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}';
  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     if($in_cs ne 'ucs@mcs'){
 174         s/(.)/&get_char_in_mcs($1,$in_cs)/ge;
 175     }
 176     s/(amp.+?;)/&de_tex_er($1)/ge;
 177 #    s/(&.+?;)/&de_tex_er($1)/ge;
 178     @chars=split(//);
 179   CHAR:
 180     for($i=0;$i<=$#chars;$i++){
 181         $char=$chars[$i];
 182         $char_id=unpack("U",$char);
 183
 184         if($char_id<=0x20){
 185             print $chars[$i];
 186             next CHAR;
 187         }elsif($char=~m/($tex_meta_re)/o){
 188             print $tex_meta{$1};
 189             next CHAR;
 190         }elsif(($char_id>0x20 and $char_id<=0x021f)
 191                # Basic Latin
 192                # Latin-1 Supplement
 193                # Latin Extended-A
 194                # Latin Extended-B (not all)
 195                or($char_id>=0x0250 and $char_id<=0x02af)
 196                # IPA Extensions
 197                or($char_id>=0x0300 and $char_id<=0x033f)
 198                or($char_id>=0x0360 and $char_id<=0x036f)
 199                # Combining Diacritical Marks
 200                or($char_id>=0x1e00 and $char_id<=0x1eff)
 201                # Latin Extended Additional
 202                or($char_id>=0x0370 and $char_id<=0x03ff)
 203                # Greek and Coptic
 204                or($char_id>=0x0400 and $char_id<=0x04ff)
 205                # Cyrillic
 206                or($char_id>=0x0530 and $char_id<=0x058f)
 207                # Armenian
 208                ){
 209             print &latin_parse();
 210             next CHAR;
 211         }elsif($char_id>=0x2ff0 and $char_id<=0x2fff){
 212             # Ideographic Description Characters
 213             print &ids_parse();
 214             next CHAR;
 215         }else{
 216             if(($out_char=&get_output_char($char,$out_cs))){
 217                 print $out_char,&add_break($i);
 218             }else{
 219                 if($opt_allow_unify){
 220                     @chars_unified=&get_chars_unified($char);
 221                     if(@chars_unified>0){
 222                         foreach $char_unified (@chars_unified){
 223                             if(($out_char
 224                                 =&get_output_char($char_unified,$out_cs))){
 225                                 print $out_char,&add_break($i);
 226                                 next CHAR;
 227                             }
 228                         }
 229                     }
 230                 }
 231                 if($opt_use_kage_for_Ext_B){
 232                     if($char_id >= 0x20000 && $char_id <=0x2a6df){
 233                         # CJK Unified Ideographs Extension B
 234                         if(not defined($ids{$char}) and $ids{$char}[1]>=0){
 235                             $ids{$char}[0]=$font_start;
 236                             $ids{$char}[1]=$ids_start;
 237                             $ids_start++;
 238                             if($ids_start>255){
 239                                 $ids_start=0;
 240                                 $font_start++;
 241                             }
 242                         }
 243                         print "{\\fontencoding{OT1}\\fontfamily{" .
 244                             sprintf("chise%03d",$ids{$char}[0]) .
