inCHISE

   1 #!/usr/bin/perl -w -CSD
   2
   3 # ver.0.2
   4
   5 use strict;
   6 use vars qw($omegadb_path
   7             $opt_protrude $opt_allow_unify
   8             $opt_use_kage_for_Ext_B
   9             %opt_order %order %order_map
  10             $opt_in_cs $opt_out_cs
  11             $opt_help $usage
  12             $in_cs $out_cs $i @chars
  13             $char $char_id $out_char
  14             $char_unified @chars_unified
  15             $ids $ids_argc %ids $idsdb $geta
  16             $idsdata_file $ids_start $font_start
  17             @CDP @HZK @GT
  18             );
  19 use Getopt::Long;
  20 use utf8;
  21 use Fcntl ':flock';
  22 use Chise_utils ':all';
  23 require 5.008;
  24
  25 ### Options ###
  26
  27 #$opt_order{'UniMulti'}='jcgk';
  28 $opt_order{'UniMulti'}='jGcgkHC';
  29 $opt_order{'UniCNS'}='c';
  30 $opt_order{'UniGB'}='g';
  31 $opt_order{'UniJIS'}='j';
  32 $opt_order{'UniKS'}='k';
  33
  34 $opt_allow_unify=1; # 1=true, 0=false.
  35 $opt_protrude=0;# 1=true, 0=false.
  36
  37 # currently does not work.
  38 $opt_use_kage_for_Ext_B=0;# 1=true, 0=false.
  39
  40 ### End ###
  41
  42 my $strictly_forbidden_after = '「【『［（〈“‘‘（〔｛《{\[\(\x{3016}｛｢';
  43 #       \x{3016} | # white 【
  44
  45 my $forbidden_after = "\x{0000}";
  46
  47 # ￥¥＄$〒♯＃#￠¢￡£＠@§
  48 my $slightly_forbidden_after = '￥¥＄$〒♯＃#￠¢￡£＠@§';
  49
  50 # $strictly_forbidden_before
  51 # All these characters are allowed to protrude
  52 # in the right margin
  53 my $strictly_forbidden_before=
  54     '!,.:;?、。！，．：；？。\)#}’”〉》」』】〕\x{3017}）］｝｝」\]';
  55 ###       \x{3017} | # white 】
  56
  57 my $forbidden_before
  58     = 'ー々ぁぃぅぇぉゃゅょっゎァィゥェォャュョッヮヵヶ';
  59
  60 my $slightly_forbidden_before
  61     = '\x{000a}\#\-‐−‰′″℃゛゜ゝゞヽヾ＂％－ﾞﾟ';
  62
  63 my $asian = '\x{1100}-\x{11FF}\x{2E80}-\x{D7AF}\x{F900}-\x{FAFF}\x{FE30}-\x{FE4F}\x{FF00}-\x{FFFFFF}'; # need to fix.
