1/********************************************************************** 2 3 encoding.c - 4 5 $Author: nagachika $ 6 created at: Thu May 24 17:23:27 JST 2007 7 8 Copyright (C) 2007 Yukihiro Matsumoto 9 10**********************************************************************/ 11 12#include "ruby/ruby.h" 13#include "ruby/encoding.h" 14#include "internal.h" 15#include "regenc.h" 16#include <ctype.h> 17#ifndef NO_LOCALE_CHARMAP 18#ifdef __CYGWIN__ 19#include <windows.h> 20#endif 21#ifdef HAVE_LANGINFO_H 22#include <langinfo.h> 23#endif 24#endif 25#include "ruby/util.h" 26 27#if defined __GNUC__ && __GNUC__ >= 4 28#pragma GCC visibility push(default) 29int rb_enc_register(const char *name, rb_encoding *encoding); 30void rb_enc_set_base(const char *name, const char *orig); 31void rb_encdb_declare(const char *name); 32int rb_encdb_replicate(const char *name, const char *orig); 33int rb_encdb_dummy(const char *name); 34int rb_encdb_alias(const char *alias, const char *orig); 35void rb_encdb_set_unicode(int index); 36#pragma GCC visibility pop 37#endif 38 39static ID id_encoding; 40VALUE rb_cEncoding; 41static VALUE rb_encoding_list; 42 43struct rb_encoding_entry { 44 const char *name; 45 rb_encoding *enc; 46 rb_encoding *base; 47}; 48 49static struct { 50 struct rb_encoding_entry *list; 51 int count; 52 int size; 53 st_table *names; 54} enc_table; 55 56void rb_enc_init(void); 57 58#define ENCODING_COUNT ENCINDEX_BUILTIN_MAX 59#define UNSPECIFIED_ENCODING INT_MAX 60 61#define ENCODING_NAMELEN_MAX 63 62#define valid_encoding_name_p(name) ((name) && strlen(name) <= ENCODING_NAMELEN_MAX) 63 64#define enc_autoload_p(enc) (!rb_enc_mbmaxlen(enc)) 65 66static int load_encoding(const char *name); 67 68static size_t 69enc_memsize(const void *p) 70{ 71 return 0; 72} 73 74static const rb_data_type_t encoding_data_type = { 75 "encoding", 76 {0, 0, enc_memsize,}, 77}; 78 79#define is_data_encoding(obj) (RTYPEDDATA_P(obj) && RTYPEDDATA_TYPE(obj) == &encoding_data_type) 80 81static VALUE 82enc_new(rb_encoding *encoding) 83{ 84 return TypedData_Wrap_Struct(rb_cEncoding, &encoding_data_type, encoding); 85} 86 87static VALUE 88rb_enc_from_encoding_index(int idx) 89{ 90 VALUE list, enc; 91 92 if (!(list = rb_encoding_list)) { 93 rb_bug("rb_enc_from_encoding_index(%d): no rb_encoding_list", idx); 94 } 95 enc = rb_ary_entry(list, idx); 96 if (NIL_P(enc)) { 97 rb_bug("rb_enc_from_encoding_index(%d): not created yet", idx); 98 } 99 return enc; 100} 101 102VALUE 103rb_enc_from_encoding(rb_encoding *encoding) 104{ 105 int idx; 106 if (!encoding) return Qnil; 107 idx = ENC_TO_ENCINDEX(encoding); 108 return rb_enc_from_encoding_index(idx); 109} 110 111static int enc_autoload(rb_encoding *); 112 113static int 114check_encoding(rb_encoding *enc) 115{ 116 int index = rb_enc_to_index(enc); 117 if (rb_enc_from_index(index) != enc) 118 return -1; 119 if (enc_autoload_p(enc)) { 120 index = enc_autoload(enc); 121 } 122 return index; 123} 124 125static int 126enc_check_encoding(VALUE obj) 127{ 128 if (SPECIAL_CONST_P(obj) || !rb_typeddata_is_kind_of(obj, &encoding_data_type)) { 129 return -1; 130 } 131 return check_encoding(RDATA(obj)->data); 132} 133 134static int 135must_encoding(VALUE enc) 136{ 137 int index = enc_check_encoding(enc); 138 if (index < 0) { 139 rb_raise(rb_eTypeError, "wrong argument type %s (expected Encoding)", 140 rb_obj_classname(enc)); 141 } 142 return index; 143} 144 145int 146rb_to_encoding_index(VALUE enc) 147{ 148 int idx; 149 150 idx = enc_check_encoding(enc); 151 if (idx >= 0) { 152 return idx; 153 } 154 else if (NIL_P(enc = rb_check_string_type(enc))) { 155 return -1; 156 } 157 if (!rb_enc_asciicompat(rb_enc_get(enc))) { 158 return -1; 159 } 160 return rb_enc_find_index(StringValueCStr(enc)); 161} 162 163/* Returns encoding index or UNSPECIFIED_ENCODING */ 164static int 165str_find_encindex(VALUE enc) 166{ 167 int idx; 168 169 StringValue(enc); 170 if (!rb_enc_asciicompat(rb_enc_get(enc))) { 171 rb_raise(rb_eArgError, "invalid name encoding (non ASCII)"); 172 } 173 idx = rb_enc_find_index(StringValueCStr(enc)); 174 return idx; 175} 176 177static int 178str_to_encindex(VALUE enc) 179{ 180 int idx = str_find_encindex(enc); 181 if (idx < 0) { 182 rb_raise(rb_eArgError, "unknown encoding name - %s", RSTRING_PTR(enc)); 183 } 184 return idx; 185} 186 187static rb_encoding * 188str_to_encoding(VALUE enc) 189{ 190 return rb_enc_from_index(str_to_encindex(enc)); 191} 192 193rb_encoding * 194rb_to_encoding(VALUE enc) 195{ 196 if (enc_check_encoding(enc) >= 0) return RDATA(enc)->data; 197 return str_to_encoding(enc); 198} 199 200rb_encoding * 201rb_find_encoding(VALUE enc) 202{ 203 int idx; 204 if (enc_check_encoding(enc) >= 0) return RDATA(enc)->data; 205 idx = str_find_encindex(enc); 206 if (idx < 0) return NULL; 207 return rb_enc_from_index(idx); 208} 209 210void 211rb_gc_mark_encodings(void) 212{ 213} 214 215static int 216enc_table_expand(int newsize) 217{ 218 struct rb_encoding_entry *ent; 219 int count = newsize; 220 221 if (enc_table.size >= newsize) return newsize; 222 newsize = (newsize + 7) / 8 * 8; 223 ent = realloc(enc_table.list, sizeof(*enc_table.list) * newsize); 224 if (!ent) return -1; 225 memset(ent + enc_table.size, 0, sizeof(*ent)*(newsize - enc_table.size)); 226 enc_table.list = ent; 227 enc_table.size = newsize; 228 return count; 229} 230 231static int 232enc_register_at(int index, const char *name, rb_encoding *encoding) 233{ 234 struct rb_encoding_entry *ent = &enc_table.list[index]; 235 VALUE list; 236 237 if (!valid_encoding_name_p(name)) return -1; 238 if (!ent->name) { 239 ent->name = name = strdup(name); 240 } 241 else if (STRCASECMP(name, ent->name)) { 242 return -1; 243 } 244 if (!ent->enc) { 245 ent->enc = xmalloc(sizeof(rb_encoding)); 246 } 247 if (encoding) { 248 *ent->enc = *encoding; 249 } 250 else { 251 memset(ent->enc, 0, sizeof(*ent->enc)); 252 } 253 encoding = ent->enc; 254 encoding->name = name; 255 encoding->ruby_encoding_index = index; 256 st_insert(enc_table.names, (st_data_t)name, (st_data_t)index); 257 list = rb_encoding_list; 258 if (list && NIL_P(rb_ary_entry(list, index))) { 259 /* initialize encoding data */ 260 rb_ary_store(list, index, enc_new(encoding)); 261 } 262 return index; 263} 264 265static int 266enc_register(const char *name, rb_encoding *encoding) 267{ 268 int index = enc_table.count; 269 270 if ((index = enc_table_expand(index + 1)) < 0) return -1; 271 enc_table.count = index; 272 return enc_register_at(index - 1, name, encoding); 