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 my $asian = '\x{2E80}-\x{312f}\x{3190}-\x{ABFF}\x{F900}-\x{FAFF}\x{FE30}-\x{FE4F}\x{FF00}-\x{FFFFFF}'; # need to fix.
  65
  66 my $space = '\x{0020}\x{0009}\x{000A}\x{000C}\x{000D}';
  67
  68 my $hangul='\x{1100}-\x{11ff}\x{3130}-\x{318f}\x{AC00}-\x{D7AF}';
  69
  70 my %tex_meta=('#'=>'\#',
  71               '$'=>'\\textdollar{}',
  72               '%'=>'\%',
  73               '&'=>'\&',
  74               '{'=>'\\textbraceleft{}',
  75               '}'=>'\\textbraceright{}',
  76               '\\'=>'\\textbackslash{}',
  77               '_'=>'\\textunderscore',
  78              );
  79
  80 my $tex_meta_re=join('|',map {quotemeta($_)} keys %tex_meta);
  81
  82 &GetOptions("in=s"=>\$opt_in_cs,
  83             "out=s"=>\$opt_out_cs,
  84             "help",\$opt_help);
  85
  86 $usage=<<EOF;
  87 Usage: $0 -i <input coding system> -o <cmap encoding>
  88     input coding system:
  89       Utf8mcs, Utf8cns, Utf8gb, Utf8jis, Utf8ks, Utf8big5
  90     cmap encoding:
  91       UniCNS, UniGB, UniJIS, UniKS, UniMulti
  92 EOF
  93
  94 if($opt_in_cs or $opt_out_cs){
  95     $in_cs=$opt_in_cs;
  96     $out_cs=$opt_out_cs;
  97 }elsif(@ARGV==0){
  98     ($in_cs,$out_cs)=($0=~/(Utf8.+)To(\w+)/);
  99 }
 100
 101 # $in_cs:
 102 #   Utf8mcs,Utf8cns,Utf8gb,Utf8jis,Utf8ks,Utf8big5
 103 # $out_cs:
 104 #   UniCNS,UniGB,UniJIS,UniKS,UniMulti
 105
 106 $in_cs=~s/Utf8/ucs\@/;
 107
 108 if($opt_help
 109    or not defined($in_cs)
 110    or not defined($out_cs)){
 111     print $usage;
 112     exit 1;
 113 }
 114
 115 $omegadb_path=~s!/$!!;
 116
 117 $idsdata_file="$omegadb_path/idsdata.pl";
 118 $ids_start=0x00;
 119 $font_start=0;
 120
 121 if(-e $idsdata_file){
 122     open(IDSDATA,"+<:utf8",$idsdata_file) or die;
 123     flock(IDSDATA,LOCK_EX);
 124     seek(IDSDATA,0,0);
 125     while(<IDSDATA>){
 126         eval $_;
 127     }
 128     seek(IDSDATA,0,0);
 129 #         require $idsdata_file;
 130 }else{
 131     open(IDSDATA,">:utf8",$idsdata_file) or die;
 132     flock(IDSDATA,LOCK_EX);
 133     seek(IDSDATA,0,0);
 134 }
 135
 136 $ids_argc=0;
 137 $ids="";
 138
 139 #$geta=pack("U",0x3013);
 140 $geta=pack("U",0xfffd);
 141
 142 @GT=(#"=gt","=gt-k",
 143      "=gt-pj-1","=gt-pj-2","=gt-pj-3","=gt-pj-4","=gt-pj-5",
 144      "=gt-pj-6","=gt-pj-7","=gt-pj-8","=gt-pj-9","=gt-pj-10",
 145      "=gt-pj-11"
 146      #,"=gt-pj-k1","=gt-pj-k2"
 147      );
 148 @HZK=("=hanziku-1","=hanziku-2","=hanziku-3","=hanziku-4",
 149       "=hanziku-5","=hanziku-6","=hanziku-7","=hanziku-8",
 150       "=hanziku-9","=hanziku-10","=hanziku-11","=hanziku-12");
 151 @CDP=("=big5-cdp");
 152
 153 %order_map=('c'=>'UniCNS',
 154             'g'=>'UniGB',
 155             'j'=>'UniJIS',
 156             'k'=>'UniKS',
 157             'G'=>'GT',
 158             'H'=>'HZK',
 159             'C'=>'CDP',
 160             );
 161
 162 foreach $out_cs ('UniCNS','UniGB','UniJIS','UniKS','UniMulti'){
 163     if(defined($opt_order{$out_cs})){
 164         if($opt_order{$out_cs}=~/^[cgjkGHC]+$/){
 165             @{$order{$out_cs}}=map {$order_map{$_}}
 166             (split(//,$opt_order{$out_cs}));
 167         }else{
 168             print STDERR "Invalid order for $out_cs!\n";
 169             exit 1;
 170         }
 171     }
 172 }
 173
 174 while(<>){
 175     utf8::decode($_);
 176
 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($1)/ge;
 182     s/([$asian])\s+/$1/go unless($out_cs eq 'UniKS');