  64
  65 my $space = '\x{0020}\x{0009}\x{000A}\x{000C}\x{000D}';
  66
  67 my %tex_meta=('#'=>'\#',
  68               '$'=>'\\textdollar{}',
  69               '%'=>'\%',
  70               '&'=>'\&',
  71               '{'=>'\\textbraceleft{}',
  72               '}'=>'\\textbraceright{}',
  73               '\\'=>'\\textbackslash{}',
  74               '_'=>'\\textunderscore',
  75              );
  76
  77 my $tex_meta_re=join('|',map {quotemeta($_)} keys %tex_meta);
  78
  79 &GetOptions("in=s"=>\$opt_in_cs,
  80             "out=s"=>\$opt_out_cs,
  81             "help",\$opt_help);
  82
  83 $usage=<<EOF;
  84 Usage: $0 -i <input coding system> -o <cmap encoding>
  85     input coding system:
  86       Utf8mcs, Utf8cns, Utf8gb, Utf8jis, Utf8ks, Utf8big5
  87     cmap encoding:
  88       UniCNS, UniGB, UniJIS, UniKS, UniMulti
  89 EOF
  90
  91 if($opt_in_cs or $opt_out_cs){
  92     $in_cs=$opt_in_cs;
  93     $out_cs=$opt_out_cs;
  94 }elsif(@ARGV==0){
  95     ($in_cs,$out_cs)=($0=~/(Utf8.+)To(\w+)/);
  96 }
  97
  98 # $in_cs:
  99 #   Utf8mcs,Utf8cns,Utf8gb,Utf8jis,Utf8ks,Utf8big5
 100 # $out_cs:
 101 #   UniCNS,UniGB,UniJIS,UniKS,UniMulti
 102
 103 $in_cs=~s/Utf8/ucs\@/;
 104
 105 if($opt_help
 106    or not defined($in_cs)
 107    or not defined($out_cs)){
 108     print $usage;
 109     exit 1;
 110 }
 111
 112 $omegadb_path=~s!/$!!;
 113
 114 $idsdata_file="$omegadb_path/idsdata.pl";
 115 $ids_start=0x00;
 116 $font_start=0;
 117
 118 if(-e $idsdata_file){
 119     open(IDSDATA,"+<:utf8",$idsdata_file) or die;
 120     flock(IDSDATA,LOCK_EX);
 121     seek(IDSDATA,0,0);
 122     while(<IDSDATA>){
 123         eval $_;
 124     }
 125     seek(IDSDATA,0,0);
 126 #         require $idsdata_file;
 127 }else{
 128     open(IDSDATA,">:utf8",$idsdata_file) or die;
 129     flock(IDSDATA,LOCK_EX);
 130     seek(IDSDATA,0,0);
 131 }
 132
 133 $ids_argc=0;
 134 $ids="";
 135
 136 #$geta=pack("U",0x3013);
 137 $geta=pack("U",0xfffd);
 138
 139 @GT=(#"=gt","=gt-k",
 140      "=gt-pj-1","=gt-pj-2","=gt-pj-3","=gt-pj-4","=gt-pj-5",
 141      "=gt-pj-6","=gt-pj-7","=gt-pj-8","=gt-pj-9","=gt-pj-10",
 142      "=gt-pj-11"
 143      #,"=gt-pj-k1","=gt-pj-k2"
 144      );
 145 @HZK=("=hanziku-1","=hanziku-2","=hanziku-3","=hanziku-4",
 146       "=hanziku-5","=hanziku-6","=hanziku-7","=hanziku-8",
 147       "=hanziku-9","=hanziku-10","=hanziku-11","=hanziku-12");
 148 @CDP=("=big5-cdp");
 149
 150 %order_map=('c'=>'UniCNS',
 151             'g'=>'UniGB',
 152             'j'=>'UniJIS',
 153             'k'=>'UniKS',
 154             'G'=>'GT',
 155             'H'=>'HZK',
 156             'C'=>'CDP',
 157             );
 158
 159 foreach $out_cs ('UniCNS','UniGB','UniJIS','UniKS','UniMulti'){
 160     if(defined($opt_order{$out_cs})){
 161         if($opt_order{$out_cs}=~/^[cgjkGHC]+$/){
 162             @{$order{$out_cs}}=map {$order_map{$_}}
 163             (split(//,$opt_order{$out_cs}));
 164         }else{
 165             print STDERR "Invalid order for $out_cs!\n";
 166             exit 1;
 167         }
 168     }
 169 }
 170
 171 while(<>){
 172     utf8::decode($_);
 173
 174     s/([$asian])\s+([$asian])/$1$2/g unless($out_cs eq 'UniKS');
 175     s/([$asian])\s*([^$asian])/$1 $2/g;
 176     s/([^$asian])\s*([$asian])/$1 $2/g;
 177     if($in_cs ne 'ucs@mcs'){
 178         s/(.)/&get_char_in_mcs($1,$in_cs)/ge;
 179     }
 180     s/(amp.+?;)/&de_tex_er($1)/ge;
 181 #    s/(&.+?;)/&de_tex_er($1)/ge;
 182     @chars=split(//);
 183   CHAR:
 184     for($i=0;$i<=$#chars;$i++){
 185         $char=$chars[$i];
 186         $char_id=unpack("U",$char);
 187
 188         if($char_id<=0x20){
 189             # add quarter space between asian and non-asian.
