = 0;
  61 my $SCHEMAS_STYLE_ARG = 0;
  62 my $RFC822DEB_STYLE_ARG = 0;
  63 my $QUOTED_STYLE_ARG = 0;
  64 my $QUIET_ARG = 0;
  65 my $PASS_THROUGH_ARG = 0;
  66 my $UTF8_ARG = 0;
  67 my $MULTIPLE_OUTPUT = 0;
  68 my $cache_file;
  69
  70 ## Handle options
  71 GetOptions
  72 (
  73  "help" => \$HELP_ARG,
  74  "version" => \$VERSION_ARG,
  75  "quiet|q" => \$QUIET_ARG,
  76  "oaf-style|o" => \$BA_STYLE_ARG, ## for compatibility
  77  "ba-style|b" => \$BA_STYLE_ARG,
  78  "xml-style|x" => \$XML_STYLE_ARG,
  79  "keys-style|k" => \$KEYS_STYLE_ARG,
  80  "desktop-style|d" => \$DESKTOP_STYLE_ARG,
  81  "schemas-style|s" => \$SCHEMAS_STYLE_ARG,
  82  "rfc822deb-style|r" => \$RFC822DEB_STYLE_ARG,
  83  "quoted-style" => \$QUOTED_STYLE_ARG,
  84  "pass-through|p" => \$PASS_THROUGH_ARG,
  85  "utf8|u" => \$UTF8_ARG,
  86  "multiple-output|m" => \$MULTIPLE_OUTPUT,
  87  "cache|c=s" => \$cache_file
  88  ) or &error;
  89
  90 my $PO_DIR;
  91 my $FILE;
  92 my $OUTFILE;
  93
  94 my %po_files_by_lang = ();
  95 my %translations = ();
  96 my $iconv = $ENV{"ICONV"} || $ENV{"INTLTOOL_ICONV"} || "/usr/bin/iconv";
  97 my $devnull = ($^O eq 'MSWin32' ? 'NUL:' : '/dev/null');
  98
  99 # Use this instead of \w for XML files to handle more possible characters.
 100 my $w = "[-A-Za-z0-9._:]";
 101
 102 # XML quoted string contents
 103 my $q = "[^\\\"]*";
 104
 105 ## Check for options.
 106
 107 if ($VERSION_ARG)
 108 {
 109         &print_version;
 110 }
 111 elsif ($HELP_ARG)
 112 {
 113         &print_help;
 114 }
 115 elsif ($BA_STYLE_ARG && @ARGV > 2)
 116 {
 117         &utf8_sanity_check;
 118         &preparation;
 119         &print_message;
 120         &ba_merge_translations;
 121         &finalize;
 122 }
 123 elsif ($XML_STYLE_ARG && @ARGV > 2)
 124 {
 125         &utf8_sanity_check;
 126         &preparation;
 127         &print_message;
 128         &xml_merge_output;
 129         &finalize;
 130 }
 131 elsif ($KEYS_STYLE_ARG && @ARGV > 2)
 132 {
 133         &utf8_sanity_check;
 134         &preparation;
 135         &print_message;
 136         &keys_merge_translations;
 137         &finalize;
 138 }
 139 elsif ($DESKTOP_STYLE_ARG && @ARGV > 2)
 140 {
 141         &utf8_sanity_check;
 142         &preparation;
 143         &print_message;
 144         &desktop_merge_translations;
 145         &finalize;
 146 }
 147 elsif ($SCHEMAS_STYLE_ARG && @ARGV > 2)
 148 {
 149         &utf8_sanity_check;
 150         &preparation;
 151         &print_message;
 152         &schemas_merge_translations;
 153         &finalize;
 154 }
 155 elsif ($RFC822DEB_STYLE_ARG && @ARGV > 2)
 156 {
 157         &preparation;
 158         &print_message;
 159         &rfc822deb_merge_translations;
 160         &finalize;
 161 }
 162 elsif ($QUOTED_STYLE_ARG && @ARGV > 2)
 163 {
 164         &utf8_sanity_check;
 165         &preparation;
 166         &print_message;
 167         &quoted_merge_translations;
 168         &finalize;
 169 }
 170 else
 171 {
 172         &print_help;
 173 }
 174
 175 exit;
 176
 177 ## Sub for printing release information
 178 sub print_version
 179 {
 180     print <<_EOF_;
 181 ${PROGRAM} (${PACKAGE}) ${VERSION}
 182 Written by Maciej Stachowiak, Darin Adler and Kenneth Christiansen.
 183
 184 Copyright (C) 2000-2003 Free Software Foundation, Inc.
 185 Copyright (C) 2000-2001 Eazel, Inc.
 186 This is free software; see the source for copying conditions.  There is NO
 187 warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
 188 _EOF_
 189     exit;
 190 }
 191
 192 ## Sub for printing usage information
 193 sub print_help
 194 {
 195     print <<_EOF_;
 196 Usage: ${PROGRAM} [OPTION]... PO_DIRECTORY FILENAME OUTPUT_FILE
 197 Generates an output file that includes some localized attributes from an
 198 untranslated source file.
 199
 200 Mandatory options: (exactly one must be specified)
 201   -b, --ba-style         includes translations in the bonobo-activation style
 202   -d, --desktop-style    includes translations in the desktop style
 203   -k, --keys-style       includes translations in the keys style
 204   -s, --schemas-style    includes translations in the schemas style
 205   -r, --rfc822deb-style  includes translations in the RFC822 style
 206       --quoted-style     includes translations in the quoted string style
 207   -x, --xml-style        includes translations in the standard xml style
 208
 209 Other options:
 210   -u, --utf8             convert all strings to UTF-8 before merging
 211                          (default for everything except RFC822 style)
 212   -p, --pass-through     deprecated, does nothing and issues a warning
 213   -m, --multiple-output  output one localized file per locale, instead of
 214                          a single file containing all localized elements
 215   -c, --cache=FILE       specify cache file name
 216                          (usually \$top_builddir/po/.intltool-merge-cache)
 217   -q, --quiet            suppress most messages
 218       --help             display this help and exit
 219       --version          output version information and exit
 220
 221 Report bugs to http://bugzilla.gnome.org/ (product name "$PACKAGE")
 222 or send email to <xml-i18n-tools\@gnome.org>.
