src/coding.h

   1 /* Header for coding system handler.
   2    Copyright (C) 2001-2019 Free Software Foundation, Inc.
   3    Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   4      2005, 2006, 2007, 2008, 2009, 2010, 2011
   5      National Institute of Advanced Industrial Science and Technology (AIST)
   6      Registration Number H14PRO021
   7    Copyright (C) 2003
   8      National Institute of Advanced Industrial Science and Technology (AIST)
   9      Registration Number H13PRO009
  10
  11 This file is part of GNU Emacs.
  12
  13 GNU Emacs is free software: you can redistribute it and/or modify
  14 it under the terms of the GNU General Public License as published by
  15 the Free Software Foundation, either version 3 of the License, or (at
  16 your option) any later version.
  17
  18 GNU Emacs is distributed in the hope that it will be useful,
  19 but WITHOUT ANY WARRANTY; without even the implied warranty of
  20 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  21 GNU General Public License for more details.
  22
  23 You should have received a copy of the GNU General Public License
  24 along with GNU Emacs.  If not, see <https://www.gnu.org/licenses/>.  */
  25
  26 #ifndef EMACS_CODING_H
  27 #define EMACS_CODING_H
  28
  29 #include "lisp.h"
  30
  31 INLINE_HEADER_BEGIN
  32
  33 /* Index to arguments of Fdefine_coding_system_internal.  */
  34
  35 enum define_coding_system_arg_index
  36   {
  37     coding_arg_name,
  38     coding_arg_mnemonic,
  39     coding_arg_coding_type,
  40     coding_arg_charset_list,
  41     coding_arg_ascii_compatible_p,
  42     coding_arg_decode_translation_table,
  43     coding_arg_encode_translation_table,
  44     coding_arg_post_read_conversion,
  45     coding_arg_pre_write_conversion,
  46     coding_arg_default_char,
  47     coding_arg_for_unibyte,
  48     coding_arg_plist,
  49     coding_arg_eol_type,
  50     coding_arg_max
  51   };
  52
  53 enum define_coding_iso2022_arg_index
  54   {
  55     coding_arg_iso2022_initial = coding_arg_max,
  56     coding_arg_iso2022_reg_usage,
  57     coding_arg_iso2022_request,
  58     coding_arg_iso2022_flags,
  59     coding_arg_iso2022_max
  60   };
  61
  62 enum define_coding_utf8_arg_index
  63   {
  64     coding_arg_utf8_bom = coding_arg_max,
  65     coding_arg_utf8_max
  66   };
  67
  68 enum define_coding_utf16_arg_index
  69   {
  70     coding_arg_utf16_bom = coding_arg_max,
  71     coding_arg_utf16_endian,
  72     coding_arg_utf16_max
  73   };
  74
  75 enum define_coding_ccl_arg_index
  76   {
  77     coding_arg_ccl_decoder = coding_arg_max,
  78     coding_arg_ccl_encoder,
  79     coding_arg_ccl_valids,
  80     coding_arg_ccl_max
  81   };
  82
  83 enum define_coding_undecided_arg_index
  84   {
  85     coding_arg_undecided_inhibit_null_byte_detection = coding_arg_max,
  86     coding_arg_undecided_inhibit_iso_escape_detection,
  87     coding_arg_undecided_prefer_utf_8,
  88     coding_arg_undecided_max
  89   };
  90
  91 /* Hash table for all coding systems.  Keys are coding system symbols
  92    and values are spec vectors of the corresponding coding system.  A
  93    spec vector has the form [ ATTRS ALIASES EOL-TYPE ].  ATTRS is a
  94    vector of attribute of the coding system.  ALIASES is a list of
  95    aliases (symbols) of the coding system.  EOL-TYPE is `unix', `dos',
  96    `mac' or a vector of coding systems (symbols).  */
  97
  98 extern Lisp_Object Vcoding_system_hash_table;
  99
 100
 101 /* Enumeration of index to an attribute vector of a coding system.  */