273} 274 275static void set_encoding_const(const char *, rb_encoding *); 276int rb_enc_registered(const char *name); 277 278int 279rb_enc_register(const char *name, rb_encoding *encoding) 280{ 281 int index = rb_enc_registered(name); 282 283 if (index >= 0) { 284 rb_encoding *oldenc = rb_enc_from_index(index); 285 if (STRCASECMP(name, rb_enc_name(oldenc))) { 286 index = enc_register(name, encoding); 287 } 288 else if (enc_autoload_p(oldenc) || !ENC_DUMMY_P(oldenc)) { 289 enc_register_at(index, name, encoding); 290 } 291 else { 292 rb_raise(rb_eArgError, "encoding %s is already registered", name); 293 } 294 } 295 else { 296 index = enc_register(name, encoding); 297 set_encoding_const(name, rb_enc_from_index(index)); 298 } 299 return index; 300} 301 302void 303rb_encdb_declare(const char *name) 304{ 305 int idx = rb_enc_registered(name); 306 if (idx < 0) { 307 idx = enc_register(name, 0); 308 } 309 set_encoding_const(name, rb_enc_from_index(idx)); 310} 311 312static void 313enc_check_duplication(const char *name) 314{ 315 if (rb_enc_registered(name) >= 0) { 316 rb_raise(rb_eArgError, "encoding %s is already registered", name); 317 } 318} 319 320static rb_encoding* 321set_base_encoding(int index, rb_encoding *base) 322{ 323 rb_encoding *enc = enc_table.list[index].enc; 324 325 enc_table.list[index].base = base; 326 if (rb_enc_dummy_p(base)) ENC_SET_DUMMY(enc); 327 return enc; 328} 329 330/* for encdb.h 331 * Set base encoding for encodings which are not replicas 332 * but not in their own files. 333 */ 334void 335rb_enc_set_base(const char *name, const char *orig) 336{ 337 int idx = rb_enc_registered(name); 338 int origidx = rb_enc_registered(orig); 339 set_base_encoding(idx, rb_enc_from_index(origidx)); 340} 341 342int 343rb_enc_replicate(const char *name, rb_encoding *encoding) 344{ 345 int idx; 346 347 enc_check_duplication(name); 348 idx = enc_register(name, encoding); 349 set_base_encoding(idx, encoding); 350 set_encoding_const(name, rb_enc_from_index(idx)); 351 return idx; 352} 353 354/* 355 * call-seq: 356 * enc.replicate(name) -> encoding 357 * 358 * Returns a replicated encoding of _enc_ whose name is _name_. 359 * The new encoding should have the same byte structure of _enc_. 360 * If _name_ is used by another encoding, raise ArgumentError. 361 * 362 */ 363static VALUE 364enc_replicate(VALUE encoding, VALUE name) 365{ 366 return rb_enc_from_encoding_index( 367 rb_enc_replicate(StringValueCStr(name), 368 rb_to_encoding(encoding))); 369} 370 371static int 372enc_replicate_with_index(const char *name, rb_encoding *origenc, int idx) 373{ 374 if (idx < 0) { 375 idx = enc_register(name, origenc); 376 } 377 else { 378 idx = enc_register_at(idx, name, origenc); 379 } 380 if (idx >= 0) { 381 set_base_encoding(idx, origenc); 382 set_encoding_const(name, rb_enc_from_index(idx)); 383 } 384 return idx; 385} 386 387int 388rb_encdb_replicate(const char *name, const char *orig) 389{ 390 int origidx = rb_enc_registered(orig); 391 int idx = rb_enc_registered(name); 392 393 if (origidx < 0) { 394 origidx = enc_register(orig, 0); 395 } 396 return enc_replicate_with_index(name, rb_enc_from_index(origidx), idx); 397} 398 399int 400rb_define_dummy_encoding(const char *name) 401{ 402 int index = rb_enc_replicate(name, rb_ascii8bit_encoding()); 403 rb_encoding *enc = enc_table.list[index].enc; 404 405 ENC_SET_DUMMY(enc); 406 return index; 407} 408 409int 410rb_encdb_dummy(const char *name) 411{ 412 int index = enc_replicate_with_index(name, rb_ascii8bit_encoding(), 413 rb_enc_registered(name)); 414 rb_encoding *enc = enc_table.list[index].enc; 415 416 ENC_SET_DUMMY(enc); 417 return index; 418} 419 420/* 421 * call-seq: 422 * enc.dummy? -> true or false 423 * 424 * Returns true for dummy encodings. 425 * A dummy encoding is an encoding for which character handling is not properly 426 * implemented. 427 * It is used for stateful encodings. 428 * 429 * Encoding::ISO_2022_JP.dummy? #=> true 430 * Encoding::UTF_8.dummy? #=> false 431 * 432 */ 433static VALUE 434enc_dummy_p(VALUE enc) 435{ 436 return ENC_DUMMY_P(enc_table.list[must_encoding(enc)].enc) ? Qtrue : Qfalse; 437} 438 439/* 440 * call-seq: 441 * enc.ascii_compatible? -> true or false 442 * 443 * Returns whether ASCII-compatible or not. 444 * 445 * Encoding::UTF_8.ascii_compatible? #=> true 446 * Encoding::UTF_16BE.ascii_compatible? #=> false 447 * 448 */ 449static VALUE 450enc_ascii_compatible_p(VALUE enc) 451{ 452 return rb_enc_asciicompat(enc_table.list[must_encoding(enc)].enc) ? Qtrue : Qfalse; 453} 454 455/* 456 * Returns 1 when the encoding is Unicode series other than UTF-7 else 0. 457 */ 458int 459rb_enc_unicode_p(rb_encoding *enc) 460{ 461 return ONIGENC_IS_UNICODE(enc); 462} 463 464static st_data_t 465enc_dup_name(st_data_t name) 466{ 467 return (st_data_t)strdup((const char *)name); 468} 469 470/* 471 * Returns copied alias name when the key is added for st_table, 472 * else returns NULL. 473 */ 474static int 475enc_alias_internal(const char *alias, int idx) 476{ 477 return st_insert2(enc_table.names, (st_data_t)alias, (st_data_t)idx, 478 enc_dup_name); 479} 480 481static int 482enc_alias(const char *alias, int idx) 483{ 484 if (!valid_encoding_name_p(alias)) return -1; 485 if (!enc_alias_internal(alias, idx)) 486 set_encoding_const(alias, rb_enc_from_index(idx)); 487 return idx; 488} 489 490int 491rb_enc_alias(const char *alias, const char *orig) 492{ 493 int idx; 494 495 enc_check_duplication(alias); 496 if (!enc_table.list) { 497 rb_enc_init(); 498 } 499 if ((idx = rb_enc_find_index(orig)) < 0) { 500 return -1; 501 } 502 return enc_alias(alias, idx); 503} 504 505int 506rb_encdb_alias(const char *alias, const char *orig) 507{ 508 int idx = rb_enc_registered(orig); 509 510 if (idx < 0) { 511 idx = enc_register(orig, 0); 512 } 513 return enc_alias(alias, idx); 514} 515 516void 517rb_encdb_set_unicode(int index) 518{ 519 rb_enc_from_index(index)->flags |= ONIGENC_FLAG_UNICODE; 520} 521 522enum { 523 ENCINDEX_ASCII, 524 ENCINDEX_UTF_8, 525 ENCINDEX_US_ASCII, 526 ENCINDEX_BUILTIN_MAX 527}; 528 529extern rb_encoding OnigEncodingUTF_8; 530extern rb_encoding OnigEncodingUS_ASCII; 531 532void 533rb_enc_init(void) 534{ 535 enc_table_expand(ENCODING_COUNT + 1); 536 if (!enc_table.names) { 537 enc_table.names = st_init_strcasetable(); 538 } 539#define ENC_REGISTER(enc) enc_register_at(ENCINDEX_##enc, rb_enc_name(&OnigEncoding##enc), &OnigEncoding##enc) 540 ENC_REGISTER(ASCII); 541 ENC_REGISTER(UTF_8); 542 ENC_REGISTER(US_ASCII); 543#undef ENC_REGISTER 544 enc_table.count = ENCINDEX_BUILTIN_MAX; 545} 546 547rb_encoding * 548rb_enc_from_index(int index) 549{ 550 if (!enc_table.list) { 551 rb_enc_init(); 552 } 553 if (index < 0 || enc_table.count <= index) { 554 return 0; 555 } 556 return enc_table.list[index].enc; 557} 558 559int 560rb_enc_registered(const char *name) 561{ 562 st_data_t idx = 0; 563 564 if (!name) return -1; 565 if (!enc_table.list) return -1; 566 if (st_lookup(enc_table.names, (st_data_t)name, &idx)) { 567 return (int)idx; 568 } 569 return -1; 570} 571 572static VALUE 573require_enc(VALUE enclib) 574{ 575 int safe = rb_safe_level(); 576 return rb_require_safe(enclib, safe > 3 ? 