 245                                 "}\\selectfont\\char$ids{$char}[1]}",&add_break($i);
 246                         next CHAR;
 247                     }
 248                 }
 249                 if($ids=&get_ids($char)){
 250                     print &get_macro_for_ids($ids),&add_break($i);
 251                 }else{
 252                     print '\rule{1ex}{1ex}',&add_break($i);
 253                 }
 254             }
 255         }
 256     }
 257 }
 258
 259 print IDSDATA 'use utf8;',"\n";
 260 foreach $ids (keys %ids){
 261     print IDSDATA '$ids{\'',$ids,'\'}='
 262     ,'[',join ",",@{$ids{$ids}},"];\n";
 263 }
 264 print IDSDATA '$font_start=',$font_start,";\n";
 265 print IDSDATA '$ids_start=',$ids_start,";\n";
 266 print IDSDATA "1;";
 267 flock(IDSDATA,LOCK_UN);
 268
 269 exit 0;
 270
 271 sub de_tex_er{
 272     my($er)=@_;
 273     my($prefix,$suffix);
 274     my($output_char,$atr,$value);
 275     $er=~/^(amp)(.*)(;)$/
 276         and $prefix=$1,$er=$2,$suffix=$3;
 277     $prefix or $prefix="",$suffix or $suffix="";
 278     if($er=~/^U[\+|\-]([a-fA-F\d]+)/){
 279         $output_char=pack("U",hex($1));
 280     }elsif($er=~/^(?:I\-)?($er_prefix_re)\-?([0-9a-fA-F]+)$/){
 281         ($atr,$value)=($1,$2);
 282         unless($er_alias{$atr}=~/daikanwa|gt/){
 283             $value=hex($value);
 284         }
 285         ($output_char)=&get_chars_matching($er_alias{$atr},$value);
 286     }
 287     if($output_char){
 288         return $output_char;
 289     }else{
 290         return $prefix.$er.$suffix;
 291     }
 292 }
 293
 294 sub add_break{
 295     my($i)=@_;
 296
 297     if($i<($#chars-1)){
 298         if(($chars[$i+1]=~m/[$strictly_forbidden_before]/o)
 299            and($chars[$i+2]=~m/[$strictly_forbidden_before]/o)){
 300             return "\\CJKunbreakablekernone ";
 301         }elsif($opt_protrude){
 302             if(($chars[$i+1]=~m/[$strictly_forbidden_before]/o)
 303                and($chars[$i+2]=~m/[^$strictly_forbidden_before]/o)){
 304                 return "\\CJKunbreakablekernone \\CJKprotrude ";
 305             }
 306         }
 307     }
 308     if(($i<$#chars)
 309        and($chars[$i+1]=~m/[$strictly_forbidden_before]/o)){
 310         return "\\CJKunbreakablekernone ";
 311     }
 312     if($chars[$i]=~m/[$strictly_forbidden_after]/o){
 313         return "\\CJKunbreakablekernone ";
 314     }
 315     if(($i<$#chars)
 316        and($chars[$i+1]=~m/[$forbidden_before]/o)){
 317         return "\\CJKunbreakablekerntwo ";
 318
 319     }
 320     if($chars[$i]=~m/[$forbidden_after]/o){
 321         return "\\CJKunbreakablekerntwo ";
 322     }
 323     if(($i<$#chars)
 324        and($chars[$i+1]=~m/[$slightly_forbidden_before]/o)){
 325         return "\\CJKunbreakablekernthree ";
 326     }
 327     if($chars[$i]=~m/[$slightly_forbidden_after]/o){
 328         return "\\CJKunbreakablekernthree ";
 329     }
 330     if($chars[$i]=~m/[$asian]/o){
 331         return "\\CJKbreakablekern ";
 332     }
 333     if(($i<$#chars)and($chars[$i+1]=~m/[$asian]/o)){
 334         return "\\CJKbreakablekern ";
 335     }
 336 }
 337
 338 sub latin_parse{
 339     # arguments: none
 340     # return: string for output with TeX macro.
 341     my($char_id);
 342     my $out_str=$chars[$i];
 343     $i++;
 344     while($i<=$#chars){
 345         $char_id=unpack("U",$chars[$i]);
 346         if(($char_id>0x20 and $char_id<=0x021f)
 347            or($char_id>=0x0250 and $char_id<=0x02af)# IPA Extensions
 348            or($char_id>=0x0300 and $char_id<=0x033f)# Combining Diacritical Marks
 349            or($char_id>=0x0360 and $char_id<=0x036f)
 350            or($char_id>=0x0370 and $char_id<=0x03ff)# Greek and Coptic
 351            or($char_id>=0x0400 and $char_id<=0x04ff)# Cyrillic
 352            or($char_id>=0x0530 and $char_id<=0x058f)# Armenian
 353            ){
 354             $out_str.=$chars[$i];
 355         }elsif($char_id>=0x1e00 and $char_id<=0x1eff){
 356             # Latin Extended Additional
 357             # 0x1e00 -> 0x0600, etc.