 102
 103 enum coding_attr_index
 104   {
 105     coding_attr_base_name,
 106     coding_attr_docstring,
 107     coding_attr_mnemonic,
 108     coding_attr_type,
 109     coding_attr_charset_list,
 110     coding_attr_ascii_compat,
 111     coding_attr_decode_tbl,
 112     coding_attr_encode_tbl,
 113     coding_attr_trans_tbl,
 114     coding_attr_post_read,
 115     coding_attr_pre_write,
 116     coding_attr_default_char,
 117     coding_attr_for_unibyte,
 118     coding_attr_plist,
 119
 120     coding_attr_category,
 121     coding_attr_safe_charsets,
 122
 123     /* The followings are extra attributes for each type.  */
 124     coding_attr_charset_valids,
 125
 126     coding_attr_ccl_decoder,
 127     coding_attr_ccl_encoder,
 128     coding_attr_ccl_valids,
 129
 130     coding_attr_iso_initial,
 131     coding_attr_iso_usage,
 132     coding_attr_iso_request,
 133     coding_attr_iso_flags,
 134
 135     coding_attr_utf_bom,
 136     coding_attr_utf_16_endian,
 137
 138     coding_attr_emacs_mule_full,
 139
 140     coding_attr_undecided_inhibit_null_byte_detection,
 141     coding_attr_undecided_inhibit_iso_escape_detection,
 142     coding_attr_undecided_prefer_utf_8,
 143
 144     coding_attr_last_index
 145   };
 146
 147
 148 /* Macros to access an element of an attribute vector.  */
 149
 150 #define CODING_ATTR_BASE_NAME(attrs)    AREF (attrs, coding_attr_base_name)
 151 #define CODING_ATTR_TYPE(attrs)         AREF (attrs, coding_attr_type)
 152 #define CODING_ATTR_CHARSET_LIST(attrs) AREF (attrs, coding_attr_charset_list)
 153 #define CODING_ATTR_MNEMONIC(attrs)     AREF (attrs, coding_attr_mnemonic)
 154 #define CODING_ATTR_DOCSTRING(attrs)    AREF (attrs, coding_attr_docstring)
 155 #define CODING_ATTR_ASCII_COMPAT(attrs) AREF (attrs, coding_attr_ascii_compat)
 156 #define CODING_ATTR_DECODE_TBL(attrs)   AREF (attrs, coding_attr_decode_tbl)
 157 #define CODING_ATTR_ENCODE_TBL(attrs)   AREF (attrs, coding_attr_encode_tbl)
 158 #define CODING_ATTR_TRANS_TBL(attrs)    AREF (attrs, coding_attr_trans_tbl)
 159 #define CODING_ATTR_POST_READ(attrs)    AREF (attrs, coding_attr_post_read)
 160 #define CODING_ATTR_PRE_WRITE(attrs)    AREF (attrs, coding_attr_pre_write)
 161 #define CODING_ATTR_DEFAULT_CHAR(attrs) AREF (attrs, coding_attr_default_char)
 162 #define CODING_ATTR_FOR_UNIBYTE(attrs)  AREF (attrs, coding_attr_for_unibyte)
 163 #define CODING_ATTR_PLIST(attrs)        AREF (attrs, coding_attr_plist)
 164 #define CODING_ATTR_CATEGORY(attrs)     AREF (attrs, coding_attr_category)
 165 #define CODING_ATTR_SAFE_CHARSETS(attrs)AREF (attrs, coding_attr_safe_charsets)
 166
 167
 168 /* Return the name of a coding system specified by ID.  */
 169 #define CODING_ID_NAME(id) \
 170   (HASH_KEY (XHASH_TABLE (Vcoding_system_hash_table), id))
 171
 172 /* Return the attribute vector of a coding system specified by ID.  */
 173
 174 #define CODING_ID_ATTRS(id)     \
 175   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 0))
 176
 177 /* Return the list of aliases of a coding system specified by ID.  */
 178
 179 #define CODING_ID_ALIASES(id)   \
 180   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 1))
 181
 182 /* Return the eol-type of a coding system specified by ID.  */
 183
 184 #define CODING_ID_EOL_TYPE(id)  \
 185   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 2))
 186
 187
 188 /* Return the spec vector of CODING_SYSTEM_SYMBOL.  */
 189
 190 #define CODING_SYSTEM_SPEC(coding_system_symbol)        \
 191   (Fgethash (coding_system_symbol, Vcoding_system_hash_table, Qnil))
 192
 193