 223 _EOF_
 224     exit;
 225 }
 226
 227
 228 ## Sub for printing error messages
 229 sub print_error
 230 {
 231     print STDERR "Try `${PROGRAM} --help' for more information.\n";
 232     exit;
 233 }
 234
 235
 236 sub print_message
 237 {
 238     print "Merging translations into $OUTFILE.\n" unless $QUIET_ARG;
 239 }
 240
 241
 242 sub preparation
 243 {
 244     $PO_DIR = $ARGV[0];
 245     $FILE = $ARGV[1];
 246     $OUTFILE = $ARGV[2];
 247
 248     &gather_po_files;
 249     &get_translation_database;
 250 }
 251
 252 # General-purpose code for looking up translations in .po files
 253
 254 sub po_file2lang
 255 {
 256     my ($tmp) = @_;
 257     $tmp =~ s/^.*\/(.*)\.po$/$1/;
 258     return $tmp;
 259 }
 260
 261 sub gather_po_files
 262 {
 263     for my $po_file (glob "$PO_DIR/*.po") {
 264         $po_files_by_lang{po_file2lang($po_file)} = $po_file;
 265     }
 266 }
 267
 268 sub get_local_charset
 269 {
 270     my ($encoding) = @_;
 271     my $alias_file = $ENV{"G_CHARSET_ALIAS"} || "/usr/local/lib/charset.alias";
 272
 273     # seek character encoding aliases in charset.alias (glib)
 274
 275     if (open CHARSET_ALIAS, $alias_file)
 276     {
 277         while (<CHARSET_ALIAS>)
 278         {
 279             next if /^\#/;
 280             return $1 if (/^\s*([-._a-zA-Z0-9]+)\s+$encoding\b/i)
 281         }
 282
 283         close CHARSET_ALIAS;
 284     }
 285
 286     # if not found, return input string
 287
 288     return $encoding;
 289 }
 290
 291 sub get_po_encoding
 292 {
 293     my ($in_po_file) = @_;
 294     my $encoding = "";
 295
 296     open IN_PO_FILE, $in_po_file or die;
 297     while (<IN_PO_FILE>)
 298     {
 299         ## example: "Content-Type: text/plain; charset=ISO-8859-1\n"
 300         if (/Content-Type\:.*charset=([-a-zA-Z0-9]+)\\n/)
 301         {
 302             $encoding = $1;
 303             last;
 304         }
 305     }
 306     close IN_PO_FILE;
 307
 308     if (!$encoding)
 309     {
 310         print STDERR "Warning: no encoding found in $in_po_file. Assuming ISO-8859-1\n" unless $QUIET_ARG;
 311         $encoding = "ISO-8859-1";
 312     }
 313
 314     system ("$iconv -f $encoding -t UTF-8 <$devnull 2>$devnull");
 315     if ($?) {
 316         $encoding = get_local_charset($encoding);
 317     }
 318
 319     return $encoding
 320 }
 321
 322 sub utf8_sanity_check
 323 {
 324     print STDERR "Warning: option --pass-through has been removed.\n" if $PASS_THROUGH_ARG;
 325     $UTF8_ARG = 1;
 326 }
 327
 328 sub get_translation_database
 329 {
 330     if ($cache_file) {
 331         &get_cached_translation_database;
 332     } else {
 333         &create_translation_database;
 334     }
 335 }
 336
 337 sub get_newest_po_age
 338 {
 339     my $newest_age;
 340
 341     foreach my $file (values %po_files_by_lang)
 342     {
 343         my $file_age = -M $file;
 344         $newest_age = $file_age if !$newest_age || $file_age < $newest_age;
 345     }
 346
 347     $newest_age = 0 if !$newest_age;
 348
 349     return $newest_age;
 350 }
 351
 352 sub create_cache
 353 {
 354     print "Generating and caching the translation database\n" unless $QUIET_ARG;
 355
 356     &create_translation_database;
 357
 358     open CACHE, ">$cache_file" || die;
 359     print CACHE join "\x01", %translations;
 360     close CACHE;
 361 }
 362
 363 sub load_cache
 364 {
 365     print "Found cached translation database\n" unless $QUIET_ARG;
 366
 367     my $contents;
 368     open CACHE, "<$cache_file" || die;
 369     {
 370         local $/;
 371         $contents = <CACHE>;
 372     }
 373     close CACHE;
 374     %translations = split "\x01", $contents;
 375 }
 376
 377 sub get_cached_translation_database
 378 {
 379     my $cache_file_age = -M $cache_file;
 380     if (defined $cache_file_age)
 381     {
 382         if ($cache_file_age <= &get_newest_po_age)
 383         {
 384             &load_cache;
 385             return;
 386         }
 387         print "Found too-old cached translation database\n" unless $QUIET_ARG;
 388     }
 389
 390     &create_cache;
 391 }
 392
 393 sub create_translation_database
 394 {
 395     for my $lang (keys %po_files_by_lang)
 396     {
 397         my $po_file = $po_files_by_lang{$lang};
 398
 399         if ($UTF8_ARG)
 400         {
 401             my $encoding = get_po_encoding ($po_file);
 402
 403             if (lc $encoding eq "utf-8")
 404             {