 183     s/\s+([$asian])/$1/go unless($out_cs eq 'UniKS');
 184     s/([$asian])\s*([^$asian$space])/$1 $2/go;
 185     s/([^$asian$idc])\s*([$asian])/$1 $2/go;
 186     s/\-\-\-/pack("U",0x2014)/geo;# EM DASH
 187     s/\-\-/pack("U",0x2013)/geo;# EN DASH
 188     s/\`\`/pack("U",0x201f)/geo;# DOUBLE HIGH-REVERSED-9 QUOTATION MARK
 189     s/\`/pack("U",0x201b)/geo;# SINGLE HIGH-REVERSED-9 QUOTATION MARK
 190     s/\'\'/pack("U",0x201d)/geo;# RIGHT DOUBLE QUOTATION MARK
 191     s/\'/pack("U",0x2019)/geo;# RIGHT DOUBLE QUOTATION MARK
 192     @chars=split(//);
 193   CHAR:
 194     for($i=0;$i<=$#chars;$i++){
 195         $char=$chars[$i];
 196         $char_id=unpack("U",$char);
 197
 198         if($char_id<=0x20){
 199             # add quarter space between asian and non-asian.
 200             if($i>0 and $i<$#chars){
 201                 if(($chars[$i-1]=~/[$asian]/
 202                     and $chars[$i+1]=~/[^$asian]/
 203                       and $chars[$i+1]=~/[$strictly_forbidden_before$slightly_forbidden_before]/)
 204                    or($chars[$i-1]=~/[^$asian]/
 205                       and $chars[$i+1]=~/[$asian]/
 206                       and $chars[$i-1]=~/[$strictly_forbidden_after$slightly_forbidden_after]/)
 207                    ){
 208                     print '';
 209                 }elsif(($chars[$i-1]=~/[$asian]/
 210                         and $chars[$i+1]=~/[^$asian]/)){
 211                     print '\unskip\kern.25ex';
 212                 }elsif($chars[$i-1]=~/[^$asian]/
 213                        and $chars[$i+1]=~/[$asian]/){
 214                     print '\kern.25ex';
 215                 }else{
 216                     print ' ';
 217                 }
 218             }
 219             next CHAR;
 220         }elsif($char=~m/($tex_meta_re)/o){
 221             print $tex_meta{$1};
 222             next CHAR;
 223         }elsif(($char_id>0x20 and $char_id<=0x021f)
 224                # Basic Latin
 225                # Latin-1 Supplement
 226                # Latin Extended-A
 227                # Latin Extended-B (not all)
 228                or($char_id>=0x0250 and $char_id<=0x02af)
 229                # IPA Extensions
 230                or($char_id>=0x0300 and $char_id<=0x033f)
 231                or($char_id>=0x0360 and $char_id<=0x036f)
 232                # Combining Diacritical Marks
 233                or($char_id>=0x1e00 and $char_id<=0x1eff)
 234                # Latin Extended Additional
 235                or($char_id>=0x0370 and $char_id<=0x03ff)
 236                # Greek and Coptic
 237                or($char_id>=0x0400 and $char_id<=0x04ff)
 238                # Cyrillic
 239                or($char_id>=0x0530 and $char_id<=0x058f)
 240                # Armenian
 241                or($char_id>=0x2010 and $char_id<=0x2046)
 242                # General Punctuation (partial)
 243                ){
 244             print &latin_parse();
 245             next CHAR;
 246         }elsif($char_id>=0x2ff0 and $char_id<=0x2fff){
 247             # Ideographic Description Characters
 248             print &ids_parse();
 249             next CHAR;
 250         }else{
 251             if(($out_char=&get_output_char($char,$out_cs))){
 252                 print $out_char,&add_break($i);
 253             }else{
 254                 if($opt_allow_unify){
 255                     @chars_unified=&get_chars_unified($char);
 256                     if(@chars_unified>0){