 194 /* Return the ID of CODING_SYSTEM_SYMBOL.  */
 195
 196 #define CODING_SYSTEM_ID(coding_system_symbol)                  \
 197   hash_lookup (XHASH_TABLE (Vcoding_system_hash_table),         \
 198                coding_system_symbol, NULL)
 199
 200 /* Return true if CODING_SYSTEM_SYMBOL is a coding system.  */
 201
 202 #define CODING_SYSTEM_P(coding_system_symbol)           \
 203   (CODING_SYSTEM_ID (coding_system_symbol) >= 0         \
 204    || (! NILP (coding_system_symbol)                    \
 205        && ! NILP (Fcoding_system_p (coding_system_symbol))))
 206
 207 /* Check if X is a coding system or not.  */
 208
 209 #define CHECK_CODING_SYSTEM(x)                          \
 210   do {                                                  \
 211     if (CODING_SYSTEM_ID (x) < 0                        \
 212         && NILP (Fcheck_coding_system (x)))             \
 213       wrong_type_argument (Qcoding_system_p, (x));      \
 214   } while (false)
 215
 216
 217 /* Check if X is a coding system or not.  If it is, set SEPC to the
 218    spec vector of the coding system.  */
 219
 220 #define CHECK_CODING_SYSTEM_GET_SPEC(x, spec)           \
 221   do {                                                  \
 222     spec = CODING_SYSTEM_SPEC (x);                      \
 223     if (NILP (spec))                                    \
 224       {                                                 \
 225         Fcheck_coding_system (x);                       \
 226         spec = CODING_SYSTEM_SPEC (x);                  \
 227       }                                                 \
 228     if (NILP (spec))                                    \
 229       wrong_type_argument (Qcoding_system_p, (x));      \
 230   } while (false)
 231
 232
 233 /* Check if X is a coding system or not.  If it is, set ID to the
 234    ID of the coding system.  */
 235
 236 #define CHECK_CODING_SYSTEM_GET_ID(x, id)                       \
 237   do                                                            \
 238     {                                                           \
 239       id = CODING_SYSTEM_ID (x);                                \
 240       if (id < 0)                                               \
 241         {                                                       \
 242           Fcheck_coding_system (x);                             \
 243           id = CODING_SYSTEM_ID (x);                            \
 244         }                                                       \
 245       if (id < 0)                                               \
 246         wrong_type_argument (Qcoding_system_p, (x));    \
 247     } while (false)
 248
 249
 250 /*** GENERAL section ***/
 251
 252 /* Enumeration of result code of code conversion.  */
 253 enum coding_result_code
 254   {
 255     CODING_RESULT_SUCCESS,
 256     CODING_RESULT_INSUFFICIENT_SRC,
 257     CODING_RESULT_INSUFFICIENT_DST,
 258     CODING_RESULT_INVALID_SRC,
 259     CODING_RESULT_INTERRUPT
 260   };
 261
 262
 263 /* Macros used for the member `mode' of the struct coding_system.  */
 264
 265 /* If set, the decoding/encoding routines treat the current data as
 266    the last block of the whole text to be converted, and do the
 267    appropriate finishing job.  */
 268 #define CODING_MODE_LAST_BLOCK                  0x01
 269
 270 /* If set, it means that the current source text is in a buffer which
 271    enables selective display.  */
 272 #define CODING_MODE_SELECTIVE_DISPLAY           0x02
 273
 274 /* This flag is used by the decoding/encoding routines on the fly.  If