 405                 open PO_FILE, "<$po_file";
 406             }
 407             else
 408             {
 409                 print "NOTICE: $po_file is not in UTF-8 but $encoding, converting...\n" unless $QUIET_ARG;;
 410
 411                 open PO_FILE, "$iconv -f $encoding -t UTF-8 $po_file|";
 412             }
 413         }
 414         else
 415         {
 416             open PO_FILE, "<$po_file";
 417         }
 418
 419         my $nextfuzzy = 0;
 420         my $inmsgid = 0;
 421         my $inmsgstr = 0;
 422         my $msgid = "";
 423         my $msgstr = "";
 424
 425         while (<PO_FILE>)
 426         {
 427             $nextfuzzy = 1 if /^#, fuzzy/;
 428
 429             if (/^msgid "((\\.|[^\\]+)*)"/ )
 430             {
 431                 $translations{$lang, $msgid} = $msgstr if $inmsgstr && $msgid && $msgstr;
 432                 $msgid = "";
 433                 $msgstr = "";
 434
 435                 if ($nextfuzzy) {
 436                     $inmsgid = 0;
 437                 } else {
 438                     $msgid = unescape_po_string($1);
 439                     $inmsgid = 1;
 440                 }
 441                 $inmsgstr = 0;
 442                 $nextfuzzy = 0;
 443             }
 444
 445             if (/^msgstr "((\\.|[^\\]+)*)"/)
 446             {
 447                 $msgstr = unescape_po_string($1);
 448                 $inmsgstr = 1;
 449                 $inmsgid = 0;
 450             }
 451
 452             if (/^"((\\.|[^\\]+)*)"/)
 453             {
 454                 $msgid .= unescape_po_string($1) if $inmsgid;
 455                 $msgstr .= unescape_po_string($1) if $inmsgstr;
 456             }
 457         }
 458         $translations{$lang, $msgid} = $msgstr if $inmsgstr && $msgid && $msgstr;
 459     }
 460 }
 461
 462 sub finalize
 463 {
 464 }
 465
 466 sub unescape_one_sequence
 467 {
 468     my ($sequence) = @_;
 469
 470     return "\\" if $sequence eq "\\\\";
 471     return "\"" if $sequence eq "\\\"";
 472     return "\n" if $sequence eq "\\n";
 473     return "\r" if $sequence eq "\\r";
 474     return "\t" if $sequence eq "\\t";
 475     return "\b" if $sequence eq "\\b";
 476     return "\f" if $sequence eq "\\f";
 477     return "\a" if $sequence eq "\\a";
 478     return chr(11) if $sequence eq "\\v"; # vertical tab, see ascii(7)
 479
 480     return chr(hex($1)) if ($sequence =~ /\\x([0-9a-fA-F]{2})/);
 481     return chr(oct($1)) if ($sequence =~ /\\([0-7]{3})/);
 482
 483     # FIXME: Is \0 supported as well? Kenneth and Rodney don't want it, see bug #48489
 484
 485     return $sequence;
 486 }
 487
 488 sub unescape_po_string
 489 {
 490     my ($string) = @_;
 491
 492     $string =~ s/(\\x[0-9a-fA-F]{2}|\\[0-7]{3}|\\.)/unescape_one_sequence($1)/eg;
 493
 494     return $string;
 495 }
 496
 497 ## NOTE: deal with < - &lt; but not > - &gt;  because it seems its ok to have
 498 ## > in the entity. For further info please look at #84738.
 499 sub entity_decode
 500 {
 501     local ($_) = @_;
 502
 503     s/&apos;/'/g; # '
 504     s/&quot;/"/g; # "
 505     s/&amp;/&/g;
 506     s/&lt;/</g;
 507
 508     return $_;
 509 }
 510
 511 # entity_encode: (string)
 512 #
 513 # Encode the given string to XML format (encode '<' etc).
 514
 515 sub entity_encode
 516 {
 517     my ($pre_encoded) = @_;
 518
 519     my @list_of_chars = unpack ('C*', $pre_encoded);
 520
 521     # with UTF-8 we only encode minimalistic
 522     return join ('', map (&entity_encode_int_minimalist, @list_of_chars));
 523 }
 524
 525 sub entity_encode_int_minimalist
 526 {
 527     return "&quot;" if $_ == 34;
 528     return "&amp;" if $_ == 38;
 529     return "&apos;" if $_ == 39;
 530     return "&lt;" if $_ == 60;
 531     return chr $_;
 532 }
 533
 534 sub entity_encoded_translation
 535 {
 536     my ($lang, $string) = @_;
 537
 538     my $translation = $translations{$lang, $string};
 539     return $string if !$translation;
 540     return entity_encode ($translation);
 541 }
 542
 543 ## XML (bonobo-activation specific) merge code
 544
 545 sub ba_merge_translations
 546 {
 547     my $source;
 548
 549     {
 550        local $/; # slurp mode
 551        open INPUT, "<$FILE" or die "can't open $FILE: $!";
 552        $source = <INPUT>;
 553        close INPUT;
 554     }
 555
 556     open OUTPUT, ">$OUTFILE" or die "can't open $OUTFILE: $!";
 557     # Binmode so that selftest works ok if using a native Win32 Perl...