 190             if($i>0 and $i<$#chars){
 191                 if((($chars[$i-1]=~/[$asian]/)
 192                     and($chars[$i+1]=~/[^$asian]/))
 193                    or(($chars[$i-1]=~/[^$asian]/)
 194                       and($chars[$i+1]=~/[$asian]/))){
 195                     print '{\selectjisfont\hspace{.25ex}}';
 196                 }else{
 197                     print " ";
 198                 }
 199             }
 200             next CHAR;
 201         }elsif($char=~m/($tex_meta_re)/o){
 202             print $tex_meta{$1};
 203             next CHAR;
 204         }elsif(($char_id>0x20 and $char_id<=0x021f)
 205                # Basic Latin
 206                # Latin-1 Supplement
 207                # Latin Extended-A
 208                # Latin Extended-B (not all)
 209                or($char_id>=0x0250 and $char_id<=0x02af)
 210                # IPA Extensions
 211                or($char_id>=0x0300 and $char_id<=0x033f)
 212                or($char_id>=0x0360 and $char_id<=0x036f)
 213                # Combining Diacritical Marks
 214                or($char_id>=0x1e00 and $char_id<=0x1eff)
 215                # Latin Extended Additional
 216                or($char_id>=0x0370 and $char_id<=0x03ff)
 217                # Greek and Coptic
 218                or($char_id>=0x0400 and $char_id<=0x04ff)
 219                # Cyrillic
 220                or($char_id>=0x0530 and $char_id<=0x058f)
 221                # Armenian
 222                ){
 223             print &latin_parse();
 224             next CHAR;
 225         }elsif($char_id>=0x2ff0 and $char_id<=0x2fff){
 226             # Ideographic Description Characters
 227             print &ids_parse();
 228             next CHAR;
 229         }else{
 230             if(($out_char=&get_output_char($char,$out_cs))){
 231                 print $out_char,&add_break($i);
 232             }else{
 233                 if($opt_allow_unify){
 234                     @chars_unified=&get_chars_unified($char);
 235                     if(@chars_unified>0){
 236                         foreach $char_unified (@chars_unified){
 237                             if(($out_char
 238                                 =&get_output_char($char_unified,$out_cs))){
 239                                 print $out_char,&add_break($i);
 240                                 next CHAR;
 241                             }
 242                         }
 243                     }
 244                 }
 245                 if($opt_use_kage_for_Ext_B){
 246                     if($char_id >= 0x20000 && $char_id <=0x2a6df){
 247                         # CJK Unified Ideographs Extension B
 248                         if(not defined($ids{$char}) and $ids{$char}[1]>=0){
 249                             $ids{$char}[0]=$font_start;
 250                             $ids{$char}[1]=$ids_start;
 251                             $ids_start++;
 252                             if($ids_start>255){
 253                                 $ids_start=0;
 254                                 $font_start++;
 255                             }
 256                         }
 257                         print "{\\fontencoding{OT1}\\fontfamily{" .
 258                             sprintf("chise%03d",$ids{$char}[0]) .