 358             $out_str.=pack("U",$char_id-0x1800);
 359         }else{
 360             $i--;
 361             last;
 362         }
 363         $i++;
 364     }
 365     return '{\fontencoding{UT1}\fontfamily{omlgc}\selectfont '.$out_str.'}';
 366 }
 367
 368 sub ids_parse{
 369     # arguments: none
 370     # return: character for output,
 371     #          TeX macro for ids,
 372     #          or GETA character if ids is invalid.
 373     my($ids,$ids_argc)=&ids_rest("",0,$chars[$i]);
 374
 375     while($ids_argc>0){
 376         # We are in IDS.
 377         $i++;
 378         if($i>$#chars){
 379             print STDERR "IDS parse error: $ids\n";
 380             return $geta;
 381         }
 382
 383         ($ids,$ids_argc)=&ids_rest($ids,$ids_argc,$chars[$i]);
 384         if($ids_argc==0){
 385             if(($char=&get_char_for_ids($ids))
 386                and($out_char=&get_output_char($char,$out_cs))){
 387                 return $out_char;
 388             }else{
 389                 return &get_macro_for_ids($ids);
 390             }
 391         }
 392     }
 393 }
 394
 395 sub ids_rest{
 396     # arguments: <ids>, <rest number of arguments for ids>, <character>
 397     # return: ids and rest number of arguments for ids.
 398     my($ids,$ids_argc,$char)=@_;
 399     my($argc);
 400     $argc=&ids_argc($char);
 401     if($argc){
 402         $ids_argc+=$ids_argc==0?$argc:$argc-1;
 403     }else{
 404         $ids_argc--;
 405     }
 406     $ids.=$char;
 407     return ($ids,$ids_argc);
 408 }
 409
 410 sub get_macro_for_ids{
 411     # argument: <ids>
 412     # return: TeX macro for ids
 413     #          or GETA character if ids is invalid for KAGE.
 414     my($ids)=@_;
 415     # $ids=&normalize_ids($ids,"UniJIS");
 416     return $geta if(($ids!~/[$idc]/)
 417                     or($ids=~/[\x{10000}-]/));
 418                     #irregular for KAGE.
 419     if(not defined($ids{$ids})){
 420         $ids{$ids}[0]=$font_start;
 421         $ids{$ids}[1]=$ids_start;
 422         $ids_start++;
 423     }
 424     if($ids_start>255){
 425         $ids_start=0;
 426         $font_start++;
 427     }
 428     return "{\\fontencoding{OT1}\\fontfamily{"
 429         .sprintf("chise%03d",$ids{$ids}[0])
 430         ."}\\selectfont\\char$ids{$ids}[1]}";
 431 }
 432
 433 sub normalize_ids{
 434     # argument: <ids>, <output coding system>
 435     # return: ids or GETA character if ids is invalid for KAGE.
 436     my($ids,$out_cs)=@_;
 437     $out_cs=~s/Uni(.+)/'ucs@'.lc($1)/e;
 438
 439     my $output_ids="";
 440     my($char,$output_char_id);
 441     while($ids=~m/(.)/g){
 442         $char=$1;
 443         if($char=~/[$idc]/){
 444             $output_ids.=$char;
 445         }elsif($output_char_id=&get_char_attribute($char,"=$out_cs")
 446            or $output_char_id=&get_char_attribute($char,"=ucs")
 447            or $output_char_id=&get_char_attribute($char,"=>$out_cs")
 448            or $output_char_id=&get_char_attribute($char,"=>ucs")
 449            or $output_char_id=&get_char_attribute($char,"=>ucs*")
 450               ){
 451             $output_ids.=pack("U",$output_char_id);
 452         }else{
 453             return $geta;
 454         }
 455     }
 456     return $output_ids;
 457 }
 458
 459 sub get_output_char{
 460     # argument: <char>
 461     # return: character in output coding system or TeX macro or undef.