 558     binmode (OUTPUT) if $^O eq 'MSWin32';
 559
 560     while ($source =~ s|^(.*?)([ \t]*<\s*$w+\s+($w+\s*=\s*"$q"\s*)+/?>)([ \t]*\n)?||s)
 561     {
 562         print OUTPUT $1;
 563
 564         my $node = $2 . "\n";
 565
 566         my @strings = ();
 567         $_ = $node;
 568         while (s/(\s)_($w+\s*=\s*"($q)")/$1$2/s) {
 569              push @strings, entity_decode($3);
 570         }
 571         print OUTPUT;
 572
 573         my %langs;
 574         for my $string (@strings)
 575         {
 576             for my $lang (keys %po_files_by_lang)
 577             {
 578                 $langs{$lang} = 1 if $translations{$lang, $string};
 579             }
 580         }
 581
 582         for my $lang (sort keys %langs)
 583         {
 584             $_ = $node;
 585             s/(\sname\s*=\s*)"($q)"/$1"$2-$lang"/s;
 586             s/(\s)_($w+\s*=\s*")($q)"/$1 . $2 . entity_encoded_translation($lang, $3) . '"'/seg;
 587             print OUTPUT;
 588         }
 589     }
 590
 591     print OUTPUT $source;
 592
 593     close OUTPUT;
 594 }
 595
 596
 597 ## XML (non-bonobo-activation) merge code
 598
 599
 600 # Process tag attributes
 601 #   Only parameter is a HASH containing attributes -> values mapping
 602 sub getAttributeString
 603 {
 604     my $sub = shift;
 605     my $do_translate = shift || 0;
 606     my $language = shift || "";
 607     my $result = "";
 608     my $translate = shift;
 609     foreach my $e (reverse(sort(keys %{ $sub }))) {
 610         my $key    = $e;
 611         my $string = $sub->{$e};
 612         my $quote = '"';
 613
 614         $string =~ s/^[\s]+//;
 615         $string =~ s/[\s]+$//;
 616
 617         if ($string =~ /^'.*'$/)
 618         {
 619             $quote = "'";
 620         }
 621         $string =~ s/^['"]//g;
 622         $string =~ s/['"]$//g;
 623
 624         if ($do_translate && $key =~ /^_/) {
 625             $key =~ s|^_||g;
 626             if ($language) {
 627                 # Handle translation
 628                 my $decode_string = entity_decode($string);
 629                 my $translation = $translations{$language, $decode_string};
 630                 if ($translation) {
 631                     $translation = entity_encode($translation);
 632                     $string = $translation;
 633                 }
 634                 $$translate = 2;
 635             } else {
 636                  $$translate = 2 if ($translate && (!$$translate)); # watch not to "overwrite" $translate
 637             }
 638         }
 639
 640         $result .= " $key=$quote$string$quote";
 641     }
 642     return $result;
 643 }
 644
 645 # Returns a translatable string from XML node, it works on contents of every node in XML::Parser tree
 646 sub getXMLstring
 647 {
 648     my $ref = shift;
 649     my $spacepreserve = shift || 0;
 650     my @list = @{ $ref };
 651     my $result = "";
 652
 653     my $count = scalar(@list);
 654     my $attrs = $list[0];
 655     my $index = 1;
 656
 657     $spacepreserve = 1 if ((exists $attrs->{"xml:space"}) && ($attrs->{"xml:space"} =~ /^["']?preserve["']?$/));
 658     $spacepreserve = 0 if ((exists $attrs->{"xml:space"}) && ($attrs->{"xml:space"} =~ /^["']?default["']?$/));
 659
 660     while ($index < $count) {
 661         my $type = $list[$index];
 662         my $content = $list[$index+1];
 663         if (! $type ) {
 664             # We've got CDATA
 665             if ($content) {
 666                 # lets strip the whitespace here, and *ONLY* here
 667                 $content =~ s/\s+/ /gs if (!$spacepreserve);
 668                 $result .= $content;
 669             }
 670         } elsif ( "$type" ne "1" ) {
 671             # We've got another element
 672             $result .= "<$type";
 673             $result .= getAttributeString(@{$content}[0], 0); # no nested translatable elements
 674             if ($content) {
 675                 my $subresult = getXMLstring($content, $spacepreserve);
 676                 if ($subresult) {
 677                     $result .= ">".$subresult . "</$type>";
 678                 } else {
 679                     $result .= "/>";
 680                 }
 681             } else {
 682                 $result .= "/>";
 683             }
 684         }
 685         $index += 2;
 686     }
 687     return $result;
 688 }
 689
 690 # Translate list of nodes if necessary
 691 sub translate_subnodes
 692 {
 693     my $fh = shift;
 694     my $content = shift;
 695     my $language = shift || "";
 696     my $singlelang = shift || 0;
 697     my $spacepreserve = shift || 0;
 698
 699     my @nodes = @{ $content };
 700
 701     my $count = scalar(@nodes);
 702     my $index = 0;
 703     while ($index < $count) {
 704         my $type = $nodes[$index];
 705         my $rest = $nodes[$index+1];
 706         if ($singlelang) {
 707             my $oldMO = $MULTIPLE_OUTPUT;
 708             $MULTIPLE_OUTPUT = 1;
 709             traverse($fh, $type, $rest, $language, $spacepreserve);
 710             $MULTIPLE_OUTPUT = $oldMO;
 711         } else {
 712             traverse($fh, $type, $rest, $language, $spacepreserve);
 713         }
 714         $index += 2;
 715     }
 716 }
 717
 718 sub isWellFormedXmlFragment
 719 {
 720     my $ret = eval 'require XML::Parser';
 721     if(!$ret) {
 722         die "You must have XML::Parser installed to run $0\n\n";
 723     }
 724
 725     my $fragment = shift;
 726     return 0 if (!$fragment);
 727
 728     $fragment = "<root>$fragment</root>";
 729     my $xp = new XML::Parser(Style => 'Tree');
 730     my $tree = 0;
 731     eval { $tree = $xp->parse($fragment); };
 732     return $tree;
 733 }
 734
 735 sub traverse
 736 {
 737     my $fh = shift;
 738     my $nodename = shift;
 739     my $content = shift;
 740     my $language = shift || "";
 741     my $spacepreserve = shift || 0;
 742
 743     if (!$nodename) {
 744         if ($content =~ /^[\s]*$/) {
 745             $leading_space .= $content;
 746         }
 747         print $fh $content;
 748     } else {
 749         # element
 750         my @all = @{ $content };
 751         my $attrs = shift @all;
 752         my $translate = 0;
 753         my $outattr = getAttributeString($attrs, 1, $language, \$translate);
 754
 755         if ($nodename =~ /^_/) {
 756             $translate = 1;
 757             $nodename =~ s/^_//;
 758         }
 759         my $lookup = '';
 760
 761         $spacepreserve = 0 if ((exists $attrs->{"xml:space"}) && ($attrs->{"xml:space"} =~ /^["']?default["']?$/));
 762         $spacepreserve = 1 if ((exists $attrs->{"xml:space"}) && ($attrs->{"xml:space"} =~ /^["']?preserve["']?$/));
 763
 764         print $fh "<$nodename", $outattr;
 765         if ($translate) {
 766             $lookup = getXMLstring($content, $spacepreserve);
 767             if (!$spacepreserve) {
 768                 $lookup =~ s/^\s+//s;
 769                 $lookup =~ s/\s+$//s;
 770             }
 771
 772             if ($lookup || $translate == 2) {
 773                 my $translation = $translations{$language, $lookup} if isWellFormedXmlFragment($translations{$language, $lookup});
 774                 if ($MULTIPLE_OUTPUT && ($translation || $translate == 2)) {
 775                     $translation = $lookup if (!$translation);
 776                     print $fh " xml:lang=\"", $language, "\"" if $language;
 777                     print $fh ">";
 778                     if ($translate == 2) {
 779                         translate_subnodes($fh, \@all, $language, 1, $spacepreserve);
 780                     } else {
 781                         print $fh $translation;
 782                     }
 783                     print $fh "</$nodename>";
 784
 785                     return; # this means there will be no same translation with xml:lang="$language"...