 259                                 "}\\selectfont\\char$ids{$char}[1]}",&add_break($i);
 260                         next CHAR;
 261                     }
 262                 }
 263                 if($ids=&get_ids($char)){
 264                     print &get_macro_for_ids($ids),&add_break($i);
 265                 }else{
 266                     print '\rule{1ex}{1ex}',&add_break($i);
 267                 }
 268             }
 269         }
 270     }
 271 }
 272
 273 print IDSDATA 'use utf8;',"\n";
 274 foreach $ids (keys %ids){
 275     print IDSDATA '$ids{\'',$ids,'\'}='
 276     ,'[',join ",",@{$ids{$ids}},"];\n";
 277 }
 278 print IDSDATA '$font_start=',$font_start,";\n";
 279 print IDSDATA '$ids_start=',$ids_start,";\n";
 280 print IDSDATA "1;";
 281 flock(IDSDATA,LOCK_UN);
 282
 283 exit 0;
 284
 285 sub de_tex_er{
 286     my($er)=@_;
 287     my($prefix,$suffix);
 288     my($output_char,$atr,$value);
 289     $er=~/^(amp)(.*)(;)$/
 290         and $prefix=$1,$er=$2,$suffix=$3;
 291     $prefix or $prefix="",$suffix or $suffix="";
 292     if($er=~/^U[\+|\-]([a-fA-F\d]+)/){
 293         $output_char=pack("U",hex($1));
 294     }elsif($er=~/^(?:I\-)?($er_prefix_re)\-?([0-9a-fA-F]+)$/){
 295         ($atr,$value)=($1,$2);
 296         unless($er_alias{$atr}=~/daikanwa|gt/){
 297             $value=hex($value);
 298         }
 299         ($output_char)=&get_chars_matching($er_alias{$atr},$value);
 300     }
 301     if($output_char){
 302         return $output_char;
 303     }else{
 304         return $prefix.$er.$suffix;
 305     }
 306 }
 307
 308 sub add_break{
 309     my($i)=@_;
 310
 311     if($i<($#chars-1)){
 312         if(($chars[$i+1]=~m/[$strictly_forbidden_before]/o)
 313            and($chars[$i+2]=~m/[$strictly_forbidden_before]/o)){
 314             return "\\CJKunbreakablekernone ";
 315         }elsif($opt_protrude){
 316             if(($chars[$i+1]=~m/[$strictly_forbidden_before]/o)
 317                and($chars[$i+2]=~m/[^$strictly_forbidden_before]/o)){
 318                 return "\\CJKunbreakablekernone \\CJKprotrude ";
 319             }
 320         }
 321     }
 322     if(($i<$#chars)
 323        and($chars[$i+1]eq" ")){
 324        # preserve space.
 325         return "";
 326     }
 327     if(($i<$#chars)
 328        and($chars[$i+1]=~m/[$strictly_forbidden_before]/o)){
 329         return "\\CJKunbreakablekernone ";
 330     }
 331     if($chars[$i]=~m/[$strictly_forbidden_after]/o){
 332         return "\\CJKunbreakablekernone ";
 333     }
 334     if(($i<$#chars)
 335        and($chars[$i+1]=~m/[$forbidden_before]/o)){
 336         return "\\CJKunbreakablekerntwo ";
 337
 338     }
 339     if($chars[$i]=~m/[$forbidden_after]/o){
 340         return "\\CJKunbreakablekerntwo ";
 341     }
 342     if(($i<$#chars)
 343        and($chars[$i+1]=~m/[$slightly_forbidden_before]/o)){
 344         return "\\CJKunbreakablekernthree ";
 345     }
 346     if($chars[$i]=~m/[$slightly_forbidden_after]/o){
 347         return "\\CJKunbreakablekernthree ";
 348     }
 349     if($chars[$i]=~m/[$asian]/o){
 350         return "\\CJKbreakablekern ";
 351     }
 352     if(($i<$#chars)and($chars[$i+1]=~m/[$asian]/o)){
 353         return "\\CJKbreakablekern ";
 354     }
 355 }
 356
 357 sub latin_parse{
 358     # arguments: none
 359     # return: string for output with TeX macro.