3 : safe); 577} 578 579static int 580load_encoding(const char *name) 581{ 582 VALUE enclib = rb_sprintf("enc/%s.so", name); 583 VALUE verbose = ruby_verbose; 584 VALUE debug = ruby_debug; 585 VALUE errinfo; 586 VALUE loaded; 587 char *s = RSTRING_PTR(enclib) + 4, *e = RSTRING_END(enclib) - 3; 588 int idx; 589 590 while (s < e) { 591 if (!ISALNUM(*s)) *s = '_'; 592 else if (ISUPPER(*s)) *s = (char)TOLOWER(*s); 593 ++s; 594 } 595 FL_UNSET(enclib, FL_TAINT|FL_UNTRUSTED); 596 OBJ_FREEZE(enclib); 597 ruby_verbose = Qfalse; 598 ruby_debug = Qfalse; 599 errinfo = rb_errinfo(); 600 loaded = rb_protect(require_enc, enclib, 0); 601 ruby_verbose = verbose; 602 ruby_debug = debug; 603 rb_set_errinfo(errinfo); 604 if (NIL_P(loaded)) return -1; 605 if ((idx = rb_enc_registered(name)) < 0) return -1; 606 if (enc_autoload_p(enc_table.list[idx].enc)) return -1; 607 return idx; 608} 609 610static int 611enc_autoload(rb_encoding *enc) 612{ 613 int i; 614 rb_encoding *base = enc_table.list[ENC_TO_ENCINDEX(enc)].base; 615 616 if (base) { 617 i = 0; 618 do { 619 if (i >= enc_table.count) return -1; 620 } while (enc_table.list[i].enc != base && (++i, 1)); 621 if (enc_autoload_p(base)) { 622 if (enc_autoload(base) < 0) return -1; 623 } 624 i = ENC_TO_ENCINDEX(enc); 625 enc_register_at(i, rb_enc_name(enc), base); 626 } 627 else { 628 i = load_encoding(rb_enc_name(enc)); 629 } 630 return i; 631} 632 633/* Return encoding index or UNSPECIFIED_ENCODING from encoding name */ 634int 635rb_enc_find_index(const char *name) 636{ 637 int i = rb_enc_registered(name); 638 rb_encoding *enc; 639 640 if (i < 0) { 641 i = load_encoding(name); 642 } 643 else if (!(enc = rb_enc_from_index(i))) { 644 if (i != UNSPECIFIED_ENCODING) { 645 rb_raise(rb_eArgError, "encoding %s is not registered", name); 646 } 647 } 648 else if (enc_autoload_p(enc)) { 649 if (enc_autoload(enc) < 0) { 650 rb_warn("failed to load encoding (%s); use ASCII-8BIT instead", 651 name); 652 return 0; 653 } 654 } 655 return i; 656} 657 658rb_encoding * 659rb_enc_find(const char *name) 660{ 661 int idx = rb_enc_find_index(name); 662 if (idx < 0) idx = 0; 663 return rb_enc_from_index(idx); 664} 665 666static inline int 667enc_capable(VALUE obj) 668{ 669 if (SPECIAL_CONST_P(obj)) return SYMBOL_P(obj); 670 switch (BUILTIN_TYPE(obj)) { 671 case T_STRING: 672 case T_REGEXP: 673 case T_FILE: 674 return TRUE; 675 case T_DATA: 676 if (is_data_encoding(obj)) return TRUE; 677 default: 678 return FALSE; 679 } 680} 681 682ID 683rb_id_encoding(void) 684{ 685 CONST_ID(id_encoding, "encoding"); 686 return id_encoding; 687} 688 689int 690rb_enc_get_index(VALUE obj) 691{ 692 int i = -1; 693 VALUE tmp; 694 695 if (SPECIAL_CONST_P(obj)) { 696 if (!SYMBOL_P(obj)) return -1; 697 obj = rb_id2str(SYM2ID(obj)); 698 } 699 switch (BUILTIN_TYPE(obj)) { 700 as_default: 701 default: 702 case T_STRING: 703 case T_REGEXP: 704 i = ENCODING_GET_INLINED(obj); 705 if (i == ENCODING_INLINE_MAX) { 706 VALUE iv; 707 708 iv = rb_ivar_get(obj, rb_id_encoding()); 709 i = NUM2INT(iv); 710 } 711 break; 712 case T_FILE: 713 tmp = rb_funcall(obj, rb_intern("internal_encoding"), 0, 0); 714 if (NIL_P(tmp)) obj = rb_funcall(obj, rb_intern("external_encoding"), 0, 0); 715 else obj = tmp; 716 if (NIL_P(obj)) break; 717 case T_DATA: 718 if (is_data_encoding(obj)) { 719 i = enc_check_encoding(obj); 720 } 721 else { 722 goto as_default; 723 } 724 break; 725 } 726 return i; 727} 728 729static void 730enc_set_index(VALUE obj, int idx) 731{ 732 if (idx < ENCODING_INLINE_MAX) { 733 ENCODING_SET_INLINED(obj, idx); 734 return; 735 } 736 ENCODING_SET_INLINED(obj, ENCODING_INLINE_MAX); 737 rb_ivar_set(obj, rb_id_encoding(), INT2NUM(idx)); 738} 739 740void 741rb_enc_set_index(VALUE obj, int idx) 742{ 743 rb_check_frozen(obj); 744 enc_set_index(obj, idx); 745} 746 747VALUE 748rb_enc_associate_index(VALUE obj, int idx) 749{ 750/* enc_check_capable(obj);*/ 751 rb_check_frozen(obj); 752 if (rb_enc_get_index(obj) == idx) 753 return obj; 754 if (SPECIAL_CONST_P(obj)) { 755 rb_raise(rb_eArgError, "cannot set encoding"); 756 } 757 if (!ENC_CODERANGE_ASCIIONLY(obj) || 758 !rb_enc_asciicompat(rb_enc_from_index(idx))) { 759 ENC_CODERANGE_CLEAR(obj); 760 } 761 enc_set_index(obj, idx); 762 return obj; 763} 764 765VALUE 766rb_enc_associate(VALUE obj, rb_encoding *enc) 767{ 768 return rb_enc_associate_index(obj, rb_enc_to_index(enc)); 769} 770 771rb_encoding* 772rb_enc_get(VALUE obj) 773{ 774 return rb_enc_from_index(rb_enc_get_index(obj)); 775} 776 777rb_encoding* 778rb_enc_check(VALUE str1, VALUE str2) 779{ 780 rb_encoding *enc = rb_enc_compatible(str1, str2); 781 if (!enc) 782 rb_raise(rb_eEncCompatError, "incompatible character encodings: %s and %s", 783 rb_enc_name(rb_enc_get(str1)), 784 rb_enc_name(rb_enc_get(str2))); 785 return enc; 786} 787 788rb_encoding* 789rb_enc_compatible(VALUE str1, VALUE str2) 790{ 791 int idx1, idx2; 792 rb_encoding *enc1, *enc2; 793 int isstr1, isstr2; 794 795 idx1 = rb_enc_get_index(str1); 796 idx2 = rb_enc_get_index(str2); 797 798 if (idx1 < 0 || idx2 < 0) 799 return 0; 800 801 if (idx1 == idx2) { 802 return rb_enc_from_index(idx1); 803 } 804 enc1 = rb_enc_from_index(idx1); 805 enc2 = rb_enc_from_index(idx2); 806 807 isstr2 = RB_TYPE_P(str2, T_STRING); 808 if (isstr2 && RSTRING_LEN(str2) == 0) 809 return enc1; 810 isstr1 = RB_TYPE_P(str1, T_STRING); 811 if (isstr1 && RSTRING_LEN(str1) == 0) 812 return (rb_enc_asciicompat(enc1) && rb_enc_str_asciionly_p(str2)) ? enc1 : enc2; 813 if (!rb_enc_asciicompat(enc1) || !rb_enc_asciicompat(enc2)) { 814 return 0; 815 } 816 817 /* objects whose encoding is the same of contents */ 818 if (!isstr2 && idx2 == ENCINDEX_US_ASCII) 819 return enc1; 820 if (!isstr1 && idx1 == ENCINDEX_US_ASCII) 821 return enc2; 822 823 if (!isstr1) { 824 VALUE tmp = str1; 825 int idx0 = idx1; 826 str1 = str2; 827 str2 = tmp; 828 idx1 = idx2; 829 idx2 = idx0; 830 idx0 = isstr1; 831 isstr1 = isstr2; 832 isstr2 = idx0; 833 } 834 if (isstr1) { 835 int cr1, cr2; 836 837 cr1 = rb_enc_str_coderange(str1); 838 if (isstr2) { 839 cr2 = rb_enc_str_coderange(str2); 840 if (cr1 != cr2) { 841 /* may need to handle ENC_CODERANGE_BROKEN */ 842 if (cr1 == ENC_CODERANGE_7BIT) return enc2; 843 if (cr2 == ENC_CODERANGE_7BIT) return enc1; 844 } 845 if (cr2 == ENC_CODERANGE_7BIT) { 846 return enc1; 847 } 848 } 849 if (cr1 == ENC_CODERANGE_7BIT) 850 return enc2; 851 } 852 return 0; 853} 854 855void 856rb_enc_copy(VALUE obj1, VALUE obj2) 857{ 858 rb_enc_associate_index(obj1, rb_enc_get_index(obj2)); 859} 860 861 862/* 863 * call-seq: 864 * obj.encoding -> encoding 865 * 866 * Returns the Encoding object that represents the encoding of obj. 867 */ 868 869VALUE 870rb_obj_encoding(VALUE obj) 871{ 872 int idx = rb_enc_get_index(obj); 873 if (idx < 0) { 874 rb_raise(rb_eTypeError, "unknown encoding"); 875 } 876 return rb_enc_from_encoding_index(idx); 877} 878 879int 880rb_enc_fast_mbclen(const char *p, const char *e, rb_encoding *enc) 881{ 882 return ONIGENC_MBC_ENC_LEN(enc, (UChar*)p, (UChar*)e); 883} 884 885int 886rb_enc_mbclen(const char *p, const char *e, rb_encoding *enc) 887{ 888 int n = ONIGENC_PRECISE_MBC_ENC_LEN(enc, (UChar*)p, (UChar*)e); 889 if (MBCLEN_CHARFOUND_P(n) && MBCLEN_CHARFOUND_LEN(n) <= e-p) 890 return MBCLEN_CHARFOUND_LEN(n); 891 else { 892 int min = rb_enc_mbminlen(enc); 893 return min <= e-p ? min : (int)(e-p); 894 } 895} 896 897int 898rb_enc_precise_mbclen(const char *p, const char *e, rb_encoding *enc) 899{ 900 int n; 901 if (e <= p) 902 return ONIGENC_CONSTRUCT_MBCLEN_NEEDMORE(1); 903 n = ONIGENC_PRECISE_MBC_ENC_LEN(enc, (UChar*)p, (UChar*)e); 904 if (e-p < n) 905 return ONIGENC_CONSTRUCT_MBCLEN_NEEDMORE(n-(int)(e-p)); 906 return n; 907} 908 909int 910rb_enc_ascget(const char *p, const char *e, int *len, rb_encoding *enc) 911{ 912 unsigned int c, l; 913 if (e <= p) 914 return -1; 915 if (rb_enc_asciicompat(enc)) { 916 c = (unsigned char)*p; 917 if (!ISASCII(c)) 918 return -1; 919 if (len) *len = 1; 920 return c; 921 } 922 l = rb_enc_precise_mbclen(p, e, enc); 923 if (!MBCLEN_CHARFOUND_P(l)) 924 return -1; 925 c = rb_enc_mbc_to_codepoint(p, e, enc); 926 if (!rb_enc_isascii(c, enc)) 927 return -1; 928 if (len) *len = l; 929 return c; 930} 931 932unsigned int 933rb_enc_codepoint_len(const char *p, const char *e, int *len_p, rb_encoding *enc) 934{ 935 int r; 936 if (e <= p) 937 rb_raise(rb_eArgError, "empty string"); 938 r = rb_enc_precise_mbclen(p, e, enc); 939 if (!MBCLEN_CHARFOUND_P(r)) { 940 rb_raise(rb_eArgError, "invalid byte sequence in %s", rb_enc_name(enc)); 941 } 942 if (len_p) *len_p = MBCLEN_CHARFOUND_LEN(r); 943 return rb_enc_mbc_to_codepoint(p, e, enc); 944} 945 946#undef rb_enc_codepoint 947unsigned int 948rb_enc_codepoint(const char *p, const char *e, rb_encoding *enc) 949{ 950 return rb_enc_codepoint_len(p, e, 0, enc); 951} 952 953int 954rb_enc_codelen(int c, rb_encoding *enc) 955{ 956 int n = ONIGENC_CODE_TO_MBCLEN(enc,c); 957 if (n == 0) { 958 rb_raise(rb_eArgError, "invalid codepoint 0x%x in %s", c, rb_enc_name(enc)); 959 } 960 return n; 961} 962 963int 964rb_enc_toupper(int c, rb_encoding *enc) 965{ 966 return (ONIGENC_IS_ASCII_CODE(c)?ONIGENC_ASCII_CODE_TO_UPPER_CASE(c):(c)); 967} 968 969int 970rb_enc_tolower(int c, rb_encoding *enc) 971{ 972 return (ONIGENC_IS_ASCII_CODE(c)?ONIGENC_ASCII_CODE_TO_LOWER_CASE(c):(c)); 973} 974 975/* 976 * call-seq: 977 * enc.inspect -> string 978 * 979 * Returns a string which represents the encoding for programmers. 980 * 981 * Encoding::UTF_8.inspect #=> "#<Encoding:UTF-8>" 982 * Encoding::ISO_2022_JP.inspect #=> "#<Encoding:ISO-2022-JP (dummy)>" 983 */ 984static VALUE 985enc_inspect(VALUE self) 986{ 987 VALUE str = rb_sprintf("#<%s:%s%s>", rb_obj_classname(self), 988 rb_enc_name((rb_encoding*)DATA_PTR(self)), 989 (enc_dummy_p(self) ? " (dummy)" : "")); 990 ENCODING_CODERANGE_SET(str, rb_usascii_encindex(), ENC_CODERANGE_7BIT); 991 return str; 992} 993 994/* 995 * call-seq: 996 * enc.name -> string 997 * 998 * Returns the name of the encoding. 999 * 1000 * Encoding::UTF_8.name #=> "UTF-8" 1001 */ 1002static VALUE 1003enc_name(VALUE self) 1004{ 1005 return rb_usascii_str_new2(rb_enc_name((rb_encoding*)DATA_PTR(self))); 1006} 1007 1008static int 1009enc_names_i(st_data_t name, st_data_t idx, st_data_t args) 1010{ 1011 VALUE *arg = (VALUE *)args; 1012 1013 if ((int)idx == (int)arg[0]) { 1014 VALUE str = rb_usascii_str_new2((char *)name); 1015 OBJ_FREEZE(str); 1016 rb_ary_push(arg[1], str); 1017 } 1018 return ST_CONTINUE; 1019} 1020 1021/* 1022 * call-seq: 1023 * enc.names -> array 1024 * 1025 * Returns the list of name and aliases of the encoding. 1026 * 1027 * Encoding::WINDOWS_31J.names #=> ["Windows-31J", "CP932", "csWindows31J"] 1028 */ 1029static VALUE 1030enc_names(VALUE self) 1031{ 1032 VALUE args[2]; 1033 1034 args[0] = (VALUE)rb_to_encoding_index(self); 1035 args[1] = rb_ary_new2(0); 1036 st_foreach(enc_table.names, enc_names_i, (st_data_t)args); 1037 return args[1]; 1038} 1039 1040/* 1041 * call-seq: 1042 * Encoding.list -> [enc1, enc2, ...] 1043 * 1044 * Returns the list of loaded encodings. 1045 * 1046 * Encoding.list 1047 * #=> [#<Encoding:ASCII-8BIT>, #<Encoding:UTF-8>, 1048 * #<Encoding:ISO-2022-JP (dummy)>] 1049 * 1050 * Encoding.find("US-ASCII") 1051 * #=> #<Encoding:US-ASCII> 1052 * 1053 * Encoding.list 1054 * #=> [#<Encoding:ASCII-8BIT>, #<Encoding:UTF-8>, 1055 * #<Encoding:US-ASCII>, #<Encoding:ISO-2022-JP (dummy)>] 1056 * 1057 */ 1058static VALUE 1059enc_list(VALUE klass) 1060{ 1061 VALUE ary = rb_ary_new2(0); 1062 rb_ary_replace(ary, rb_encoding_list); 1063 return ary; 1064} 1065 1066/* 1067 * call-seq: 1068 * Encoding.find(string) -> enc 1069 * Encoding.find(symbol) -> enc 1070 * 1071 * Search the encoding with specified <i>name</i>. 1072 * <i>name</i> should be a string or symbol. 1073 * 1074 * Encoding.find("US-ASCII") #=> #<Encoding:US-ASCII> 1075 * Encoding.find(:Shift_JIS) #=> #<Encoding:Shift_JIS> 1076 * 1077 * Names which this method accept are encoding names and aliases 1078 * including following special aliases 1079 * 1080 * "external":: default external encoding 1081 * "internal":: default internal encoding 1082 * "locale":: locale encoding 1083 * "filesystem":: filesystem encoding 1084 * 1085 * An ArgumentError is raised when no encoding with <i>name</i>. 