 275    set, it means that right-to-left text is being processed.  */
 276 #define CODING_MODE_DIRECTION                   0x04
 277
 278 #define CODING_MODE_FIXED_DESTINATION           0x08
 279
 280 /* If set, it means that the encoding routines produces some safe
 281    ASCII characters (usually '?') for unsupported characters.  */
 282 #define CODING_MODE_SAFE_ENCODING               0x10
 283
 284   /* For handling composition sequence.  */
 285 #include "composite.h"
 286
 287 enum composition_state
 288   {
 289     COMPOSING_NO,
 290     COMPOSING_CHAR,
 291     COMPOSING_RULE,
 292     COMPOSING_COMPONENT_CHAR,
 293     COMPOSING_COMPONENT_RULE
 294   };
 295
 296 /* Structure for the current composition status.  */
 297 struct composition_status
 298 {
 299   enum composition_state state;
 300   enum composition_method method;
 301   bool old_form;          /* true if pre-21 form */
 302   int length;             /* number of elements produced in charbuf */
 303   int nchars;             /* number of characters composed */
 304   int ncomps;             /* number of composition components */
 305   /* Maximum carryover is for the case of COMPOSITION_WITH_RULE_ALTCHARS.
 306      See the comment in coding.c.  */
 307   int carryover[4               /* annotation header */
 308                 + MAX_COMPOSITION_COMPONENTS * 3 - 2 /* ALTs and RULEs */
 309                 + 2                                  /* intermediate -1 -1 */
 310                 + MAX_COMPOSITION_COMPONENTS         /* CHARs */
 311                 ];
 312 };
 313
 314
 315 /* Structure of the field `spec.iso_2022' in the structure
 316    `coding_system'.  */
 317 struct iso_2022_spec
 318 {
 319   /* Bit-wise-or of CODING_ISO_FLAG_XXX.  */
 320   unsigned flags;
 321
 322   /* The current graphic register invoked to each graphic plane.  */
 323   int current_invocation[2];
 324
 325   /* The current charset designated to each graphic register.  The
 326      value -1 means that not charset is designated, -2 means that
 327      there was an invalid designation previously.  */
 328   int current_designation[4];
 329
 330   /* If positive, we are now scanning CTEXT extended segment.  */
 331   int ctext_extended_segment_len;
 332
 333   /* True temporarily only when graphic register 2 or 3 is invoked by
 334      single-shift while encoding.  */
 335   bool_bf single_shifting : 1;
 336
 337   /* True temporarily only when processing at beginning of line.  */
 338   bool_bf bol : 1;
 339
 340   /* If true, we are now scanning embedded UTF-8 sequence.  */
 341   bool_bf embedded_utf_8 : 1;
 342
 343   /* The current composition.  */
 344   struct composition_status cmp_status;
 345 };
 346
 347 struct emacs_mule_spec
 348 {
 349   struct composition_status cmp_status;
 350 };
 351
 352 struct undecided_spec
 353 {
 354   /* Inhibit null byte detection.  1 means always inhibit,
 355      -1 means do not inhibit, 0 means rely on user variable.  */
 356   int inhibit_nbd;
 357
 358   /* Inhibit ISO escape detection.  -1, 0, 1 as above.  */
 359   int inhibit_ied;
 360
 361   /* Prefer UTF-8 when the input could be other encodings.  */
 362   bool prefer_utf_8;
 363 };
 364
 365 enum utf_bom_type
 366   {
 367     utf_detect_bom,
 368     utf_without_bom,
 369     utf_with_bom
 370   };
 371
 372 enum utf_16_endian_type
 373   {
 374     utf_16_big_endian,
 375     utf_16_little_endian
 376   };
 377
 378 struct utf_16_spec
 379 {
 380   enum utf_bom_type bom;
 381   enum utf_16_endian_type endian;
 382   int surrogate;
 383 };
 384