 360     my($char_id);
 361     my $out_str=$chars[$i];
 362     $i++;
 363     while($i<=$#chars){
 364         $char_id=unpack("U",$chars[$i]);
 365         if(($char_id>0x20 and $char_id<=0x021f)
 366            or($char_id>=0x0250 and $char_id<=0x02af)# IPA Extensions
 367            or($char_id>=0x0300 and $char_id<=0x033f)# Combining Diacritical Marks
 368            or($char_id>=0x0360 and $char_id<=0x036f)
 369            or($char_id>=0x0370 and $char_id<=0x03ff)# Greek and Coptic
 370            or($char_id>=0x0400 and $char_id<=0x04ff)# Cyrillic
 371            or($char_id>=0x0530 and $char_id<=0x058f)# Armenian
 372            ){
 373             $out_str.=$chars[$i];
 374         }elsif($char_id>=0x1e00 and $char_id<=0x1eff){
 375             # Latin Extended Additional
 376             # 0x1e00 -> 0x0600, etc.
 377             $out_str.=pack("U",$char_id-0x1800);
 378         }else{
 379             $i--;
 380             last;
 381         }
 382         $i++;
 383     }
 384     return '{\fontencoding{UT1}\fontfamily{omlgc}\selectfont '.$out_str.'}';
 385 }
 386
 387 sub ids_parse{
 388     # arguments: none
 389     # return: character for output,
 390     #          TeX macro for ids,
 391     #          or GETA character if ids is invalid.
 392     my($ids,$ids_argc)=&ids_rest("",0,$chars[$i]);
 393
 394     while($ids_argc>0){
 395         # We are in IDS.
 396         $i++;
 397         if($i>$#chars){
 398             print STDERR "IDS parse error: $ids\n";
 399             return $geta;
 400         }
 401
 402         ($ids,$ids_argc)=&ids_rest($ids,$ids_argc,$chars[$i]);
 403         if($ids_argc==0){
 404             if(($char=&get_char_for_ids($ids))
 405                and($out_char=&get_output_char($char,$out_cs))){
 406                 return $out_char;
 407             }else{
 408                 return &get_macro_for_ids($ids);
 409             }
 410         }
 411     }
 412 }
 413
 414 sub ids_rest{
 415     # arguments: <ids>, <rest number of arguments for ids>, <character>
 416     # return: ids and rest number of arguments for ids.
 417     my($ids,$ids_argc,$char)=@_;
 418     my($argc);
 419     $argc=&ids_argc($char);
 420     if($argc){
 421         $ids_argc+=$ids_argc==0?$argc:$argc-1;
 422     }else{
 423         $ids_argc--;
 424     }
 425     $ids.=$char;
 426     return ($ids,$ids_argc);
 427 }
 428
 429 sub get_macro_for_ids{
 430     # argument: <ids>
 431     # return: TeX macro for ids
 432     #          or GETA character if ids is invalid for KAGE.
 433     my($ids)=@_;
 434     # $ids=&normalize_ids($ids,"UniJIS");
 435     return $geta if(($ids!~/[$idc]/)
 436                     or($ids=~/[\x{10000}-]/));
 437                     #irregular for KAGE.
 438     if(not defined($ids{$ids})){
 439         $ids{$ids}[0]=$font_start;
 440         $ids{$ids}[1]=$ids_start;
 441         $ids_start++;
 442     }
 443     if($ids_start>255){
 444         $ids_start=0;
 445         $font_start++;
 446     }
 447     return "{\\fontencoding{OT1}\\fontfamily{"
 448         .sprintf("chise%03d",$ids{$ids}[0])
 449         ."}\\selectfont\\char$ids{$ids}[1]}";
 450 }
 451
 452 sub normalize_ids{
 453     # argument: <ids>, <output coding system>
 454     # return: ids or GETA character if ids is invalid for KAGE.
 455     my($ids,$out_cs)=@_;
 456     $out_cs=~s/Uni(.+)/'ucs@'.lc($1)/e;
 457
 458     my $output_ids="";
 459     my($char,$output_char_id);
 460     while($ids=~m/(.)/g){
 461         $char=$1;
 462         if($char=~/[$idc]/){
 463             $output_ids.=$char;
 464         }elsif($output_char_id=&get_char_attribute($char,"=$out_cs")
 465            or $output_char_id=&get_char_attribute($char,"=ucs")
 466            or $output_char_id=&get_char_attribute($char,"=>$out_cs")
 467            or $output_char_id=&get_char_attribute($char,"=>ucs")
 468            or $output_char_id=&get_char_attribute($char,"=>ucs*")
 469               ){
 470             $output_ids.=pack("U",$output_char_id);
 471         }else{
 472             return $geta;
 473         }
 474     }
 475     return $output_ids;
 476 }
 477
 478 sub get_output_char{
 479     # argument: <char>
 480     # return: character in output coding system or TeX macro or undef.
