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_preserve_spaces
  12             $opt_help $usage
  13             $in_cs $out_cs $i @chars
  14             $char $char_id $out_char
  15             $char_unified @chars_unified
  16             $ids $ids_argc %ids $idsdb $geta
  17             $idsdata_file $ids_start $font_start
  18             @CDP @HZK @GT
  19             );
  20 use Getopt::Long;
  21 use utf8;
  22 use Fcntl ':flock';
  23 use Chise_utils ':all';
  24 require 5.008;
  25
  26 ### Options ###
  27
  28 $opt_order{'UniMulti'}='jGcgk';
  29 #$opt_order{'UniMulti'}='jGcgkHC';
  30 $opt_order{'UniCNS'}='c';
  31 $opt_order{'UniGB'}='g';
  32 $opt_order{'UniJIS'}='j';
  33 $opt_order{'UniKS'}='k';
  34
  35 $opt_allow_unify=1; # 1=true, 0=false.
  36 $opt_protrude=0;# 1=true, 0=false.
  37
  38 # currently does not work.
  39 $opt_use_kage_for_Ext_B=0;# 1=true, 0=false.
  40
  41 ### End ###
  42
  43 $/="";
  44
  45 my $strictly_forbidden_after = '「【『［（〈‘‛“‟（〔｛《{\[\(〖｛｢';
  46
  47 my $forbidden_after = "\x{0000}";
  48
  49 # ￥¥＄$〒♯＃#￠¢￡£＠@§
  50 my $slightly_forbidden_after = '￥¥＄$〒♯＃#￠¢￡£＠@§';
  51
  52 # $strictly_forbidden_before
  53 # All these characters are allowed to protrude
  54 # in the right margin
  55 my $strictly_forbidden_before=
  56     '!,.:;?、。！，．：；？。\)\]}’”〉》」』】〕〗）］｝」';
  57
  58 my $forbidden_before
  59     = 'ー々ぁぃぅぇぉゃゅょっゎァィゥェォャュョッヮヵヶ';
  60
  61 my $slightly_forbidden_before
  62     = '\x{000a}\#\-‐−‰′″℃゛゜ゝゞヽヾ＂％－ﾞﾟ';
  63
  64 #my $asian = '\x{1100}-\x{11FF}\x{2E80}-\x{D7AF}\x{F900}-\x{FAFF}\x{FE30}-\x{FE4F}\x{FF00}-\x{FFFFFF}'; # need to be fixed.
  65 my $asian = '\x{2E80}-\x{312f}\x{3190}-\x{ABFF}\x{F900}-\x{FAFF}\x{FE30}-\x{FE4F}\x{FF00}-\x{FFFFFF}'; # need to be fixed.
  66
  67 my $space = '\x{0020}\x{0009}\x{000A}\x{000C}\x{000D}';
  68
  69 my $hangul='\x{1100}-\x{11ff}\x{3130}-\x{318f}\x{AC00}-\x{D7AF}';
  70
  71 my %tex_meta=('#'=>'\#',
  72               '$'=>'\\textdollar{}',
  73               '%'=>'\%',
  74               '&'=>'\&',
  75               '{'=>'\\textbraceleft{}',
  76               '}'=>'\\textbraceright{}',
  77               '\\'=>'\\textbackslash{}',
  78               '_'=>'\\textunderscore',
  79              );
  80
  81 my $tex_meta_re=join('|',map {quotemeta($_)} keys %tex_meta);
  82
  83 &GetOptions("in=s"=>\$opt_in_cs,
  84             "out=s"=>\$opt_out_cs,
  85             "help",\$opt_help);
  86
  87 $usage=<<EOF;
  88 Usage: $0 -i <input coding system> -o <cmap encoding>
  89     input coding system:
  90       Utf8mcs, Utf8cns, Utf8gb, Utf8jis, Utf8ks, Utf8big5
  91     cmap encoding:
  92       UniCNS, UniGB, UniJIS, UniKS, UniMulti
  93 EOF
  94
  95 if($opt_in_cs or $opt_out_cs){
  96     $in_cs=$opt_in_cs;
  97     $out_cs=$opt_out_cs;
  98 }elsif(@ARGV==0){
  99     ($in_cs
 100      ,$out_cs
 101      ,$opt_preserve_spaces)
 102         =($0=~/(Utf8mcs|Utf8cns|Utf8gb|Utf8jis|Utf8ks|Utf8big5)
 103           To (UniCNS|UniGB|UniJIS|UniKS|UniMulti)
 104           (Sp)?/ox);
 105 }
 106
 107 # $in_cs:
 108 #   Utf8mcs,Utf8cns,Utf8gb,Utf8jis,Utf8ks,Utf8big5
 109 # $out_cs:
 110 #   UniCNS,UniGB,UniJIS,UniKS,UniMulti
 111
 112 $in_cs=~s/Utf8/ucs\@/;
 113
 114 if($opt_help