 786                             # if we want them both, just remove this "return"
 787                 } else {
 788                     print $fh ">";
 789                     if ($translate == 2) {
 790                         translate_subnodes($fh, \@all, $language, 1, $spacepreserve);
 791                     } else {
 792                         print $fh $lookup;
 793                     }
 794                     print $fh "</$nodename>";
 795                 }
 796             } else {
 797                 print $fh "/>";
 798             }
 799
 800             for my $lang (sort keys %po_files_by_lang) {
 801                     if ($MULTIPLE_OUTPUT && $lang ne "$language") {
 802                         next;
 803                     }
 804                     if ($lang) {
 805                         # Handle translation
 806                         #
 807                         my $translate = 0;
 808                         my $localattrs = getAttributeString($attrs, 1, $lang, \$translate);
 809                         my $translation = $translations{$lang, $lookup} if isWellFormedXmlFragment($translations{$lang, $lookup});
 810                         if ($translate && !$translation) {
 811                             $translation = $lookup;
 812                         }
 813
 814                         if ($translation || $translate) {
 815                             print $fh "\n";
 816                             $leading_space =~ s/.*\n//g;
 817                             print $fh $leading_space;
 818                             print $fh "<", $nodename, " xml:lang=\"", $lang, "\"", $localattrs, ">";
 819                             if ($translate == 2) {
 820                                translate_subnodes($fh, \@all, $lang, 1, $spacepreserve);
 821                             } else {
 822                                 print $fh $translation;
 823                             }
 824                             print $fh "</$nodename>";
 825                         }
 826                     }
 827             }
 828
 829         } else {
 830             my $count = scalar(@all);
 831             if ($count > 0) {
 832                 print $fh ">";
 833                 my $index = 0;
 834                 while ($index < $count) {
 835                     my $type = $all[$index];
 836                     my $rest = $all[$index+1];
 837                     traverse($fh, $type, $rest, $language, $spacepreserve);
 838                     $index += 2;
 839                 }
 840                 print $fh "</$nodename>";
 841             } else {
 842                 print $fh "/>";
 843             }
 844         }
 845     }
 846 }
 847
 848 sub intltool_tree_comment
 849 {
 850     my $expat = shift;
 851     my $data  = shift;
 852     my $clist = $expat->{Curlist};
 853     my $pos   = $#$clist;
 854
 855     push @$clist, 1 => $data;
 856 }
 857
 858 sub intltool_tree_cdatastart
 859 {
 860     my $expat    = shift;
 861     my $clist = $expat->{Curlist};
 862     my $pos   = $#$clist;
 863
 864     push @$clist, 0 => $expat->original_string();
 865 }
 866
 867 sub intltool_tree_cdataend
 868 {
 869     my $expat    = shift;
 870     my $clist = $expat->{Curlist};
 871     my $pos   = $#$clist;
 872
 873     $clist->[$pos] .= $expat->original_string();
 874 }
 875
 876 sub intltool_tree_char
 877 {
 878     my $expat = shift;
 879     my $text  = shift;
 880     my $clist = $expat->{Curlist};
 881     my $pos   = $#$clist;
 882
 883     # Use original_string so that we retain escaped entities
 884     # in CDATA sections.
 885     #
 886     if ($pos > 0 and $clist->[$pos - 1] eq '0') {
 887         $clist->[$pos] .= $expat->original_string();
 888     } else {
 889         push @$clist, 0 => $expat->original_string();
 890     }
 891 }
 892
 893 sub intltool_tree_start
 894 {
 895     my $expat    = shift;
 896     my $tag      = shift;
 897     my @origlist = ();
 898
 899     # Use original_string so that we retain escaped entities
 900     # in attribute values.  We must convert the string to an
 901     # @origlist array to conform to the structure of the Tree
 902     # Style.
 903     #
 904     my @original_array = split /\x/, $expat->original_string();
 905     my $source         = $expat->original_string();
 906
 907     # Remove leading tag.
 908     #
 909     $source =~ s|^\s*<\s*(\S+)||s;
 910
 911     # Grab attribute key/value pairs and push onto @origlist array.