 385 struct coding_detection_info
 386 {
 387   /* Values of these members are bitwise-OR of CATEGORY_MASK_XXXs.  */
 388   /* Which categories are already checked.  */
 389   int checked;
 390   /* Which categories are strongly found.  */
 391   int found;
 392   /* Which categories are rejected.  */
 393   int rejected;
 394 };
 395
 396
 397 struct coding_system
 398 {
 399   /* ID number of the coding system.  This is an index to
 400      Vcoding_system_hash_table.  This value is set by
 401      setup_coding_system.  At the early stage of building time, this
 402      value is -1 in the array coding_categories to indicate that no
 403      coding-system of that category is yet defined.  */
 404   ptrdiff_t id;
 405
 406   /* Flag bits of the coding system.  The meaning of each bit is common
 407      to all types of coding systems.  */
 408   unsigned common_flags : 14;
 409
 410   /* Mode bits of the coding system.  See the comments of the macros
 411      CODING_MODE_XXX.  */
 412   unsigned mode : 5;
 413
 414   /* The following two members specify how binary 8-bit code 128..255
 415      are represented in source and destination text respectively.  True
 416      means they are represented by 2-byte sequence, false means they are
 417      represented by 1-byte as is (see the comment in character.h).  */
 418   bool_bf src_multibyte : 1;
 419   bool_bf dst_multibyte : 1;
 420
 421   /* True if the source of conversion is not in the member
 422      `charbuf', but at `src_object'.  */
 423   bool_bf chars_at_source : 1;
 424
 425   /* Nonzero if the result of conversion is in `destination'
 426      buffer rather than in `dst_object'.  */
 427   bool_bf raw_destination : 1;
 428
 429   /* Set to true if charbuf contains an annotation.  */
 430   bool_bf annotated : 1;
 431
 432   /* Used internally in coding.c.  See the comment of detect_ascii.  */
 433   unsigned eol_seen : 3;
 434
 435   /* Finish status of code conversion.  */
 436   ENUM_BF (coding_result_code) result : 3;
 437
 438   int max_charset_id;
 439
 440   /* Detailed information specific to each type of coding system.  */
 441   union
 442     {
 443       struct iso_2022_spec iso_2022;
 444       struct ccl_spec *ccl;     /* Defined in ccl.h.  */
 445       struct utf_16_spec utf_16;
 446       enum utf_bom_type utf_8_bom;
 447       struct emacs_mule_spec emacs_mule;
 448       struct undecided_spec undecided;
 449     } spec;
 450
 451   unsigned char *safe_charsets;
 452
 453   /* How many heading bytes we can skip for decoding.  This is set to
 454      -1 in setup_coding_system, and updated by detect_coding.  So,
 455      when this is equal to the byte length of the text being
 456      converted, we can skip the actual conversion process except for
 457      the eol format.  */
 458   ptrdiff_t head_ascii;
 459
 460   /* How many bytes/chars at the source are detected as valid utf-8
 461      sequence.  Set by detect_coding_utf_8.  */
 462   ptrdiff_t detected_utf8_bytes, detected_utf8_chars;
 463
 464   /* The following members are set by encoding/decoding routine.  */
 465   ptrdiff_t produced, produced_char, consumed, consumed_char;
 466
 467   ptrdiff_t src_pos, src_pos_byte, src_chars, src_bytes;
 468   Lisp_Object src_object;
 469   const unsigned char *source;
 470
 471   ptrdiff_t dst_pos, dst_pos_byte, dst_bytes;
 472   Lisp_Object dst_object;
 473   unsigned char *destination;
 474
 475   /* If an element is non-negative, it is a character code.
 476
 477      If it is in the range -128..-1, it is a 8-bit character code
 478      minus 256.