 462     my($char,$out_cs)=@_;
 463     my($out_char_id,$suffix);
 464     my($gt,$hzk,$cdp);
 465
 466     foreach $out_cs (@{$order{$out_cs}}){
 467         if($out_cs eq 'UniJIS'
 468            and &get_char_attribute($char,"vnd-adobe-cid-unijis-utf16-h")){
 469             if($out_char_id=&get_char_attribute($char,'=ucs@jis')
 470                or $out_char_id=&get_char_attribute($char,'=ucs')
 471                or $out_char_id=&get_char_attribute($char,'=>ucs@jis')
 472                or $out_char_id=&get_char_attribute($char,'=>ucs')
 473                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 474                ){
 475                 return '{\selectjisfont\char'.$out_char_id.'}';
 476             }
 477         }elsif($out_cs eq 'UniGB'
 478                and &get_char_attribute($char,"vnd-adobe-cid-unigb-ucs2-h")){
 479             if($out_char_id=&get_char_attribute($char,'=ucs@gb')
 480                or $out_char_id=&get_char_attribute($char,'=ucs')
 481                or $out_char_id=&get_char_attribute($char,'=>ucs@gb')
 482                or $out_char_id=&get_char_attribute($char,'=>ucs')
 483                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 484                ){
 485                 return '{\selectgbsfont\char'.$out_char_id.'}';
 486             }
 487         }elsif($out_cs eq 'UniCNS'
 488                and &get_char_attribute($char,"vnd-adobe-cid-unicns-ucs2-h")){
 489             if($out_char_id=&get_char_attribute($char,'=ucs@cns')
 490                or $out_char_id=&get_char_attribute($char,'=ucs')
 491                or $out_char_id=&get_char_attribute($char,'=>ucs@cns')
 492                or $out_char_id=&get_char_attribute($char,'=>ucs')
 493                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 494                ){
 495                 return '{\selectcnsfont\char'.$out_char_id.'}';
 496             }
 497         }elsif($out_cs eq 'UniKS'
 498                and &get_char_attribute($char,"vnd-adobe-cid-uniks-ucs2-h")){
 499             if($out_char_id=&get_char_attribute($char,'=ucs@ks')
 500                or $out_char_id=&get_char_attribute($char,'=ucs')
 501                or $out_char_id=&get_char_attribute($char,'=>ucs@ks')
 502                or $out_char_id=&get_char_attribute($char,'=>ucs')
 503                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 504                ){
 505                 return '{\selectksxfont\char'.$out_char_id.'}';
 506             }
 507         }elsif($out_cs eq 'GT'){
 508             return $gt if($gt=&get_macro_for_GT($char));
 509         }elsif($out_cs eq 'HZK'){
 510             return $hzk if($hzk=&get_macro_for_HZK($char));
 511         }elsif($out_cs eq 'CDP'){
 512             return $cdp if($cdp=&get_macro_for_CDP($char));
 513         }
 514     }
 515     return undef;
 516 }
 517
 518 sub get_ids{
 519     # argument: <character>
 520     # return: ids
 521     my($char)=@_;
 522     my $ids="";
 523     $ids=&get_char_attribute($char,"ids-aggregated")
 524         or $ids=&get_char_attribute($char,"ids");
 525 #    $ids=&get_char_attribute($char,"ids-decomposed")
 526 #       or $ids=&get_char_attribute($char,"ids");
 527 #         or $ids=&get_char_attribute($char,"ideographic-structure");
 528 #    $ids=~s/[? ()]//g;
 529     return $ids;
 530 }
 531
 532 sub get_char_for_ids{
 533     # argument: <ideographic description sequence>
 534     # return: char or undef.