 115    or not defined($in_cs)
 116    or not defined($out_cs)){
 117     print $usage;
 118     exit 1;
 119 }
 120
 121 $omegadb_path=~s!/$!!;
 122
 123 $idsdata_file="$omegadb_path/idsdata.pl";
 124 $ids_start=0x00;
 125 $font_start=0;
 126
 127 if(-e $idsdata_file){
 128     open(IDSDATA,"+<:utf8",$idsdata_file) or die;
 129     flock(IDSDATA,LOCK_EX);
 130     seek(IDSDATA,0,0);
 131     while(<IDSDATA>){
 132         eval $_;
 133     }
 134     seek(IDSDATA,0,0);
 135 #         require $idsdata_file;
 136 }else{
 137     open(IDSDATA,">:utf8",$idsdata_file) or die;
 138     flock(IDSDATA,LOCK_EX);
 139     seek(IDSDATA,0,0);
 140 }
 141
 142 $ids_argc=0;
 143 $ids="";
 144
 145 #$geta=pack("U",0x3013);
 146 $geta=pack("U",0xfffd);
 147
 148 @GT=(#"=gt","=gt-k",
 149      "=gt-pj-1","=gt-pj-2","=gt-pj-3","=gt-pj-4","=gt-pj-5",
 150      "=gt-pj-6","=gt-pj-7","=gt-pj-8","=gt-pj-9","=gt-pj-10",
 151      "=gt-pj-11"
 152      #,"=gt-pj-k1","=gt-pj-k2"
 153      );
 154 @HZK=("=hanziku-1","=hanziku-2","=hanziku-3","=hanziku-4",
 155       "=hanziku-5","=hanziku-6","=hanziku-7","=hanziku-8",
 156       "=hanziku-9","=hanziku-10","=hanziku-11","=hanziku-12");
 157 @CDP=("=big5-cdp");
 158
 159 %order_map=('c'=>'UniCNS',
 160             'g'=>'UniGB',
 161             'j'=>'UniJIS',
 162             'k'=>'UniKS',
 163             'G'=>'GT',
 164             'H'=>'HZK',
 165             'C'=>'CDP',
 166             );
 167
 168 foreach $out_cs ('UniCNS','UniGB','UniJIS','UniKS','UniMulti'){
 169     if(defined($opt_order{$out_cs})){
 170         if($opt_order{$out_cs}=~/^[cgjkGHC]+$/){
 171             @{$order{$out_cs}}=map {$order_map{$_}}
 172             (split(//,$opt_order{$out_cs}));
 173         }else{
 174             print STDERR "Invalid order for $out_cs!\n";
 175             exit 1;
 176         }
 177     }
 178 }
 179
 180 while(<>){
 181     utf8::decode($_);
 182
 183     print '{\relax{}';
 184
 185     if($in_cs ne 'ucs@mcs'){
 186         s/(.)/&get_char_in_mcs($1,$in_cs)/ge;
 187     }
 188     s/(amp.+?;)/&de_tex_er($1)/ge;
 189 #    s/(&.+?;)/&de_tex_er($1)/ge;
 190     s/([$asian])[$space]+/$1/go unless($opt_preserve_spaces);
 191     s/[$space]+([$asian])/$1/go unless($opt_preserve_spaces);
 192     s/([$asian])\s*([^$asian$space])/$1 $2/go;
 193     s/([^$asian$idc])\s*([$asian])/$1 $2/go;
 194     s/\-\-\-/pack("U",0x2014)/geo;# EM DASH
 195     s/\-\-/pack("U",0x2013)/geo;# EN DASH
 196 #    s/\`\`/pack("U",0x201f)/geo;# DOUBLE HIGH-REVERSED-9 QUOTATION MARK
 197     s/\`\`/pack("U",0x201c)/geo;# _TeX compatible_
 198 #    s/\`/pack("U",0x201b)/geo;# SINGLE HIGH-REVERSED-9 QUOTATION MARK
 199     s/\`/pack("U",0x2018)/geo;# _TeX compatible_
 200     s/\'\'/pack("U",0x201d)/geo;# RIGHT DOUBLE QUOTATION MARK
 201     s/\'/pack("U",0x2019)/geo;# RIGHT DOUBLE QUOTATION MARK
 202     @chars=split(//);
 203
 204   CHAR:
 205     for($i=0;$i<=$#chars;$i++){
 206         $char=$chars[$i];
 207         if($char_id=&get_char_attribute($char,'=ucs@unicode')){
 208             $char=pack("U",$char_id);
 209         }else{
 210             $char_id=unpack("U",$char);
 211         }
 212
 213         if($char_id<=0x20){
 214             # add quarter space between asian and non-asian.