1086 * Only <code>Encoding.find("internal")</code> however returns nil 1087 * when no encoding named "internal", in other words, when Ruby has no 1088 * default internal encoding. 1089 */ 1090static VALUE 1091enc_find(VALUE klass, VALUE enc) 1092{ 1093 int idx; 1094 if (RB_TYPE_P(enc, T_DATA) && is_data_encoding(enc)) 1095 return enc; 1096 idx = str_to_encindex(enc); 1097 if (idx == UNSPECIFIED_ENCODING) return Qnil; 1098 return rb_enc_from_encoding_index(idx); 1099} 1100 1101/* 1102 * call-seq: 1103 * Encoding.compatible?(obj1, obj2) -> enc or nil 1104 * 1105 * Checks the compatibility of two objects. 1106 * 1107 * If the objects are both strings they are compatible when they are 1108 * concatenatable. The encoding of the concatenated string will be returned 1109 * if they are compatible, nil if they are not. 1110 * 1111 * Encoding.compatible?("\xa1".force_encoding("iso-8859-1"), "b") 1112 * #=> #<Encoding:ISO-8859-1> 1113 * 1114 * Encoding.compatible?( 1115 * "\xa1".force_encoding("iso-8859-1"), 1116 * "\xa1\xa1".force_encoding("euc-jp")) 1117 * #=> nil 1118 * 1119 * If the objects are non-strings their encodings are compatible when they 1120 * have an encoding and: 1121 * * Either encoding is US-ASCII compatible 1122 * * One of the encodings is a 7-bit encoding 1123 * 1124 */ 1125static VALUE 1126enc_compatible_p(VALUE klass, VALUE str1, VALUE str2) 1127{ 1128 rb_encoding *enc; 1129 1130 if (!enc_capable(str1)) return Qnil; 1131 if (!enc_capable(str2)) return Qnil; 1132 enc = rb_enc_compatible(str1, str2); 1133 if (!enc) return Qnil; 1134 return rb_enc_from_encoding(enc); 1135} 1136 1137/* :nodoc: */ 1138static VALUE 1139enc_dump(int argc, VALUE *argv, VALUE self) 1140{ 1141 rb_scan_args(argc, argv, "01", 0); 1142 return enc_name(self); 1143} 1144 1145/* :nodoc: */ 1146static VALUE 1147enc_load(VALUE klass, VALUE str) 1148{ 1149 return enc_find(klass, str); 1150} 1151 1152rb_encoding * 1153rb_ascii8bit_encoding(void) 1154{ 1155 if (!enc_table.list) { 1156 rb_enc_init(); 1157 } 1158 return enc_table.list[ENCINDEX_ASCII].enc; 1159} 1160 1161int 1162rb_ascii8bit_encindex(void) 1163{ 1164 return ENCINDEX_ASCII; 1165} 1166 1167rb_encoding * 1168rb_utf8_encoding(void) 1169{ 1170 if (!enc_table.list) { 1171 rb_enc_init(); 1172 } 1173 return enc_table.list[ENCINDEX_UTF_8].enc; 1174} 1175 1176int 1177rb_utf8_encindex(void) 1178{ 1179 return ENCINDEX_UTF_8; 1180} 1181 1182rb_encoding * 1183rb_usascii_encoding(void) 1184{ 1185 if (!enc_table.list) { 1186 rb_enc_init(); 1187 } 1188 return enc_table.list[ENCINDEX_US_ASCII].enc; 1189} 1190 1191int 1192rb_usascii_encindex(void) 1193{ 1194 return ENCINDEX_US_ASCII; 1195} 1196 1197int 1198rb_locale_encindex(void) 1199{ 1200 VALUE charmap = rb_locale_charmap(rb_cEncoding); 1201 int idx; 1202 1203 if (NIL_P(charmap)) 1204 idx = rb_usascii_encindex(); 1205 else if ((idx = rb_enc_find_index(StringValueCStr(charmap))) < 0) 1206 idx = rb_ascii8bit_encindex(); 1207 1208 if (rb_enc_registered("locale") < 0) enc_alias_internal("locale", idx); 1209 1210 return idx; 1211} 1212 1213rb_encoding * 1214rb_locale_encoding(void) 1215{ 1216 return rb_enc_from_index(rb_locale_encindex()); 1217} 1218 1219static int 1220enc_set_filesystem_encoding(void) 1221{ 1222 int idx; 1223#if defined NO_LOCALE_CHARMAP 1224 idx = rb_enc_to_index(rb_default_external_encoding()); 1225#elif defined _WIN32 || defined __CYGWIN__ 1226 char cp[sizeof(int) * 8 / 3 + 4]; 1227 snprintf(cp, sizeof cp, "CP%d", AreFileApisANSI() ? GetACP() : GetOEMCP()); 1228 idx = rb_enc_find_index(cp); 1229 if (idx < 0) idx = rb_ascii8bit_encindex(); 1230#else 1231 idx = rb_enc_to_index(rb_default_external_encoding()); 1232#endif 1233 1234 enc_alias_internal("filesystem", idx); 1235 return idx; 1236} 1237 1238int 1239rb_filesystem_encindex(void) 1240{ 1241 int idx = rb_enc_registered("filesystem"); 1242 if (idx < 0) 1243 idx = rb_ascii8bit_encindex(); 1244 return idx; 1245} 1246 1247rb_encoding * 1248rb_filesystem_encoding(void) 1249{ 1250 return rb_enc_from_index(rb_filesystem_encindex()); 1251} 1252 1253struct default_encoding { 1254 int index; /* -2 => not yet set, -1 => nil */ 1255 rb_encoding *enc; 1256}; 1257 1258static struct default_encoding default_external = {0}; 1259 1260static int 1261enc_set_default_encoding(struct default_encoding *def, VALUE encoding, const char *name) 1262{ 1263 int overridden = FALSE; 1264 1265 if (def->index != -2) 1266 /* Already set */ 1267 overridden = TRUE; 1268 1269 if (NIL_P(encoding)) { 1270 def->index = -1; 1271 def->enc = 0; 1272 st_insert(enc_table.names, (st_data_t)strdup(name), 1273 (st_data_t)UNSPECIFIED_ENCODING); 1274 } 1275 else { 1276 def->index = rb_enc_to_index(rb_to_encoding(encoding)); 1277 def->enc = 0; 1278 enc_alias_internal(name, def->index); 1279 } 1280 1281 if (def == &default_external) 1282 enc_set_filesystem_encoding(); 1283 1284 return overridden; 1285} 1286 1287rb_encoding * 1288rb_default_external_encoding(void) 1289{ 1290 if (default_external.enc) return default_external.enc; 1291 1292 if (default_external.index >= 0) { 1293 default_external.enc = rb_enc_from_index(default_external.index); 1294 return default_external.enc; 1295 } 1296 else { 1297 return rb_locale_encoding(); 1298 } 1299} 1300 1301VALUE 1302rb_enc_default_external(void) 1303{ 1304 return rb_enc_from_encoding(rb_default_external_encoding()); 1305} 1306 1307/* 1308 * call-seq: 1309 * Encoding.default_external -> enc 1310 * 1311 * Returns default external encoding. 1312 * 1313 * The default external encoding is used by default for strings created from 1314 * the following locations: 1315 * 1316 * * CSV 1317 * * File data read from disk 1318 * * SDBM 1319 * * StringIO 1320 * * Zlib::GzipReader 1321 * * Zlib::GzipWriter 1322 * * String#inspect 1323 * * Regexp#inspect 1324 * 1325 * While strings created from these locations will have this encoding, the 1326 * encoding may not be valid. Be sure to check String#valid_encoding?. 1327 * 1328 * File data written to disk will be transcoded to the default external 1329 * encoding when written. 1330 * 1331 * The default external encoding is initialized by the locale or -E option. 