 535     my($ids)=@_;
 536     my($output_char);
 537
 538     if(($output_char)=&get_chars_matching("ids",$ids)){
 539         return $output_char;
 540     }else{
 541         return undef;
 542     }
 543 }
 544
 545 sub get_char_in_mcs{
 546     # argument: <char>, <input coding system>
 547     # return:   char in ucs@mcs.
 548     my($char,$in_cs)=@_;
 549     my($output_char);
 550
 551     return $char if($in_cs eq 'ucs@mcs');
 552
 553     if(($output_char)=&get_chars_matching("=$in_cs",unpack("U",$char))){
 554         return $output_char;
 555     }else{
 556         return $char;
 557     }
 558 }
 559
 560 sub get_chars_unified{
 561     my($char)=@_;
 562     my($chars,$ucs,$char_ucs);
 563     my(@chars);
 564
 565     if($chars=&get_char_attribute($char,'->ucs-unified')){
 566         $chars=~s/^\((.*)\)$/$1/;
 567         return (split(/\s*\?/,$chars));
 568     }elsif($ucs=&get_char_attribute($char,'=>ucs*')
 569           or $ucs=&get_char_attribute($char,'=>ucs')){
 570         $char_ucs=pack("U",$ucs);
 571         if($chars=&get_char_attribute($char_ucs,'->ucs-unified')){
 572             $chars=~s/^\((.*)\)$/$1/;
 573             @chars=grep {not /^$char$/}
 574                 (split(/\s*\?/,$chars));
 575             push(@chars,$char_ucs);
 576             return @chars;
 577         }
 578     }
 579 }
 580
 581 sub get_macro_for_GT{
 582     # argument: <char>
 583     # return: TeX macro for GT fonts or undef.
 584     my($char)=@_;
 585     my($gt,$GT);
 586     foreach (@GT){
 587         if($gt=&get_char_attribute($char,$_)){
 588             m/gt\-pj\-(\d+)/ and $GT=$1;
 589             last;
 590         }
 591     }
 592     if($gt){
 593         return "{\\fontencoding{OT1}\\fontfamily{"
 594             .sprintf("gt%02d",$GT)
 595             ."}\\selectfont\\char".($gt|0x8080)."}";
 596     }else{
 597         return undef;
 598     }
 599 }
 600
 601 sub get_macro_for_HZK{
 602     # argument: <char>
 603     # return: TeX macro for Hanziku fonts or undef.
 604     my($char)=@_;
 605     my($hzk,$HZK);
 606     foreach (@HZK){
 607         if($hzk=&get_char_attribute($char,$_)){
 608             m/hanziku\-(\d+)/ and $HZK=$1;
 609             last;
 610         }
 611     }
 612     if($hzk){
 613         return "{\\fontencoding{OT1}\\fontfamily{".sprintf("hzk%02d",$HZK)."}\\selectfont\\char".$hzk."}";
 614     }else{
 615         return undef;
 616     }
 617 }
 618
 619 sub get_macro_for_CDP{
 620     # argument: <char>
 621     # return: TeX macro for CDP fonts or undef.
 622     my($char)=@_;
 623     my($cdp,$ucs);
 624     foreach (@CDP){
 625         if($cdp=&get_char_attribute($char,$_)){
 626             last;
 627         }
 628     }
 629     if($cdp){
 630         $ucs=&get_char_attribute(&get_chars_matching("=big5-pua",$cdp),"=ucs");
 631         if($ucs){
 632             return "{\\fontencoding{OT1}\\fontfamily{cdp}\\selectfont\\char"
 633                 .$ucs.
 634                     "}";
 635         }else{
 636             print STDERR "This should not happen.\n";
 637             print STDERR "ucs code point of CDP: $cdp not found.\n";
 638         }
 639     }else{
 640         return undef;
 641     }
 642 }