 215             if($i>0 and $i<$#chars){
 216                 if(($chars[$i-1]=~/[$asian]/
 217                     and $chars[$i+1]=~/[^$asian]/
 218                       and $chars[$i+1]=~/[$strictly_forbidden_before$slightly_forbidden_before]/)
 219                    or($chars[$i-1]=~/[^$asian]/
 220                       and $chars[$i+1]=~/[$asian]/
 221                       and $chars[$i-1]=~/[$strictly_forbidden_after$slightly_forbidden_after]/)
 222                    ){
 223                     print '';
 224                 }elsif(($chars[$i-1]=~/[$asian]/
 225                         and $chars[$i+1]=~/[^$asian]/)){
 226                     print '\unskip\kern.25ex';
 227                 }elsif($chars[$i-1]=~/[^$asian]/
 228                        and $chars[$i+1]=~/[$asian]/){
 229                     print '\kern.25ex';
 230                 }else{
 231                     print ' ';
 232                 }
 233             }else{
 234                 print ' ';
 235             }
 236             next CHAR;
 237         }elsif(($char_id>0x20 and $char_id<=0x021f)
 238                # Basic Latin
 239                # Latin-1 Supplement
 240                # Latin Extended-A
 241                # Latin Extended-B (not all)
 242                or($char_id>=0x0250 and $char_id<=0x02af)
 243                # IPA Extensions
 244                or($char_id>=0x0300 and $char_id<=0x033f)
 245                or($char_id>=0x0360 and $char_id<=0x036f)
 246                # Combining Diacritical Marks
 247                or($char_id>=0x1e00 and $char_id<=0x1eff)
 248                # Latin Extended Additional
 249                or($char_id>=0x0370 and $char_id<=0x03ff)
 250                # Greek and Coptic
 251                or($char_id>=0x0400 and $char_id<=0x04ff)
 252                # Cyrillic
 253                or($char_id>=0x0530 and $char_id<=0x058f)
 254                # Armenian
 255                or($char_id>=0x2010 and $char_id<=0x2046)
 256                # General Punctuation (partial)
 257                ){
 258             print &latin_parse();
 259             next CHAR;
 260         }elsif($char_id>=0x2ff0 and $char_id<=0x2fff){
 261             # Ideographic Description Characters
 262             print &ids_parse();
 263             next CHAR;
 264         }else{
 265             if(($out_char=&get_output_char($char,$out_cs))){
 266                 print $out_char,&add_break($i);
 267             }else{
 268                 if($opt_allow_unify){
 269                     @chars_unified=&get_chars_unified($char);
 270                     if(@chars_unified>0){
 271                         foreach $char_unified (@chars_unified){
 272                             if(($out_char
 273                                 =&get_output_char($char_unified,$out_cs))){
 274                                 print $out_char,&add_break($i);
 275                                 next CHAR;
 276                             }
 277                         }
 278                     }
 279                 }
 280                 if($opt_use_kage_for_Ext_B){
 281                     if($char_id >= 0x20000 && $char_id <=0x2a6df){
 282                         # CJK Unified Ideographs Extension B
 283                         if(not defined($ids{$char}) and $ids{$char}[1]>=0){
 284                             $ids{$char}[0]=$font_start;
 285                             $ids{$char}[1]=$ids_start;
 286                             $ids_start++;
 287                             if($ids_start>255){
 288                                 $ids_start=0;
 289                                 $font_start++;