 479
 480      If it is less than -128, it specifies the start of an annotation
 481      chunk.  The length of the chunk is -128 minus the value of the
 482      element.  The following elements are OFFSET, ANNOTATION-TYPE, and
 483      a sequence of actual data for the annotation.  OFFSET is a
 484      character position offset from dst_pos or src_pos,
 485      ANNOTATION-TYPE specifies the meaning of the annotation and how to
 486      handle the following data..  */
 487   int *charbuf;
 488   int charbuf_size, charbuf_used;
 489
 490   unsigned char carryover[64];
 491   int carryover_bytes;
 492
 493   int default_char;
 494
 495   bool (*detector) (struct coding_system *, struct coding_detection_info *);
 496   void (*decoder) (struct coding_system *);
 497   bool (*encoder) (struct coding_system *);
 498 };
 499
 500 /* Meanings of bits in the member `common_flags' of the structure
 501    coding_system.  The lowest 8 bits are reserved for various kind of
 502    annotations (currently two of them are used).  */
 503 #define CODING_ANNOTATION_MASK                  0x00FF
 504 #define CODING_ANNOTATE_COMPOSITION_MASK        0x0001
 505 #define CODING_ANNOTATE_DIRECTION_MASK          0x0002
 506 #define CODING_ANNOTATE_CHARSET_MASK            0x0003
 507 #define CODING_FOR_UNIBYTE_MASK                 0x0100
 508 #define CODING_REQUIRE_FLUSHING_MASK            0x0200
 509 #define CODING_REQUIRE_DECODING_MASK            0x0400
 510 #define CODING_REQUIRE_ENCODING_MASK            0x0800
 511 #define CODING_REQUIRE_DETECTION_MASK           0x1000
 512 #define CODING_RESET_AT_BOL_MASK                0x2000
 513
 514 /* Return nonzero if the coding context CODING requires annotation
 515    handling.  */
 516 #define CODING_REQUIRE_ANNOTATION(coding) \
 517   ((coding)->common_flags & CODING_ANNOTATION_MASK)
 518
 519 /* Return nonzero if the coding context CODING prefers decoding into
 520    unibyte.  */
 521 #define CODING_FOR_UNIBYTE(coding) \
 522   ((coding)->common_flags & CODING_FOR_UNIBYTE_MASK)
 523
 524 /* Return nonzero if the coding context CODING requires specific code to be
 525    attached at the tail of converted text.  */
 526 #define CODING_REQUIRE_FLUSHING(coding) \
 527   ((coding)->common_flags & CODING_REQUIRE_FLUSHING_MASK)
 528
 529 /* Return nonzero if the coding context CODING requires code conversion on
 530    decoding.  */
 531 #define CODING_REQUIRE_DECODING(coding) \
 532   ((coding)->dst_multibyte              \
 533    || (coding)->common_flags & CODING_REQUIRE_DECODING_MASK)
 534
 535
 536 /* Return nonzero if the coding context CODING requires code conversion on
 537    encoding.