 912     #
 913     while ($source)
 914     {
 915        if ($source =~ /^\s*([\w:-]+)\s*[=]\s*["]/)
 916        {
 917            $source =~ s|^\s*([\w:-]+)\s*[=]\s*["]([^"]*)["]||s;
 918            push @origlist, $1;
 919            push @origlist, '"' . $2 . '"';
 920        }
 921        elsif ($source =~ /^\s*([\w:-]+)\s*[=]\s*[']/)
 922        {
 923            $source =~ s|^\s*([\w:-]+)\s*[=]\s*[']([^']*)[']||s;
 924            push @origlist, $1;
 925            push @origlist, "'" . $2 . "'";
 926        }
 927        else
 928        {
 929            last;
 930        }
 931     }
 932
 933     my $ol = [ { @origlist } ];
 934
 935     push @{ $expat->{Lists} }, $expat->{Curlist};
 936     push @{ $expat->{Curlist} }, $tag => $ol;
 937     $expat->{Curlist} = $ol;
 938 }
 939
 940 sub readXml
 941 {
 942     my $filename = shift || return;
 943     if(!-f $filename) {
 944         die "ERROR Cannot find filename: $filename\n";
 945     }
 946
 947     my $ret = eval 'require XML::Parser';
 948     if(!$ret) {
 949         die "You must have XML::Parser installed to run $0\n\n";
 950     }
 951     my $xp = new XML::Parser(Style => 'Tree');
 952     $xp->setHandlers(Char => \&intltool_tree_char);
 953     $xp->setHandlers(Start => \&intltool_tree_start);
 954     $xp->setHandlers(CdataStart => \&intltool_tree_cdatastart);
 955     $xp->setHandlers(CdataEnd => \&intltool_tree_cdataend);
 956     my $tree = $xp->parsefile($filename);
 957
 958 # <foo><head id="a">Hello <em>there</em></head><bar>Howdy<ref/></bar>do</foo>
 959 # would be:
 960 # [foo, [{}, head, [{id => "a"}, 0, "Hello ",  em, [{}, 0, "there"]], bar, [{},
 961 # 0, "Howdy",  ref, [{}]], 0, "do" ] ]
 962
 963     return $tree;
 964 }
 965
 966 sub print_header
 967 {
 968     my $infile = shift;
 969     my $fh = shift;
 970     my $source;
 971
 972     if(!-f $infile) {
 973         die "ERROR Cannot find filename: $infile\n";
 974     }
 975
 976     print $fh qq{<?xml version="1.0" encoding="UTF-8"?>\n};
 977     {
 978         local $/;
 979         open DOCINPUT, "<${FILE}" or die;
 980         $source = <DOCINPUT>;
 981         close DOCINPUT;
 982     }
 983     if ($source =~ /(<!DOCTYPE.*\[.*\]\s*>)/s)
 984     {
 985         print $fh "$1\n";
 986     }
 987     elsif ($source =~ /(<!DOCTYPE[^>]*>)/s)
 988     {
 989         print $fh "$1\n";
 990     }
 991 }
 992
 993 sub parseTree
 994 {
 995     my $fh        = shift;
 996     my $ref       = shift;
 997     my $language  = shift || "";
 998
 999     my $name = shift @{ $ref };
1000     my $cont = shift @{ $ref };
1001
1002     while (!$name || "$name" eq "1") {
1003         $name = shift @{ $ref };
1004         $cont = shift @{ $ref };
1005     }
1006
1007     my $spacepreserve = 0;
1008     my $attrs = @{$cont}[0];
1009     $spacepreserve = 1 if ((exists $attrs->{"xml:space"}) && ($attrs->{"xml:space"} =~ /^["']?preserve["']?$/));
1010
1011     traverse($fh, $name, $cont, $language, $spacepreserve);
1012 }
1013
1014 sub xml_merge_output
1015 {
1016     my $source;
1017
1018     if ($MULTIPLE_OUTPUT) {
1019         for my $lang (sort keys %po_files_by_lang) {
1020             if ( ! -d $lang ) {
1021                 mkdir $lang or -d $lang or die "Cannot create subdirectory $lang: $!\n";
1022             }
1023             open OUTPUT, ">$lang/$OUTFILE" or die "Cannot open $lang/$OUTFILE: $!\n";
1024             binmode (OUTPUT) if $^O eq 'MSWin32';
1025             my $tree = readXml($FILE);
1026             print_header($FILE, \*OUTPUT);
1027             parseTree(\*OUTPUT, $tree, $lang);
1028             close OUTPUT;
1029             print "CREATED $lang/$OUTFILE\n" unless $QUIET_ARG;
1030         }
1031     }
1032     open OUTPUT, ">$OUTFILE" or die "Cannot open $OUTFILE: $!\n";
1033     binmode (OUTPUT) if $^O eq 'MSWin32';
1034     my $tree = readXml($FILE);
1035     print_header($FILE, \*OUTPUT);
1036     parseTree(\*OUTPUT, $tree);
1037     close OUTPUT;
1038     print "CREATED $OUTFILE\n" unless $QUIET_ARG;
1039 }
1040
1041 sub keys_merge_translations
1042 {
1043     open INPUT, "<${FILE}" or die;
1044     open OUTPUT, ">${OUTFILE}" or die;
1045     binmode (OUTPUT) if $^O eq 'MSWin32';
1046
1047     while (<INPUT>)
1048     {
1049         if (s/^(\s*)_(\w+=(.*))/$1$2/)
1050         {
1051             my $string = $3;
1052
1053             print OUTPUT;
1054
1055             my $non_translated_line = $_;
1056
1057             for my $lang (sort keys %po_files_by_lang)
1058             {
1059                 my $translation = $translations{$lang, $string};
1060                 next if !$translation;
1061
1062                 $_ = $non_translated_line;
1063                 s/(\w+)=.*/[$lang]$1=$translation/;
1064                 print OUTPUT;
1065             }
1066         }
1067         else
1068         {
1069             print OUTPUT;
1070         }
1071     }
1072
1073     close OUTPUT;
1074     close INPUT;
1075 }
1076
1077 sub desktop_merge_translations
1078 {
1079     open INPUT, "<${FILE}" or die;
1080     open OUTPUT, ">${OUTFILE}" or die;
1081     binmode (OUTPUT) if $^O eq 'MSWin32';
1082
1083     while (<INPUT>)
1084     {
1085         if (s/^(\s*)_(\w+=(.*))/$1$2/)
1086         {
1087             my $string = $3;
1088
1089             print OUTPUT;
1090
1091             my $non_translated_line = $_;
1092
1093             for my $lang (sort keys %po_files_by_lang)
1094             {
1095                 my $translation = $translations{$lang, $string};
1096                 next if !$translation;
1097
1098                 $_ = $non_translated_line;
1099                 s/(\w+)=.*/${1}[$lang]=$translation/;
1100                 print OUTPUT;
1101             }
1102         }
1103         else
1104         {
1105             print OUTPUT;
1106         }
1107     }
1108
1109     close OUTPUT;
1110     close INPUT;
1111 }
1112
1113 sub schemas_merge_translations