 290                             }
 291                         }
 292                         print "{\\fontencoding{OT1}\\fontfamily{",
 293                             sprintf("chise%03d",$ids{$char}[0]) ,
 294                                 "}\\selectfont\\char$ids{$char}[1]}",&add_break($i);
 295                         next CHAR;
 296                     }
 297                 }
 298                 if($ids=&get_ids($char)){
 299                     print &get_macro_for_ids($ids),&add_break($i);
 300                 }else{
 301                     print '\rule{1ex}{1ex}',&add_break($i);
 302                 }
 303             }
 304         }
 305     }
 306     print '}';
 307 }
 308
 309 print IDSDATA 'use utf8;',"\n";
 310 foreach $ids (keys %ids){
 311     print IDSDATA '$ids{\'',$ids,'\'}='
 312     ,'[',join ",",@{$ids{$ids}},"];\n";
 313 }
 314 print IDSDATA '$font_start=',$font_start,";\n";
 315 print IDSDATA '$ids_start=',$ids_start,";\n";
 316 print IDSDATA "1;";
 317 flock(IDSDATA,LOCK_UN);
 318
 319 exit 0;
 320
 321 sub de_tex_er{
 322     my($er)=@_;
 323     my($prefix,$suffix);
 324     my($output_char,$atr,$value);
 325     $er=~/^(amp)(.*)(;)$/
 326         and $prefix=$1,$er=$2,$suffix=$3;
 327     $prefix or $prefix="",$suffix or $suffix="";
 328     if($er=~/^U[\+|\-]([a-fA-F\d]+)/){
 329         $output_char=pack("U",hex($1));
 330     }elsif($er=~/^(?:I\-)?($er_prefix_re)\-?([0-9a-fA-F]+)$/){
 331         ($atr,$value)=($1,$2);
 332         unless($er_alias{$atr}=~/daikanwa|gt/){
 333             $value=hex($value);
 334         }
 335         ($output_char)=&get_chars_matching($er_alias{$atr},$value);
 336     }
 337     if($output_char){
 338         return $output_char;
 339     }else{
 340         return $prefix.$er.$suffix;
 341     }
 342 }
 343
 344 sub add_break{
 345     my($i)=@_;
 346
 347     if($i<($#chars-1)){
 348         if(($chars[$i+1]=~m/[$strictly_forbidden_before]/o)
 349            and($chars[$i+2]=~m/[$strictly_forbidden_before]/o)){
 350             return "\\CJKunbreakablekernone{}";
 351         }elsif($opt_protrude){
 352             if(($chars[$i+1]=~m/[$strictly_forbidden_before]/o)
 353                and($chars[$i+2]=~m/[^$strictly_forbidden_before]/o)){
 354                 return "\\CJKunbreakablekernone \\CJKprotrude{}";
 355             }
 356         }
 357     }
 358     if(($i<$#chars)
 359        and($chars[$i+1]eq" ")){
 360        # preserve space.
 361         return "";
 362     }
 363     if(($i<$#chars)
 364        and($chars[$i+1]=~m/[$strictly_forbidden_before]/o)){
 365         return "\\CJKunbreakablekernone{}";
 366     }
 367     if($chars[$i]=~m/[$strictly_forbidden_after]/o){
 368         return "\\CJKunbreakablekernone{}";
 369     }
 370     if(($i<$#chars)
 371        and($chars[$i+1]=~m/[$forbidden_before]/o)){
 372         return "\\CJKunbreakablekerntwo{}";
 373
 374     }
 375     if($chars[$i]=~m/[$forbidden_after]/o){
 376         return "\\CJKunbreakablekerntwo{}";
 377     }
 378     if(($i<$#chars)
 379        and($chars[$i+1]=~m/[$slightly_forbidden_before]/o)){
 380         return "\\CJKunbreakablekernthree{}";
 381     }
 382     if($chars[$i]=~m/[$slightly_forbidden_after]/o){
 383         return "\\CJKunbreakablekernthree{}";
 384     }
 385     if($chars[$i]=~m/[$asian]/o){
 386         return "\\CJKbreakablekern{}";
 387     }
 388     if(($i<$#chars)and($chars[$i+1]=~m/[$asian]/o)){
 389         return "\\CJKbreakablekern{}";
 390     }
 391 }
 392
 393 sub latin_parse{
 394     # arguments: none
 395     # return: string for output with TeX macro.