1332 */ 1333static VALUE 1334get_default_external(VALUE klass) 1335{ 1336 return rb_enc_default_external(); 1337} 1338 1339void 1340rb_enc_set_default_external(VALUE encoding) 1341{ 1342 if (NIL_P(encoding)) { 1343 rb_raise(rb_eArgError, "default external can not be nil"); 1344 } 1345 enc_set_default_encoding(&default_external, encoding, 1346 "external"); 1347} 1348 1349/* 1350 * call-seq: 1351 * Encoding.default_external = enc 1352 * 1353 * Sets default external encoding. You should not set 1354 * Encoding::default_external in ruby code as strings created before changing 1355 * the value may have a different encoding from strings created after the value 1356 * was changed., instead you should use <tt>ruby -E</tt> to invoke ruby with 1357 * the correct default_external. 1358 * 1359 * See Encoding::default_external for information on how the default external 1360 * encoding is used. 1361 */ 1362static VALUE 1363set_default_external(VALUE klass, VALUE encoding) 1364{ 1365 rb_warning("setting Encoding.default_external"); 1366 rb_enc_set_default_external(encoding); 1367 return encoding; 1368} 1369 1370static struct default_encoding default_internal = {-2}; 1371 1372rb_encoding * 1373rb_default_internal_encoding(void) 1374{ 1375 if (!default_internal.enc && default_internal.index >= 0) { 1376 default_internal.enc = rb_enc_from_index(default_internal.index); 1377 } 1378 return default_internal.enc; /* can be NULL */ 1379} 1380 1381VALUE 1382rb_enc_default_internal(void) 1383{ 1384 /* Note: These functions cope with default_internal not being set */ 1385 return rb_enc_from_encoding(rb_default_internal_encoding()); 1386} 1387 1388/* 1389 * call-seq: 1390 * Encoding.default_internal -> enc 1391 * 1392 * Returns default internal encoding. Strings will be transcoded to the 1393 * default internal encoding in the following places if the default internal 1394 * encoding is not nil: 1395 * 1396 * * CSV 1397 * * Etc.sysconfdir and Etc.systmpdir 1398 * * File data read from disk 1399 * * File names from Dir 1400 * * Integer#chr 1401 * * String#inspect and Regexp#inspect 1402 * * Strings returned from Curses 1403 * * Strings returned from Readline 1404 * * Strings returned from SDBM 1405 * * Time#zone 1406 * * Values from ENV 1407 * * Values in ARGV including $PROGRAM_NAME 1408 * * __FILE__ 1409 * 1410 * Additionally String#encode and String#encode! use the default internal 1411 * encoding if no encoding is given. 1412 * 1413 * The locale encoding (__ENCODING__), not default_internal, is used as the 1414 * encoding of created strings. 1415 * 1416 * Encoding::default_internal is initialized by the source file's 1417 * internal_encoding or -E option. 1418 */ 1419static VALUE 1420get_default_internal(VALUE klass) 1421{ 1422 return rb_enc_default_internal(); 1423} 1424 1425void 1426rb_enc_set_default_internal(VALUE encoding) 1427{ 1428 enc_set_default_encoding(&default_internal, encoding, 1429 "internal"); 1430} 1431 1432/* 1433 * call-seq: 1434 * Encoding.default_internal = enc or nil 1435 * 1436 * Sets default internal encoding or removes default internal encoding when 1437 * passed nil. You should not set Encoding::default_internal in ruby code as 1438 * strings created before changing the value may have a different encoding 1439 * from strings created after the change. Instead you should use 1440 * <tt>ruby -E</tt> to invoke ruby with the correct default_internal. 1441 * 1442 * See Encoding::default_internal for information on how the default internal 1443 * encoding is used. 1444 */ 1445static VALUE 1446set_default_internal(VALUE klass, VALUE encoding) 1447{ 1448 rb_warning("setting Encoding.default_internal"); 1449 rb_enc_set_default_internal(encoding); 1450 return encoding; 1451} 1452 1453/* 1454 * call-seq: 1455 * Encoding.locale_charmap -> string 1456 * 1457 * Returns the locale charmap name. 1458 * It returns nil if no appropriate information. 1459 * 1460 * Debian GNU/Linux 1461 * LANG=C 1462 * Encoding.locale_charmap #=> "ANSI_X3.4-1968" 1463 * LANG=ja_JP.EUC-JP 1464 * Encoding.locale_charmap #=> "EUC-JP" 1465 * 1466 * SunOS 5 1467 * LANG=C 1468 * Encoding.locale_charmap #=> "646" 1469 * LANG=ja 1470 * Encoding.locale_charmap #=> "eucJP" 1471 * 1472 * The result is highly platform dependent. 1473 * So Encoding.find(Encoding.locale_charmap) may cause an error. 1474 * If you need some encoding object even for unknown locale, 1475 * Encoding.find("locale") can be used. 1476 * 1477 */ 1478VALUE 1479rb_locale_charmap(VALUE klass) 1480{ 1481#if defined NO_LOCALE_CHARMAP 1482 return rb_usascii_str_new2("ASCII-8BIT"); 1483#elif defined _WIN32 || defined __CYGWIN__ 1484 const char *codeset = 0; 1485 char cp[sizeof(int) * 3 + 4]; 1486# ifdef __CYGWIN__ 1487 const char *nl_langinfo_codeset(void); 1488 codeset = nl_langinfo_codeset(); 1489# endif 1490 if (!codeset) { 1491 UINT codepage = GetConsoleCP(); 1492 if (!codepage) codepage = GetACP(); 1493 snprintf(cp, sizeof(cp), "CP%d", codepage); 1494 codeset = cp; 1495 } 1496 return rb_usascii_str_new2(codeset); 1497#elif defined HAVE_LANGINFO_H 1498 char *codeset; 1499 codeset = nl_langinfo(CODESET); 1500 return rb_usascii_str_new2(codeset); 1501#else 1502 return Qnil; 1503#endif 1504} 1505 1506static void 1507set_encoding_const(const char *name, rb_encoding *enc) 1508{ 1509 VALUE encoding = rb_enc_from_encoding(enc); 1510 char *s = (char *)name; 1511 int haslower = 0, hasupper = 0, valid = 0; 1512 1513 if (ISDIGIT(*s)) return; 1514 if (ISUPPER(*s)) { 1515 hasupper = 1; 1516 while (*++s && (ISALNUM(*s) || *s == '_')) { 1517 if (ISLOWER(*s)) haslower = 1; 1518 } 1519 } 1520 if (!*s) { 1521 if (s - name > ENCODING_NAMELEN_MAX) return; 1522 valid = 1; 1523 rb_define_const(rb_cEncoding, name, encoding); 1524 } 1525 if (!valid || haslower) { 1526 size_t len = s - name; 1527 if (len > ENCODING_NAMELEN_MAX) return; 1528 if (!haslower || !hasupper) { 1529 do { 1530 if (ISLOWER(*s)) haslower = 1; 1531 if (ISUPPER(*s)) hasupper = 1; 1532 } while (*++s && (!haslower || !hasupper)); 1533 len = s - name; 1534 } 1535 len += strlen(s); 1536 if (len++ > ENCODING_NAMELEN_MAX) return; 1537 MEMCPY(s = ALLOCA_N(char, len), name, char, len); 1538 name = s; 1539 if (!valid) { 1540 if (ISLOWER(*s)) *s = ONIGENC_ASCII_CODE_TO_UPPER_CASE((int)*s); 1541 for (; *s; ++s) { 1542 if (!ISALNUM(*s)) *s = '_'; 1543 } 1544 if (hasupper) { 1545 rb_define_const(rb_cEncoding, name, encoding); 1546 } 1547 } 1548 if (haslower) { 1549 for (s = (char *)name; *s; ++s) { 1550 if (ISLOWER(*s)) *s = ONIGENC_ASCII_CODE_TO_UPPER_CASE((int)*s); 1551 } 1552 rb_define_const(rb_cEncoding, name, encoding); 1553 } 1554 } 1555} 1556 1557static int 1558rb_enc_name_list_i(st_data_t name, st_data_t idx, st_data_t arg) 1559{ 1560 VALUE ary = (VALUE)arg; 1561 VALUE str = rb_usascii_str_new2((char *)name); 1562 OBJ_FREEZE(str); 1563 rb_ary_push(ary, str); 1564 return ST_CONTINUE; 1565} 1566 1567/* 1568 * call-seq: 1569 * Encoding.name_list -> ["enc1", "enc2", ...] 