 257                         foreach $char_unified (@chars_unified){
 258                             if(($out_char
 259                                 =&get_output_char($char_unified,$out_cs))){
 260                                 print $out_char,&add_break($i);
 261                                 next CHAR;
 262                             }
 263                         }
 264                     }
 265                 }
 266                 if($opt_use_kage_for_Ext_B){
 267                     if($char_id >= 0x20000 && $char_id <=0x2a6df){
 268                         # CJK Unified Ideographs Extension B
 269                         if(not defined($ids{$char}) and $ids{$char}[1]>=0){
 270                             $ids{$char}[0]=$font_start;
 271                             $ids{$char}[1]=$ids_start;
 272                             $ids_start++;
 273                             if($ids_start>255){
 274                                 $ids_start=0;
 275                                 $font_start++;
 276                             }
 277                         }
 278                         print "{\\fontencoding{OT1}\\fontfamily{",
 279                             sprintf("chise%03d",$ids{$char}[0]) ,
 280                                 "}\\selectfont\\char$ids{$char}[1]}",&add_break($i);
 281                         next CHAR;
 282                     }
 283                 }
 284                 if($ids=&get_ids($char)){
 285                     print &get_macro_for_ids($ids),&add_break($i);
 286                 }else{
 287                     print '\rule{1ex}{1ex}',&add_break($i);
 288                 }
 289             }
 290         }
 291     }
 292 }
 293
 294 print IDSDATA 'use utf8;',"\n";
 295 foreach $ids (keys %ids){
 296     print IDSDATA '$ids{\'',$ids,'\'}='
 297     ,'[',join ",",@{$ids{$ids}},"];\n";
 298 }
 299 print IDSDATA '$font_start=',$font_start,";\n";
 300 print IDSDATA '$ids_start=',$ids_start,";\n";
 301 print IDSDATA "1;";
 302 flock(IDSDATA,LOCK_UN);
 303
 304 exit 0;
 305
 306 sub de_tex_er{
 307     my($er)=@_;
 308     my($prefix,$suffix);
 309     my($output_char,$atr,$value);
 310     $er=~/^(amp)(.*)(;)$/
 311         and $prefix=$1,$er=$2,$suffix=$3;
 312     $prefix or $prefix="",$suffix or $suffix="";
 313     if($er=~/^U[\+|\-]([a-fA-F\d]+)/){
 314         $output_char=pack("U",hex($1));
 315     }elsif($er=~/^(?:I\-)?($er_prefix_re)\-?([0-9a-fA-F]+)$/){
 316         ($atr,$value)=($1,$2);
 317         unless($er_alias{$atr}=~/daikanwa|gt/){
 318             $value=hex($value);
 319         }
 320         ($output_char)=&get_chars_matching($er_alias{$atr},$value);
 321     }
 322     if($output_char){
 323         return $output_char;
 324     }else{
 325         return $prefix.$er.$suffix;
 326     }
 327 }
 328
 329 sub add_break{
 330     my($i)=@_;
 331
 332     if($i<($#chars-1)){
 333         if(($chars[$i+1]=~m/[$strictly_forbidden_before]/o)
 334            and($chars[$i+2]=~m/[$strictly_forbidden_before]/o)){
 335             return "\\CJKunbreakablekernone{}";
 336         }elsif($opt_protrude){
 337             if(($chars[$i+1]=~m/[$strictly_forbidden_before]/o)
 338                and($chars[$i+2]=~m/[^$strictly_forbidden_before]/o)){
 339                 return "\\CJKunbreakablekernone \\CJKprotrude{}";
 340             }
 341         }
 342     }
 343     if(($i<$#chars)
 344        and($chars[$i+1]eq" ")){
 345        # preserve space.