 396     my($char_id);
 397     my $out_str="";
 398     while($i<=$#chars){
 399         $char_id=unpack("U",$chars[$i]);
 400         if($chars[$i]=~m/($tex_meta_re)/o){
 401             $out_str.=$tex_meta{$chars[$i]};
 402         }elsif(($char_id>0x20 and $char_id<=0x021f)
 403            or($char_id>=0x0250 and $char_id<=0x02af)# IPA Extensions
 404            or($char_id>=0x0300 and $char_id<=0x033f)# Combining Diacritical Marks
 405            or($char_id>=0x0360 and $char_id<=0x036f)
 406            or($char_id>=0x0370 and $char_id<=0x03ff)# Greek and Coptic
 407            or($char_id>=0x0400 and $char_id<=0x04ff)# Cyrillic
 408            or($char_id>=0x0530 and $char_id<=0x058f)# Armenian
 409            ){
 410             $out_str.=$chars[$i];
 411         }elsif($char_id>=0x1e00 and $char_id<=0x1eff){
 412             # Latin Extended Additional
 413             # 0x1e00 -> 0x0600, etc.
 414             $out_str.=pack("U",$char_id-0x1800);
 415         }elsif($char_id>=0x2010 and $char_id<=0x2046){
 416             # General Punctuation (partial)
 417             $out_str.=pack("U",$char_id-0x1000);
 418         }else{
 419             $i--;
 420             last;
 421         }
 422         $i++;
 423     }
 424     return '{\fontencoding{OT1}\fontfamily{omlgc}\selectfont '.$out_str.'}';###UT1?
 425 }
 426
 427 sub ids_parse{
 428     # arguments: none
 429     # return: character for output,
 430     #          TeX macro for ids,
 431     #          or GETA character if ids is invalid.
 432     my($ids,$ids_argc)=&ids_rest("",0,$chars[$i]);
 433
 434     while($ids_argc>0){
 435         # We are in IDS.
 436         $i++;
 437         if($i>$#chars){
 438             print STDERR "IDS parse error: $ids\n";
 439             return $geta;
 440         }
 441
 442         ($ids,$ids_argc)=&ids_rest($ids,$ids_argc,$chars[$i]);
 443         if($ids_argc==0){
 444             if(($char=&get_char_for_ids($ids))
 445                and($out_char=&get_output_char($char,$out_cs))){
 446                 return $out_char;
 447             }else{
 448                 return &get_macro_for_ids($ids);
 449             }
 450         }
 451     }
 452 }
 453
 454 sub ids_rest{
 455     # arguments: <ids>, <rest number of arguments for ids>, <character>
 456     # return: ids and rest number of arguments for ids.
 457     my($ids,$ids_argc,$char)=@_;
 458     my($argc);
 459     $argc=&ids_argc($char);
 460     if($argc){
 461         $ids_argc+=$ids_argc==0?$argc:$argc-1;
 462     }else{
 463         $ids_argc--;
 464     }
 465     $ids.=$char;
 466     return ($ids,$ids_argc);
 467 }
 468
 469 sub get_macro_for_ids{
 470     # argument: <ids>
 471     # return: TeX macro for ids
 472     #          or GETA character if ids is invalid for KAGE.
 473     my($ids)=@_;
 474     # $ids=&normalize_ids($ids,"UniJIS");
 475     return $geta if(($ids!~/[$idc]/)
 476                     or($ids=~/[\x{10000}-]/));
 477                     #irregular for KAGE.
 478     if(not defined($ids{$ids})){
 479         $ids{$ids}[0]=$font_start;
 480         $ids{$ids}[1]=$ids_start;
 481         $ids_start++;
 482     }
 483     if($ids_start>255){
 484         $ids_start=0;
 485         $font_start++;
 486     }
 487     return "{\\fontencoding{OT1}\\fontfamily{"
 488         .sprintf("chise%03d",$ids{$ids}[0])
 489         ."}\\selectfont\\char$ids{$ids}[1]}\\relax{}";
 490 }
 491
 492 sub normalize_ids{
 493     # argument: <ids>, <output coding system>
 494     # return: ids or GETA character if ids is invalid for KAGE.