1570 * 1571 * Returns the list of available encoding names. 1572 * 1573 * Encoding.name_list 1574 * #=> ["US-ASCII", "ASCII-8BIT", "UTF-8", 1575 * "ISO-8859-1", "Shift_JIS", "EUC-JP", 1576 * "Windows-31J", 1577 * "BINARY", "CP932", "eucJP"] 1578 * 1579 */ 1580 1581static VALUE 1582rb_enc_name_list(VALUE klass) 1583{ 1584 VALUE ary = rb_ary_new2(enc_table.names->num_entries); 1585 st_foreach(enc_table.names, rb_enc_name_list_i, (st_data_t)ary); 1586 return ary; 1587} 1588 1589static int 1590rb_enc_aliases_enc_i(st_data_t name, st_data_t orig, st_data_t arg) 1591{ 1592 VALUE *p = (VALUE *)arg; 1593 VALUE aliases = p[0], ary = p[1]; 1594 int idx = (int)orig; 1595 VALUE key, str = rb_ary_entry(ary, idx); 1596 1597 if (NIL_P(str)) { 1598 rb_encoding *enc = rb_enc_from_index(idx); 1599 1600 if (!enc) return ST_CONTINUE; 1601 if (STRCASECMP((char*)name, rb_enc_name(enc)) == 0) { 1602 return ST_CONTINUE; 1603 } 1604 str = rb_usascii_str_new2(rb_enc_name(enc)); 1605 OBJ_FREEZE(str); 1606 rb_ary_store(ary, idx, str); 1607 } 1608 key = rb_usascii_str_new2((char *)name); 1609 OBJ_FREEZE(key); 1610 rb_hash_aset(aliases, key, str); 1611 return ST_CONTINUE; 1612} 1613 1614/* 1615 * call-seq: 1616 * Encoding.aliases -> {"alias1" => "orig1", "alias2" => "orig2", ...} 1617 * 1618 * Returns the hash of available encoding alias and original encoding name. 1619 * 1620 * Encoding.aliases 1621 * #=> {"BINARY"=>"ASCII-8BIT", "ASCII"=>"US-ASCII", "ANSI_X3.4-1986"=>"US-ASCII", 1622 * "SJIS"=>"Shift_JIS", "eucJP"=>"EUC-JP", "CP932"=>"Windows-31J"} 1623 * 1624 */ 1625 1626static VALUE 1627rb_enc_aliases(VALUE klass) 1628{ 1629 VALUE aliases[2]; 1630 aliases[0] = rb_hash_new(); 1631 aliases[1] = rb_ary_new(); 1632 st_foreach(enc_table.names, rb_enc_aliases_enc_i, (st_data_t)aliases); 1633 return aliases[0]; 1634} 1635 1636/* 1637 * An Encoding instance represents a character encoding usable in Ruby. It is 1638 * defined as a constant under the Encoding namespace. It has a name and 1639 * optionally, aliases: 1640 * 1641 * Encoding::ISO_8859_1.name 1642 * #=> #<Encoding:ISO-8859-1> 1643 * 1644 * Encoding::ISO_8859_1.names 1645 * #=> ["ISO-8859-1", "ISO8859-1"] 1646 * 1647 * Ruby methods dealing with encodings return or accept Encoding instances as 1648 * arguments (when a method accepts an Encoding instance as an argument, it 1649 * can be passed an Encoding name or alias instead). 1650 * 1651 * "some string".encoding 1652 * #=> #<Encoding:UTF-8> 1653 * 1654 * string = "some string".encode(Encoding::ISO_8859_1) 1655 * #=> "some string" 1656 * string.encoding 1657 * #=> #<Encoding:ISO-8859-1> 1658 * 1659 * "some string".encode "ISO-8859-1" 1660 * #=> "some string" 1661 * 1662 * <code>Encoding::ASCII_8BIT</code> is a special encoding that is usually 1663 * used for a byte string, not a character string. But as the name insists, 1664 * its characters in the range of ASCII are considered as ASCII characters. 1665 * This is useful when you use ASCII-8BIT characters with other ASCII 1666 * compatible characters. 1667 * 1668 * == Changing an encoding 1669 * 1670 * The associated Encoding of a String can be changed in two different ways. 1671 * 1672 * First, it is possible to set the Encoding of a string to a new Encoding 1673 * without changing the internal byte representation of the string, with 1674 * String#force_encoding. This is how you can tell Ruby the correct encoding 1675 * of a string. 1676 * 1677 * string 1678 * #=> "R\xC3\xA9sum\xC3\xA9" 1679 * string.encoding 1680 * #=> #<Encoding:ISO-8859-1> 1681 * string.force_encoding(Encoding::UTF-8) 1682 * #=> "R\u00E9sum\u00E9" 1683 * 1684 * Second, it is possible to transcode a string, i.e. translate its internal 1685 * byte representation to another encoding. Its associated encoding is also 1686 * set to the other encoding. See String#encode for the various forms of 1687 * transcoding, and the Encoding::Converter class for additional control over 1688 * the transcoding process. 1689 * 1690 * string 1691 * #=> "R\u00E9sum\u00E9" 1692 * string.encoding 1693 * #=> #<Encoding:UTF-8> 1694 * string = string.encode!(Encoding::ISO_8859_1) 1695 * #=> "R\xE9sum\xE9" 1696 * string.encoding 1697 * #=> #<Encoding::ISO-8859-1> 1698 * 1699 * == Script encoding 1700 * 1701 * All Ruby script code has an associated Encoding which any String literal 1702 * created in the source code will be associated to. 1703 * 1704 * The default script encoding is <code>Encoding::US-ASCII</code>, but it can 1705 * be changed by a magic comment on the first line of the source code file (or 1706 * second line, if there is a shebang line on the first). The comment must 1707 * contain the word <code>coding</code> or <code>encoding</code>, followed 1708 * by a colon, space and the Encoding name or alias: 1709 * 1710 * # encoding: UTF-8 1711 * 1712 * "some string".encoding 1713 * #=> #<Encoding:UTF-8> 1714 * 1715 * The <code>__ENCODING__</code> keyword returns the script encoding of the file 1716 * which the keyword is written: 1717 * 1718 * # encoding: ISO-8859-1 1719 * 1720 * __ENCODING__ 1721 * #=> #<Encoding:ISO-8859-1> 1722 * 1723 * <code>ruby -K</code> will change the default locale encoding, but this is 1724 * not recommended. Ruby source files should declare its script encoding by a 1725 * magic comment even when they only depend on US-ASCII strings or regular 1726 * expressions. 1727 * 1728 * == Locale encoding 1729 * 1730 * The default encoding of the environment. Usually derived from locale. 1731 * 1732 * see Encoding.locale_charmap, Encoding.find('locale') 1733 * 1734 * == Filesystem encoding 1735 * 1736 * The default encoding of strings from the filesystem of the environment. 1737 * This is used for strings of file names or paths. 1738 * 1739 * see Encoding.find('filesystem') 1740 * 1741 * == External encoding 1742 * 1743 * Each IO object has an external encoding which indicates the encoding that 1744 * Ruby will use to read its data. By default Ruby sets the external encoding 1745 * of an IO object to the default external encoding. The default external 1746 * encoding is set by locale encoding or the interpreter <code>-E</code> option. 1747 * Encoding.default_external returns the current value of the external 1748 * encoding. 