 481     my($char,$out_cs)=@_;
 482     my($out_char_id,$suffix);
 483     my($gt,$hzk,$cdp);
 484
 485     foreach $out_cs (@{$order{$out_cs}}){
 486         if($out_cs eq 'UniJIS'
 487            and &get_char_attribute($char,"vnd-adobe-cid-unijis-utf16-h")){
 488             if($out_char_id=&get_char_attribute($char,'=ucs@jis')
 489                or $out_char_id=&get_char_attribute($char,'=ucs')
 490                or $out_char_id=&get_char_attribute($char,'=>ucs@jis')
 491                or $out_char_id=&get_char_attribute($char,'=>ucs')
 492                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 493                ){
 494                 return '{\selectjisfont\char'.$out_char_id.'}';
 495             }
 496         }elsif($out_cs eq 'UniGB'
 497                and &get_char_attribute($char,"vnd-adobe-cid-unigb-ucs2-h")){
 498             if($out_char_id=&get_char_attribute($char,'=ucs@gb')
 499                or $out_char_id=&get_char_attribute($char,'=ucs')
 500                or $out_char_id=&get_char_attribute($char,'=>ucs@gb')
 501                or $out_char_id=&get_char_attribute($char,'=>ucs')
 502                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 503                ){
 504                 return '{\selectgbsfont\char'.$out_char_id.'}';
 505             }
 506         }elsif($out_cs eq 'UniCNS'
 507                and &get_char_attribute($char,"vnd-adobe-cid-unicns-ucs2-h")){
 508             if($out_char_id=&get_char_attribute($char,'=ucs@cns')
 509                or $out_char_id=&get_char_attribute($char,'=ucs')
 510                or $out_char_id=&get_char_attribute($char,'=>ucs@cns')
 511                or $out_char_id=&get_char_attribute($char,'=>ucs')
 512                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 513                ){
 514                 return '{\selectcnsfont\char'.$out_char_id.'}';
 515             }
 516         }elsif($out_cs eq 'UniKS'
 517                and &get_char_attribute($char,"vnd-adobe-cid-uniks-ucs2-h")){
 518             if($out_char_id=&get_char_attribute($char,'=ucs@ks')
 519                or $out_char_id=&get_char_attribute($char,'=ucs')
 520                or $out_char_id=&get_char_attribute($char,'=>ucs@ks')
 521                or $out_char_id=&get_char_attribute($char,'=>ucs')
 522                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 523                ){
 524                 return '{\selectksxfont\char'.$out_char_id.'}';
 525             }
 526         }elsif($out_cs eq 'GT'){
 527             return $gt if($gt=&get_macro_for_GT($char));
 528         }elsif($out_cs eq 'HZK'){
 529             return $hzk if($hzk=&get_macro_for_HZK($char));
 530         }elsif($out_cs eq 'CDP'){
 531             return $cdp if($cdp=&get_macro_for_CDP($char));
 532         }
 533     }
 534     return undef;
 535 }
 536
 537 sub get_ids{
 538     # argument: <character>
 539     # return: ids
 540     my($char)=@_;
 541     my $ids="";
 542     $ids=&get_char_attribute($char,"ids-aggregated")
 543         or $ids=&get_char_attribute($char,"ids");
 544 #    $ids=&get_char_attribute($char,"ids-decomposed")
 545 #       or $ids=&get_char_attribute($char,"ids");
 546 #         or $ids=&get_char_attribute($char,"ideographic-structure");
 547 #    $ids=~s/[? ()]//g;
 548     return $ids;
 549 }
 550
 551 sub get_char_for_ids{
 552     # argument: <ideographic description sequence>
 553     # return: char or undef.