 495     my($ids,$out_cs)=@_;
 496     $out_cs=~s/Uni(.+)/'ucs@'.lc($1)/e;
 497
 498     my $output_ids="";
 499     my($char,$output_char_id);
 500     while($ids=~m/(.)/g){
 501         $char=$1;
 502         if($char=~/[$idc]/){
 503             $output_ids.=$char;
 504         }elsif($output_char_id=&get_char_attribute($char,"=$out_cs")
 505            or $output_char_id=&get_char_attribute($char,"=ucs")
 506            or $output_char_id=&get_char_attribute($char,"=>$out_cs")
 507            or $output_char_id=&get_char_attribute($char,"=>ucs")
 508            or $output_char_id=&get_char_attribute($char,"=>ucs*")
 509               ){
 510             $output_ids.=pack("U",$output_char_id);
 511         }else{
 512             return $geta;
 513         }
 514     }
 515     return $output_ids;
 516 }
 517
 518 sub get_output_char{
 519     # argument: <char>
 520     # return: character in output coding system or TeX macro or undef.
 521     my($char,$out_cs)=@_;
 522     my($out_char_id,$suffix);
 523     my($gt,$hzk,$cdp);
 524
 525     foreach $out_cs (@{$order{$out_cs}}){
 526         if($out_cs eq 'UniJIS'
 527            and &get_char_attribute($char,"vnd-adobe-cid-unijis-utf16-h")){
 528             if($out_char_id=&get_char_attribute($char,'=ucs@jis')
 529                or $out_char_id=&get_char_attribute($char,'=ucs')
 530                or $out_char_id=&get_char_attribute($char,'=>ucs@jis')
 531                or $out_char_id=&get_char_attribute($char,'=>ucs')
 532                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 533                ){
 534                 return "{\\selectjisfont\\char$out_char_id}";
 535             }
 536         }elsif($out_cs eq 'UniGB'
 537                and &get_char_attribute($char,"vnd-adobe-cid-unigb-ucs2-h")){
 538             if($out_char_id=&get_char_attribute($char,'=ucs@gb')
 539                or $out_char_id=&get_char_attribute($char,'=ucs')
 540                or $out_char_id=&get_char_attribute($char,'=>ucs@gb')
 541                or $out_char_id=&get_char_attribute($char,'=>ucs')
 542                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 543                ){
 544                 return "{\\selectgbsfont\\char$out_char_id}";
 545             }
 546         }elsif($out_cs eq 'UniCNS'
 547                and &get_char_attribute($char,"vnd-adobe-cid-unicns-ucs2-h")){
 548             if($out_char_id=&get_char_attribute($char,'=ucs@cns')
 549                or $out_char_id=&get_char_attribute($char,'=ucs')
 550                or $out_char_id=&get_char_attribute($char,'=>ucs@cns')
 551                or $out_char_id=&get_char_attribute($char,'=>ucs')
 552                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 553                ){
 554                 return "{\\selectcnsfont\\char$out_char_id}";
 555             }
 556         }elsif($out_cs eq 'UniKS'
 557                and &get_char_attribute($char,"vnd-adobe-cid-uniks-ucs2-h")){
 558             if($out_char_id=&get_char_attribute($char,'=ucs@ks')
 559                or $out_char_id=&get_char_attribute($char,'=ucs')
 560                or $out_char_id=&get_char_attribute($char,'=>ucs@ks')
 561                or $out_char_id=&get_char_attribute($char,'=>ucs')
 562                or $out_char_id=&get_char_attribute($char,'=>ucs*')
 563                ){
 564                 return "{\\selectksxfont\\char$out_char_id}";
 565             }
 566         }elsif($out_cs eq 'GT'){
 567             return $gt if($gt=&get_macro_for_GT($char));
 568         }elsif($out_cs eq 'HZK'){
 569             return $hzk if($hzk=&get_macro_for_HZK($char));
 570         }elsif($out_cs eq 'CDP'){
 571             return $cdp if($cdp=&get_macro_for_CDP($char));
 572         }
 573     }
 574     return undef;
 575 }
 576
 577 sub get_ids{
 578     # argument: <character>
 579     # return: ids
 580     my($char)=@_;
 581     my $ids="";
 582 #    $ids=&get_char_attribute($char,"ids-aggregated")
 583 #       or $ids=&get_char_attribute($char,"ids");
 584     $ids=&get_char_attribute($char,"ids-decomposed")
 585         or $ids=&get_char_attribute($char,"ids")
 586           or $ids=&get_char_attribute($char,"ideographic-structure");
 587     $ids=~s/[? ()]//g;
 588     return $ids;
 589 }
 590
 591 sub get_char_for_ids{
 592     # argument: <ideographic description sequence>
 593     # return: char or undef.