1749 * 1750 * ENV["LANG"] 1751 * #=> "UTF-8" 1752 * Encoding.default_external 1753 * #=> #<Encoding:UTF-8> 1754 * 1755 * $ ruby -E ISO-8859-1 -e "p Encoding.default_external" 1756 * #<Encoding:ISO-8859-1> 1757 * 1758 * $ LANG=C ruby -e 'p Encoding.default_external' 1759 * #<Encoding:US-ASCII> 1760 * 1761 * The default external encoding may also be set through 1762 * Encoding.default_external=, but you should not do this as strings created 1763 * before and after the change will have inconsistent encodings. Instead use 1764 * <code>ruby -E</code> to invoke ruby with the correct external encoding. 1765 * 1766 * When you know that the actual encoding of the data of an IO object is not 1767 * the default external encoding, you can reset its external encoding with 1768 * IO#set_encoding or set it at IO object creation (see IO.new options). 1769 * 1770 * == Internal encoding 1771 * 1772 * To process the data of an IO object which has an encoding different 1773 * from its external encoding, you can set its internal encoding. Ruby will use 1774 * this internal encoding to transcode the data when it is read from the IO 1775 * object. 1776 * 1777 * Conversely, when data is written to the IO object it is transcoded from the 1778 * internal encoding to the external encoding of the IO object. 1779 * 1780 * The internal encoding of an IO object can be set with 1781 * IO#set_encoding or at IO object creation (see IO.new options). 1782 * 1783 * The internal encoding is optional and when not set, the Ruby default 1784 * internal encoding is used. If not explicitly set this default internal 1785 * encoding is +nil+ meaning that by default, no transcoding occurs. 1786 * 1787 * The default internal encoding can be set with the interpreter option 1788 * <code>-E</code>. Encoding.default_internal returns the current internal 1789 * encoding. 1790 * 1791 * $ ruby -e 'p Encoding.default_internal' 1792 * nil 1793 * 1794 * $ ruby -E ISO-8859-1:UTF-8 -e "p [Encoding.default_external, \ 1795 * Encoding.default_internal]" 1796 * [#<Encoding:ISO-8859-1>, #<Encoding:UTF-8>] 1797 * 1798 * The default internal encoding may also be set through 1799 * Encoding.default_internal=, but you should not do this as strings created 1800 * before and after the change will have inconsistent encodings. Instead use 1801 * <code>ruby -E</code> to invoke ruby with the correct internal encoding. 1802 * 1803 * == IO encoding example 1804 * 1805 * In the following example a UTF-8 encoded string "R\u00E9sum\u00E9" is transcoded for 1806 * output to ISO-8859-1 encoding, then read back in and transcoded to UTF-8: 1807 * 1808 * string = "R\u00E9sum\u00E9" 1809 * 1810 * open("transcoded.txt", "w:ISO-8859-1") do |io| 1811 * io.write(string) 1812 * end 1813 * 1814 * puts "raw text:" 1815 * p File.binread("transcoded.txt") 1816 * puts 1817 * 1818 * open("transcoded.txt", "r:ISO-8859-1:UTF-8") do |io| 1819 * puts "transcoded text:" 1820 * p io.read 1821 * end 1822 * 1823 * While writing the file, the internal encoding is not specified as it is 1824 * only necessary for reading. While reading the file both the internal and 1825 * external encoding must be specified to obtain the correct result. 1826 * 1827 * $ ruby t.rb 1828 * raw text: 1829 * "R\xE9sum\xE9" 1830 * 1831 * transcoded text: 1832 * "R\u00E9sum\u00E9" 1833 * 1834 */ 1835 1836void 1837Init_Encoding(void) 1838{ 1839#undef rb_intern 1840#define rb_intern(str) rb_intern_const(str) 1841 VALUE list; 1842 int i; 1843 1844 rb_cEncoding = rb_define_class("Encoding", rb_cObject); 1845 rb_undef_alloc_func(rb_cEncoding); 1846 rb_undef_method(CLASS_OF(rb_cEncoding), "new"); 1847 rb_define_method(rb_cEncoding, "to_s", enc_name, 0); 1848 rb_define_method(rb_cEncoding, "inspect", enc_inspect, 0); 1849 rb_define_method(rb_cEncoding, "name", enc_name, 0); 1850 rb_define_method(rb_cEncoding, "names", enc_names, 0); 1851 rb_define_method(rb_cEncoding, "dummy?", enc_dummy_p, 0); 1852 rb_define_method(rb_cEncoding, "ascii_compatible?", enc_ascii_compatible_p, 0); 1853 rb_define_method(rb_cEncoding, "replicate", enc_replicate, 1); 1854 rb_define_singleton_method(rb_cEncoding, "list", enc_list, 0); 1855 rb_define_singleton_method(rb_cEncoding, "name_list", rb_enc_name_list, 0); 1856 rb_define_singleton_method(rb_cEncoding, "aliases", rb_enc_aliases, 0); 1857 rb_define_singleton_method(rb_cEncoding, "find", enc_find, 1); 1858 rb_define_singleton_method(rb_cEncoding, "compatible?", enc_compatible_p, 2); 1859 1860 rb_define_method(rb_cEncoding, "_dump", enc_dump, -1); 1861 rb_define_singleton_method(rb_cEncoding, "_load", enc_load, 1); 1862 1863 rb_define_singleton_method(rb_cEncoding, "default_external", get_default_external, 0); 1864 rb_define_singleton_method(rb_cEncoding, "default_external=", set_default_external, 1); 1865 rb_define_singleton_method(rb_cEncoding, "default_internal", get_default_internal, 0); 1866 rb_define_singleton_method(rb_cEncoding, "default_internal=", set_default_internal, 1); 1867 rb_define_singleton_method(rb_cEncoding, "locale_charmap", rb_locale_charmap, 0); 1868 1869 list = rb_ary_new2(enc_table.count); 1870 RBASIC(list)->klass = 0; 1871 rb_encoding_list = list; 1872 rb_gc_register_mark_object(list); 1873 1874 for (i = 0; i < enc_table.count; ++i) { 1875 rb_ary_push(list, enc_new(enc_table.list[i].enc)); 1876 } 1877} 1878 1879/* locale insensitive ctype functions */ 1880 1881#define ctype_test(c, ctype) \ 1882 (rb_isascii(c) && ONIGENC_IS_ASCII_CODE_CTYPE((c), (ctype))) 1883 1884int rb_isalnum(int c) { return ctype_test(c, ONIGENC_CTYPE_ALNUM); } 1885int rb_isalpha(int c) { return ctype_test(c, ONIGENC_CTYPE_ALPHA); } 1886int rb_isblank(int c) { return ctype_test(c, ONIGENC_CTYPE_BLANK); } 1887int rb_iscntrl(int c) { return ctype_test(c, ONIGENC_CTYPE_CNTRL); } 1888int rb_isdigit(int c) { return ctype_test(c, ONIGENC_CTYPE_DIGIT); } 1889int rb_isgraph(int c) { return ctype_test(c, ONIGENC_CTYPE_GRAPH); } 1890int rb_islower(int c) { return ctype_test(c, ONIGENC_CTYPE_LOWER); } 1891int rb_isprint(int c) { return ctype_test(c, ONIGENC_CTYPE_PRINT); } 1892int rb_ispunct(int c) { return ctype_test(c, ONIGENC_CTYPE_PUNCT); } 1893int rb_isspace(int c) { return ctype_test(c, ONIGENC_CTYPE_SPACE); } 1894int rb_isupper(int c) { return ctype_test(c, ONIGENC_CTYPE_UPPER); } 1895int rb_isxdigit(int c) { return ctype_test(c, ONIGENC_CTYPE_XDIGIT); } 1896 1897int 1898rb_tolower(int c) 1899{ 1900 return rb_isascii(c) ? ONIGENC_ASCII_CODE_TO_LOWER_CASE(c) : c; 1901} 1902 1903int 1904rb_toupper(int c) 1905{ 1906 return rb_isascii(c) ? ONIGENC_ASCII_CODE_TO_UPPER_CASE(c) : c; 1907} 1908 1909