 346         return "";
 347     }
 348     if(($i<$#chars)
 349        and($chars[$i+1]=~m/[$strictly_forbidden_before]/o)){
 350         return "\\CJKunbreakablekernone{}";
 351     }
 352     if($chars[$i]=~m/[$strictly_forbidden_after]/o){
 353         return "\\CJKunbreakablekernone{}";
 354     }
 355     if(($i<$#chars)
 356        and($chars[$i+1]=~m/[$forbidden_before]/o)){
 357         return "\\CJKunbreakablekerntwo{}";
 358
 359     }
 360     if($chars[$i]=~m/[$forbidden_after]/o){
 361         return "\\CJKunbreakablekerntwo{}";
 362     }
 363     if(($i<$#chars)
 364        and($chars[$i+1]=~m/[$slightly_forbidden_before]/o)){
 365         return "\\CJKunbreakablekernthree{}";
 366     }
 367     if($chars[$i]=~m/[$slightly_forbidden_after]/o){
 368         return "\\CJKunbreakablekernthree{}";
 369     }
 370     if($chars[$i]=~m/[$asian]/o){
 371         return "\\CJKbreakablekern{}";
 372     }
 373     if(($i<$#chars)and($chars[$i+1]=~m/[$asian]/o)){
 374         return "\\CJKbreakablekern{}";
 375     }
 376 }
 377
 378 sub latin_parse{
 379     # arguments: none
 380     # return: string for output with TeX macro.
 381     my($char_id);
 382     my $out_str="";
 383     while($i<=$#chars){
 384         $char_id=unpack("U",$chars[$i]);
 385         if(($char_id>0x20 and $char_id<=0x021f)
 386            or($char_id>=0x0250 and $char_id<=0x02af)# IPA Extensions
 387            or($char_id>=0x0300 and $char_id<=0x033f)# Combining Diacritical Marks
 388            or($char_id>=0x0360 and $char_id<=0x036f)
 389            or($char_id>=0x0370 and $char_id<=0x03ff)# Greek and Coptic
 390            or($char_id>=0x0400 and $char_id<=0x04ff)# Cyrillic
 391            or($char_id>=0x0530 and $char_id<=0x058f)# Armenian
 392            ){
 393             $out_str.=$chars[$i];
 394         }elsif($char_id>=0x1e00 and $char_id<=0x1eff){
 395             # Latin Extended Additional
 396             # 0x1e00 -> 0x0600, etc.
 397             $out_str.=pack("U",$char_id-0x1800);
 398         }elsif($char_id>=0x2010 and $char_id<=0x2046){
 399             # General Punctuation (partial)
 400             $out_str.=pack("U",$char_id-0x1000);
 401         }else{
 402             $i--;
 403             last;
 404         }
 405         $i++;
 406     }
 407     return '{\fontencoding{OT1}\fontfamily{omlgc}\selectfont '.$out_str.'}';###UT1?
 408 }
 409
 410 sub ids_parse{
 411     # arguments: none
 412     # return: character for output,
 413     #          TeX macro for ids,
 414     #          or GETA character if ids is invalid.
 415     my($ids,$ids_argc)=&ids_rest("",0,$chars[$i]);
 416
 417     while($ids_argc>0){
 418         # We are in IDS.
 419         $i++;
 420         if($i>$#chars){
 421             print STDERR "IDS parse error: $ids\n";
 422             return $geta;
 423         }
 424
 425         ($ids,$ids_argc)=&ids_rest($ids,$ids_argc,$chars[$i]);
 426         if($ids_argc==0){
 427             if(($char=&get_char_for_ids($ids))
 428                and($out_char=&get_output_char($char,$out_cs))){
 429                 return $out_char;
 430             }else{
 431                 return &get_macro_for_ids($ids);
 432             }
 433         }
 434     }
 435 }
 436
 437 sub ids_rest{
 438     # arguments: <ids>, <rest number of arguments for ids>, <character>
 439     # return: ids and rest number of arguments for ids.
 440     my($ids,$ids_argc,$char)=@_;
 441     my($argc);
 442     $argc=&ids_argc($char);
 443     if($argc){
 444         $ids_argc+=$ids_argc==0?$argc:$argc-1;
 445     }else{
 446         $ids_argc--;
 447     }
 448     $ids.=$char;
 449     return ($ids,$ids_argc);
 450 }
 451
 452 sub get_macro_for_ids{
 453     # argument: <ids>
 454     # return: TeX macro for ids
 455     #          or GETA character if ids is invalid for KAGE.
 456     my($ids)=@_;
 457     # $ids=&normalize_ids($ids,"UniJIS");
 458     return $geta if(($ids!~/[$idc]/)
 459                     or($ids=~/[\x{10000}-]/));
 460                     #irregular for KAGE.