 594     my($ids)=@_;
 595     my($output_char);
 596
 597     if(($output_char)=&get_chars_matching("ids",$ids)){
 598         return $output_char;
 599     }else{
 600         return undef;
 601     }
 602 }
 603
 604 sub get_char_in_mcs{
 605     # argument: <char>, <input coding system>
 606     # return:   char in ucs@mcs.
 607     my($char,$in_cs)=@_;
 608     my($output_char);
 609
 610     return $char if($in_cs eq 'ucs@mcs');
 611
 612     if(($output_char)=&get_chars_matching("=$in_cs",unpack("U",$char))){
 613         return $output_char;
 614     }else{
 615         return $char;
 616     }
 617 }
 618
 619 sub get_chars_unified{
 620     my($char)=@_;
 621     my($chars,$ucs,$char_ucs,$char_sub);
 622     my(@chars);
 623
 624     if($chars=&get_char_attribute($char,'->ucs-unified')){
 625         $chars=~s/^\((.*)\)$/$1/;
 626         return (split(/\s*\?/,$chars));
 627     }elsif($ucs=&get_char_attribute($char,'=>ucs*')
 628            or $ucs=&get_char_attribute($char,'=>ucs')){
 629         $char_ucs=pack("U",$ucs);
 630         if($chars=&get_char_attribute($char_ucs,'->ucs-unified')){
 631             $chars=~s/^\((.*)\)$/$1/;
 632             @chars=grep {not /^$char$/}
 633                 (split(/\s*\?/,$chars));
 634             push(@chars,$char_ucs);
 635             return @chars;
 636         }
 637     }elsif($char_sub=&get_char_attribute($char,'<-subsumptive')){
 638         $char_sub=~s/[? ()]//g;
 639         $chars=&get_char_attribute($char_sub,'->subsumptive');
 640         @chars=grep {not /^$char$/} (split(/\s*\?/,$chars));
 641         push(@chars,$char_sub);
 642         return @chars;
 643     }else{
 644         return ();
 645     }
 646 }
 647
 648 sub get_macro_for_GT{
 649     # argument: <char>
 650     # return: TeX macro for GT fonts or undef.
 651     my($char)=@_;
 652     my($gt,$GT);
 653     foreach (@GT){
 654         if($gt=&get_char_attribute($char,$_)){
 655             m/gt\-pj\-(\d+)/ and $GT=$1;
 656             last;
 657         }
 658     }
 659     if($gt){
 660         return "{\\fontencoding{OT1}\\fontfamily{"
 661             .sprintf("gt%02d",$GT)
 662             ."}\\selectfont\\char".($gt|0x8080)."}";
 663     }else{
 664         return undef;
 665     }
 666 }
 667
 668 sub get_macro_for_HZK{
 669     # argument: <char>
 670     # return: TeX macro for Hanziku fonts or undef.
 671     my($char)=@_;
 672     my($hzk,$HZK);
 673     foreach (@HZK){
 674         if($hzk=&get_char_attribute($char,$_)){
 675             m/hanziku\-(\d+)/ and $HZK=$1;
 676             last;
 677         }
 678     }
 679     if($hzk){
 680         return "{\\fontencoding{OT1}\\fontfamily{"
 681             .sprintf("hzk%02d",$HZK)
 682             ."}\\selectfont\\char".$hzk."}";
 683     }else{
 684         return undef;
 685     }
 686 }
 687
 688 sub get_macro_for_CDP{
 689     # argument: <char>
 690     # return: TeX macro for CDP fonts or undef.
 691     my($char)=@_;
 692     my($cdp,$ucs);
 693     foreach (@CDP){
 694         if($cdp=&get_char_attribute($char,$_)){
 695             last;
 696         }
 697     }
 698     if($cdp){
 699         $ucs=&get_char_attribute(&get_chars_matching("=big5-pua",$cdp),"=ucs");
 700         if($ucs){
 701             return "{\\fontencoding{OT1}\\fontfamily{cdp}\\selectfont\\char$ucs}";
 702         }else{
 703             print STDERR "This should not happen.\n";
 704             print STDERR "ucs code point of CDP: $cdp not found.\n";
 705         }
 706     }else{
 707         return undef;
 708     }
 709 }