1114 {
1115     my $source;
1116
1117     {
1118        local $/; # slurp mode
1119        open INPUT, "<$FILE" or die "can't open $FILE: $!";
1120        $source = <INPUT>;
1121        close INPUT;
1122     }
1123
1124     open OUTPUT, ">$OUTFILE" or die;
1125     binmode (OUTPUT) if $^O eq 'MSWin32';
1126
1127     # FIXME: support attribute translations
1128
1129     # Empty nodes never need translation, so unmark all of them.
1130     # For example, <_foo/> is just replaced by <foo/>.
1131     $source =~ s|<\s*_($w+)\s*/>|<$1/>|g;
1132
1133     while ($source =~ s/
1134                         (.*?)
1135                         (\s+)(<locale\ name="C">(\s*)
1136                             (<default>\s*(?:<!--[^>]*?-->\s*)?(.*?)\s*<\/default>)?(\s*)
1137                             (<short>\s*(?:<!--[^>]*?-->\s*)?(.*?)\s*<\/short>)?(\s*)
1138                             (<long>\s*(?:<!--[^>]*?-->\s*)?(.*?)\s*<\/long>)?(\s*)
1139                         <\/locale>)
1140                        //sx)
1141     {
1142         print OUTPUT $1;
1143
1144         my $locale_start_spaces = $2 ? $2 : '';
1145         my $default_spaces = $4 ? $4 : '';
1146         my $short_spaces = $7 ? $7 : '';
1147         my $long_spaces = $10 ? $10 : '';
1148         my $locale_end_spaces = $13 ? $13 : '';
1149         my $c_default_block = $3 ? $3 : '';
1150         my $default_string = $6 ? $6 : '';
1151         my $short_string = $9 ? $9 : '';
1152         my $long_string = $12 ? $12 : '';
1153
1154         print OUTPUT "$locale_start_spaces$c_default_block";
1155
1156         $default_string =~ s/\s+/ /g;
1157         $default_string = entity_decode($default_string);
1158         $short_string =~ s/\s+/ /g;
1159         $short_string = entity_decode($short_string);
1160         $long_string =~ s/\s+/ /g;
1161         $long_string = entity_decode($long_string);
1162
1163         for my $lang (sort keys %po_files_by_lang)
1164         {
1165             my $default_translation = $translations{$lang, $default_string};
1166             my $short_translation = $translations{$lang, $short_string};
1167             my $long_translation  = $translations{$lang, $long_string};
1168
1169             next if (!$default_translation && !$short_translation &&
1170                      !$long_translation);
1171
1172             print OUTPUT "\n$locale_start_spaces<locale name=\"$lang\">";
1173
1174         print OUTPUT "$default_spaces";
1175
1176         if ($default_translation)
1177         {
1178             $default_translation = entity_encode($default_translation);
1179             print OUTPUT "<default>$default_translation</default>";
1180         }
1181
1182             print OUTPUT "$short_spaces";
1183
1184             if ($short_translation)
1185             {
1186                         $short_translation = entity_encode($short_translation);
1187                         print OUTPUT "<short>$short_translation</short>";
1188             }
1189
1190             print OUTPUT "$long_spaces";
1191
1192             if ($long_translation)
1193             {
1194                         $long_translation = entity_encode($long_translation);
1195                         print OUTPUT "<long>$long_translation</long>";
1196             }
1197
1198             print OUTPUT "$locale_end_spaces</locale>";
1199         }
1200     }
1201
1202     print OUTPUT $source;
1203
1204     close OUTPUT;
1205 }
1206
1207 sub rfc822deb_merge_translations
1208 {
1209     my %encodings = ();
1210     for my $lang (keys %po_files_by_lang) {
1211         $encodings{$lang} = ($UTF8_ARG ? 'UTF-8' : get_po_encoding($po_files_by_lang{$lang}));
1212     }
1213
1214     my $source;
1215
1216     $Text::Wrap::huge = 'overflow';
1217     $Text::Wrap::break = qr/\n|\s(?=\S)/;
1218
1219     {
1220        local $/; # slurp mode
1221        open INPUT, "<$FILE" or die "can't open $FILE: $!";
1222        $source = <INPUT>;
1223        close INPUT;
1224     }
1225
1226     open OUTPUT, ">${OUTFILE}" or die;
1227     binmode (OUTPUT) if $^O eq 'MSWin32';
1228
1229     while ($source =~ /(^|\n+)(_*)([^:\s]+)(:[ \t]*)(.*?)(?=\n[\S\n]|$)/sg)
1230     {
1231             my $sep = $1;
1232             my $non_translated_line = $3.$4;
1233             my $string = $5;
1234             my $underscore = length($2);
1235             next if $underscore eq 0 && $non_translated_line =~ /^#/;
1236             #  Remove [] dummy strings
1237             my $stripped = $string;
1238             $stripped =~ s/\[\s[^\[\]]*\],/,/g if $underscore eq 2;
1239             $stripped =~ s/\[\s[^\[\]]*\]$//;
1240             $non_translated_line .= $stripped;
1241
1242             print OUTPUT $sep.$non_translated_line;
1243
1244             if ($underscore)
1245             {
1246                 my @str_list = rfc822deb_split($underscore, $string);
1247
1248                 for my $lang (sort keys %po_files_by_lang)
1249                 {
1250                     my $is_translated = 1;
1251                     my $str_translated = '';
1252                     my $first = 1;
1253
1254                     for my $str (@str_list)
1255                     {
1256                         my $translation = $translations{$lang, $str};
1257
1258                         if (!$translation)
1259                         {
1260                             $is_translated = 0;
1261                             last;
1262                         }
1263
1264                         #  $translation may also contain [] dummy
1265                         #  strings, mostly to indicate an empty string
1266                         $translation =~ s/\[\s[^\[\]]*\]$//;
1267
1268                         if ($first)
1269                         {
1270                             if ($underscore eq 2)
1271                             {
1272                                 $str_translated .= $translation;