 461     if(not defined($ids{$ids})){
 462         $ids{$ids}[0]=$font_start;
 463         $ids{$ids}[1]=$ids_start;
 464         $ids_start++;
 465     }
 466     if($ids_start>255){
 467         $ids_start=0;
 468         $font_start++;
 469     }
 470     return "{\\fontencoding{OT1}\\fontfamily{"
 471         .sprintf("chise%03d",$ids{$ids}[0])
 472         ."}\\selectfont\\char$ids{$ids}[1]}";
 473 }
 474
 475 sub normalize_ids{
 476     # argument: <ids>, <output coding system>
 477     # return: ids or GETA character if ids is invalid for KAGE.
 478     my($ids,$out_cs)=@_;
 479     $out_cs=~s/Uni(.+)/'ucs@'.lc($1)/e;
 480
 481     my $output_ids="";
 482     my($char,$output_char_id);
 483     while($ids=~m/(.)/g){
 484         $char=$1;
 485         if($char=~/[$idc]/){
 486             $output_ids.=$char;
 487         }elsif($output_char_id=&get_char_attribute($char,"=$out_cs")
 488            or $output_char_id=&get_char_attribute($char,"=ucs")
 489            or $output_char_id=&get_char_attribute($char,"=>$out_cs")
 490            or $output_char_id=&get_char_attribute($char,"=>ucs")
 491            or $output_char_id=&get_char_attribute($char,"=>ucs*")
 492               ){
 493             $output_ids.=pack("U",$output_char_id);
 494         }else{
 495             return $geta;
 496         }
 497     }
 498     return $output_ids;
 499 }
 500
 501 sub get_output_char{
 502     # argument: <char>
 503     # return: character in output coding system or TeX macro or undef.
 504     my($char,$out_cs)=@_;
 505     my($out_char_id,$suffix);
 506     my($gt,$hzk,$cdp);
 507
 508     foreach $out_cs (@{$order{$out_cs}}){
 509         if($out_cs eq 'UniJIS'
 510            and &get_char_attribute($char,"vnd-adobe-cid-unijis-utf16-h")){
 511             if($out_char_id=&get_char_attribute($char,'=ucs@jis')
 512                or $out_char_id=&get_char_attribute($char,'=ucs')
 513                or $out_char_id=&get_char_attribute($char,'=>ucs@jis')
 514                or $out_char_id=&get_char_attribute($char,'=>ucs')
 515                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 516                ){
 517                 return "{\\selectjisfont\\char$out_char_id}";
 518             }
 519         }elsif($out_cs eq 'UniGB'
 520                and &get_char_attribute($char,"vnd-adobe-cid-unigb-ucs2-h")){
 521             if($out_char_id=&get_char_attribute($char,'=ucs@gb')
 522                or $out_char_id=&get_char_attribute($char,'=ucs')
 523                or $out_char_id=&get_char_attribute($char,'=>ucs@gb')
 524                or $out_char_id=&get_char_attribute($char,'=>ucs')
 525                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 526                ){
 527                 return "{\\selectgbsfont\\char$out_char_id}";
 528             }
 529         }elsif($out_cs eq 'UniCNS'
 530                and &get_char_attribute($char,"vnd-adobe-cid-unicns-ucs2-h")){
 531             if($out_char_id=&get_char_attribute($char,'=ucs@cns')
 532                or $out_char_id=&get_char_attribute($char,'=ucs')
 533                or $out_char_id=&get_char_attribute($char,'=>ucs@cns')
 534                or $out_char_id=&get_char_attribute($char,'=>ucs')
 535                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 536                ){
 537                 return "{\\selectcnsfont\\char$out_char_id}";
 538             }
 539         }elsif($out_cs eq 'UniKS'
 540                and &get_char_attribute($char,"vnd-adobe-cid-uniks-ucs2-h")){
 541             if($out_char_id=&get_char_attribute($char,'=ucs@ks')
 542                or $out_char_id=&get_char_attribute($char,'=ucs')
 543                or $out_char_id=&get_char_attribute($char,'=>ucs@ks')
 544                or $out_char_id=&get_char_attribute($char,'=>ucs')
 545                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 546                ){
 547                 return "{\\selectksxfont\\char$out_char_id}";
 548             }
 549         }elsif($out_cs eq 'GT'){
 550             return $gt if($gt=&get_macro_for_GT($char));
 551         }elsif($out_cs eq 'HZK'){
 552             return $hzk if($hzk=&get_macro_for_HZK($char));
 553         }elsif($out_cs eq 'CDP'){
 554             return $cdp if($cdp=&get_macro_for_CDP($char));
 555         }
 556     }
 557     return undef;
 558 }
 559
 560 sub get_ids{
 561     # argument: <character>
 562     # return: ids
 563     my($char)=@_;
 564     my $ids="";
 565     $ids=&get_char_attribute($char,"ids-aggregated")
 566         or $ids=&get_char_attribute($char,"ids");
 567 #    $ids=&get_char_attribute($char,"ids-decomposed")
 568 #       or $ids=&get_char_attribute($char,"ids");
 569 #         or $ids=&get_char_attribute($char,"ideographic-structure");
 570 #    $ids=~s/[? ()]//g;
 571     return $ids;
 572 }
 573
 574 sub get_char_for_ids{
 575     # argument: <ideographic description sequence>
 576     # return: char or undef.