 554     my($ids)=@_;
 555     my($output_char);
 556
 557     if(($output_char)=&get_chars_matching("ids",$ids)){
 558         return $output_char;
 559     }else{
 560         return undef;
 561     }
 562 }
 563
 564 sub get_char_in_mcs{
 565     # argument: <char>, <input coding system>
 566     # return:   char in ucs@mcs.
 567     my($char,$in_cs)=@_;
 568     my($output_char);
 569
 570     return $char if($in_cs eq 'ucs@mcs');
 571
 572     if(($output_char)=&get_chars_matching("=$in_cs",unpack("U",$char))){
 573         return $output_char;
 574     }else{
 575         return $char;
 576     }
 577 }
 578
 579 sub get_chars_unified{
 580     my($char)=@_;
 581     my($chars,$ucs,$char_ucs);
 582     my(@chars);
 583
 584     if($chars=&get_char_attribute($char,'->ucs-unified')){
 585         $chars=~s/^\((.*)\)$/$1/;
 586         return (split(/\s*\?/,$chars));
 587     }elsif($ucs=&get_char_attribute($char,'=>ucs*')
 588           or $ucs=&get_char_attribute($char,'=>ucs')){
 589         $char_ucs=pack("U",$ucs);
 590         if($chars=&get_char_attribute($char_ucs,'->ucs-unified')){
 591             $chars=~s/^\((.*)\)$/$1/;
 592             @chars=grep {not /^$char$/}
 593                 (split(/\s*\?/,$chars));
 594             push(@chars,$char_ucs);
 595             return @chars;
 596         }
 597     }
 598 }
 599
 600 sub get_macro_for_GT{
 601     # argument: <char>
 602     # return: TeX macro for GT fonts or undef.
 603     my($char)=@_;
 604     my($gt,$GT);
 605     foreach (@GT){
 606         if($gt=&get_char_attribute($char,$_)){
 607             m/gt\-pj\-(\d+)/ and $GT=$1;
 608             last;
 609         }
 610     }
 611     if($gt){
 612         return "{\\fontencoding{OT1}\\fontfamily{"
 613             .sprintf("gt%02d",$GT)
 614             ."}\\selectfont\\char".($gt|0x8080)."}";
 615     }else{
 616         return undef;
 617     }
 618 }
 619
 620 sub get_macro_for_HZK{
 621     # argument: <char>
 622     # return: TeX macro for Hanziku fonts or undef.
 623     my($char)=@_;
 624     my($hzk,$HZK);
 625     foreach (@HZK){
 626         if($hzk=&get_char_attribute($char,$_)){
 627             m/hanziku\-(\d+)/ and $HZK=$1;
 628             last;
 629         }
 630     }
 631     if($hzk){
 632         return "{\\fontencoding{OT1}\\fontfamily{".sprintf("hzk%02d",$HZK)."}\\selectfont\\char".$hzk."}";
 633     }else{
 634         return undef;
 635     }
 636 }
 637
 638 sub get_macro_for_CDP{
 639     # argument: <char>
 640     # return: TeX macro for CDP fonts or undef.
 641     my($char)=@_;
 642     my($cdp,$ucs);
 643     foreach (@CDP){
 644         if($cdp=&get_char_attribute($char,$_)){
 645             last;
 646         }
 647     }
 648     if($cdp){
 649         $ucs=&get_char_attribute(&get_chars_matching("=big5-pua",$cdp),"=ucs");
 650         if($ucs){
 651             return "{\\fontencoding{OT1}\\fontfamily{cdp}\\selectfont\\char"
 652                 .$ucs.
 653                     "}";
 654         }else{
 655             print STDERR "This should not happen.\n";
 656             print STDERR "ucs code point of CDP: $cdp not found.\n";
 657         }
 658     }else{
 659         return undef;
 660     }
 661 }