1273                             }
1274                             else
1275                             {
1276                                 $str_translated .=
1277                                     Text::Tabs::expand($translation) .
1278                                     "\n";
1279                             }
1280                         }
1281                         else
1282                         {
1283                             if ($underscore eq 2)
1284                             {
1285                                 $str_translated .= ', ' . $translation;
1286                             }
1287                             else
1288                             {
1289                                 $str_translated .= Text::Tabs::expand(
1290                                     Text::Wrap::wrap(' ', ' ', $translation)) .
1291                                     "\n .\n";
1292                             }
1293                         }
1294                         $first = 0;
1295
1296                         #  To fix some problems with Text::Wrap::wrap
1297                         $str_translated =~ s/(\n )+\n/\n .\n/g;
1298                     }
1299                     next unless $is_translated;
1300
1301                     $str_translated =~ s/\n \.\n$//;
1302                     $str_translated =~ s/\s+$//;
1303
1304                     $_ = $non_translated_line;
1305                     s/^(\w+):\s*.*/$sep${1}-$lang.$encodings{$lang}: $str_translated/s;
1306                     print OUTPUT;
1307                 }
1308             }
1309     }
1310     print OUTPUT "\n";
1311
1312     close OUTPUT;
1313     close INPUT;
1314 }
1315
1316 sub rfc822deb_split
1317 {
1318     # Debian defines a special way to deal with rfc822-style files:
1319     # when a value contain newlines, it consists of
1320     #   1.  a short form (first line)
1321     #   2.  a long description, all lines begin with a space,
1322     #       and paragraphs are separated by a single dot on a line
1323     # This routine returns an array of all paragraphs, and reformat
1324     # them.
1325     # When first argument is 2, the string is a comma separated list of
1326     # values.
1327     my $type = shift;
1328     my $text = shift;
1329     $text =~ s/^[ \t]//mg;
1330     return (split(/, */, $text, 0)) if $type ne 1;
1331     return ($text) if $text !~ /\n/;
1332
1333     $text =~ s/([^\n]*)\n//;
1334     my @list = ($1);
1335     my $str = '';
1336
1337     for my $line (split (/\n/, $text))
1338     {
1339         chomp $line;
1340         if ($line =~ /^\.\s*$/)
1341         {
1342             #  New paragraph
1343             $str =~ s/\s*$//;
1344             push(@list, $str);
1345             $str = '';
1346         }
1347         elsif ($line =~ /^\s/)
1348         {
1349             #  Line which must not be reformatted
1350             $str .= "\n" if length ($str) && $str !~ /\n$/;
1351             $line =~ s/\s+$//;
1352             $str .= $line."\n";
1353         }
1354         else
1355         {
1356             #  Continuation line, remove newline
1357             $str .= " " if length ($str) && $str !~ /\n$/;
1358             $str .= $line;
1359         }
1360     }
1361
1362     $str =~ s/\s*$//;
1363     push(@list, $str) if length ($str);
1364
1365     return @list;
1366 }
1367
1368 sub quoted_translation
1369 {
1370     my ($lang, $string) = @_;
1371
1372     $string =~ s/\\\"/\"/g;
1373
1374     my $translation = $translations{$lang, $string};
1375     $translation = $string if !$translation;
1376
1377     $translation =~ s/\"/\\\"/g;
1378     return $translation
1379 }
1380
1381 sub quoted_merge_translations
1382 {
1383     if (!$MULTIPLE_OUTPUT) {
1384         print "Quoted only supports Multiple Output.\n";
1385         exit(1);
1386     }
1387
1388     for my $lang (sort keys %po_files_by_lang) {
1389         if ( ! -d $lang ) {
1390             mkdir $lang or -d $lang or die "Cannot create subdirectory $lang: $!\n";
1391         }
1392         open INPUT, "<${FILE}" or die;
1393         open OUTPUT, ">$lang/$OUTFILE" or die "Cannot open $lang/$OUTFILE: $!\n";
1394         binmode (OUTPUT) if $^O eq 'MSWin32';
1395         while (<INPUT>)
1396         {
1397             s/\"(([^\"]|\\\")*[^\\\"])\"/"\"" . &quoted_translation($lang, $1) . "\""/ge;
1398             print OUTPUT;
1399         }
1400         close OUTPUT;
1401         close INPUT;
1402     }
1403 }