 577     my($ids)=@_;
 578     my($output_char);
 579
 580     if(($output_char)=&get_chars_matching("ids",$ids)){
 581         return $output_char;
 582     }else{
 583         return undef;
 584     }
 585 }
 586
 587 sub get_char_in_mcs{
 588     # argument: <char>, <input coding system>
 589     # return:   char in ucs@mcs.
 590     my($char,$in_cs)=@_;
 591     my($output_char);
 592
 593     return $char if($in_cs eq 'ucs@mcs');
 594
 595     if(($output_char)=&get_chars_matching("=$in_cs",unpack("U",$char))){
 596         return $output_char;
 597     }else{
 598         return $char;
 599     }
 600 }
 601
 602 sub get_chars_unified{
 603     my($char)=@_;
 604     my($chars,$ucs,$char_ucs);
 605     my(@chars);
 606
 607     if($chars=&get_char_attribute($char,'->ucs-unified')){
 608         $chars=~s/^\((.*)\)$/$1/;
 609         return (split(/\s*\?/,$chars));
 610     }elsif($ucs=&get_char_attribute($char,'=>ucs*')
 611           or $ucs=&get_char_attribute($char,'=>ucs')){
 612         $char_ucs=pack("U",$ucs);
 613         if($chars=&get_char_attribute($char_ucs,'->ucs-unified')){
 614             $chars=~s/^\((.*)\)$/$1/;
 615             @chars=grep {not /^$char$/}
 616                 (split(/\s*\?/,$chars));
 617             push(@chars,$char_ucs);
 618             return @chars;
 619         }
 620     }
 621 }
 622
 623 sub get_macro_for_GT{
 624     # argument: <char>
 625     # return: TeX macro for GT fonts or undef.
 626     my($char)=@_;
 627     my($gt,$GT);
 628     foreach (@GT){
 629         if($gt=&get_char_attribute($char,$_)){
 630             m/gt\-pj\-(\d+)/ and $GT=$1;
 631             last;
 632         }
 633     }
 634     if($gt){
 635         return "{\\fontencoding{OT1}\\fontfamily{"
 636             .sprintf("gt%02d",$GT)
 637             ."}\\selectfont\\char".($gt|0x8080)."}";
 638     }else{
 639         return undef;
 640     }
 641 }
 642
 643 sub get_macro_for_HZK{
 644     # argument: <char>
 645     # return: TeX macro for Hanziku fonts or undef.
 646     my($char)=@_;
 647     my($hzk,$HZK);
 648     foreach (@HZK){
 649         if($hzk=&get_char_attribute($char,$_)){
 650             m/hanziku\-(\d+)/ and $HZK=$1;
 651             last;
 652         }
 653     }
 654     if($hzk){
 655         return "{\\fontencoding{OT1}\\fontfamily{"
 656             .sprintf("hzk%02d",$HZK)
 657             ."}\\selectfont\\char".$hzk."}";
 658     }else{
 659         return undef;
 660     }
 661 }
 662
 663 sub get_macro_for_CDP{
 664     # argument: <char>
 665     # return: TeX macro for CDP fonts or undef.
 666     my($char)=@_;
 667     my($cdp,$ucs);
 668     foreach (@CDP){
 669         if($cdp=&get_char_attribute($char,$_)){
 670             last;
 671         }
 672     }
 673     if($cdp){
 674         $ucs=&get_char_attribute(&get_chars_matching("=big5-pua",$cdp),"=ucs");
 675         if($ucs){
 676             return "{\\fontencoding{OT1}\\fontfamily{cdp}\\selectfont\\char$ucs}";
 677         }else{
 678             print STDERR "This should not happen.\n";
 679             print STDERR "ucs code point of CDP: $cdp not found.\n";
 680         }
 681     }else{
 682         return undef;
 683     }
 684 }