 538    The non-multibyte part of the condition is to support encoding of
 539    unibyte strings/buffers generated by string-as-unibyte or
 540    (set-buffer-multibyte nil) from multibyte strings/buffers.  */
 541 #define CODING_REQUIRE_ENCODING(coding)                         \
 542   ((coding)->src_multibyte                                      \
 543    || (coding)->common_flags & CODING_REQUIRE_ENCODING_MASK     \
 544    || (coding)->mode & CODING_MODE_SELECTIVE_DISPLAY)
 545
 546
 547 /* Return nonzero if the coding context CODING requires some kind of code
 548    detection.  */
 549 #define CODING_REQUIRE_DETECTION(coding) \
 550   ((coding)->common_flags & CODING_REQUIRE_DETECTION_MASK)
 551
 552 /* Return nonzero if the coding context CODING requires code conversion on
 553    decoding or some kind of code detection.  */
 554 #define CODING_MAY_REQUIRE_DECODING(coding)     \
 555   (CODING_REQUIRE_DECODING (coding)             \
 556    || CODING_REQUIRE_DETECTION (coding))
 557
 558 /* Macros to decode or encode a character of JISX0208 in SJIS.  S1 and
 559    S2 are the 1st and 2nd position-codes of JISX0208 in SJIS coding
 560    system.  C1 and C2 are the 1st and 2nd position codes of Emacs'
 561    internal format.  */
 562
 563 #define SJIS_TO_JIS(code)                               \
 564   do {                                                  \
 565     int s1, s2, j1, j2;                                 \
 566                                                         \
 567     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 568                                                         \
 569     if (s2 >= 0x9F)                                     \
 570       (j1 = s1 * 2 - (s1 >= 0xE0 ? 0x160 : 0xE0),       \
 571        j2 = s2 - 0x7E);                                 \
 572     else                                                \
 573       (j1 = s1 * 2 - ((s1 >= 0xE0) ? 0x161 : 0xE1),     \
 574        j2 = s2 - ((s2 >= 0x7F) ? 0x20 : 0x1F));         \
 575     (code) = (j1 << 8) | j2;                            \
 576   } while (false)
 577
 578 #define SJIS_TO_JIS2(code)                              \
 579   do {                                                  \
 580     int s1, s2, j1, j2;                                 \
 581                                                         \
 582     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 583                                                         \
 584     if (s2 >= 0x9F)                                     \
 585       {                                                 \
 586         j1 = (s1 == 0xF0 ? 0x28                         \
 587               : s1 == 0xF1 ? 0x24                       \
 588               : s1 == 0xF2 ? 0x2C                       \
 589               : s1 == 0xF3 ? 0x2E                       \
 590               : 0x6E + (s1 - 0xF4) * 2);                \
 591         j2 = s2 - 0x7E;                                 \
 592       }                                                 \
 593     else                                                \
 594       {                                                 \
 595         j1 = (s1 <= 0xF2 ? 0x21 + (s1 - 0xF0) * 2       \
 596               : s1 <= 0xF4 ? 0x2D + (s1 - 0xF3) * 2     \
 597               : 0x6F + (s1 - 0xF5) * 2);                \
 598         j2 = s2 - ((s2 >= 0x7F ? 0x20 : 0x1F));         \
 599       }                                                 \
 600     (code) = (j1 << 8) | j2;                            \
 601   } while (false)
 602
 603
 604 #define JIS_TO_SJIS(code)                               \
 605   do {                                                  \
 606     int s1, s2, j1, j2;                                 \
 607                                                         \
 608     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 609     if (j1 & 1)                                         \
 610       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x71 : 0xB1),       \
 611        s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F));         \
 612     else                                                \
 613       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x70 : 0xB0),       \
 614        s2 = j2 + 0x7E);                                 \
 615     (code) = (s1 << 8) | s2;                            \
 616   } while (false)
 617
 618 #define JIS_TO_SJIS2(code)                              \
 619   do {                                                  \
 620     int s1, s2, j1, j2;                                 \
 621                                                         \
 622     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 623     if (j1 & 1)                                         \
 624       {                                                 \
 625         s1 = (j1 <= 0x25 ? 0xF0 + (j1 - 0x21) / 2       \
 626               : j1 <= 0x2F ? 0xF3 + (j1 - 0x2D) / 2     \
 627               : 0xF5 + (j1 - 0x6F) / 2);                \
 628         s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F);         \
 629       }                                                 \
 630     else                                                \
 631       {                                                 \
 632         s1 = (j1 == 0x28 ? 0xF0                         \
 633               : j1 == 0x24 ? 0xF1                       \
 634               : j1 == 0x2C ? 0xF2                       \
 635               : j1 == 0x2E ? 0xF3                       \
 636               : 0xF4 + (j1 - 0x6E) / 2);                \
 637         s2 = j2 + 0x7E;                                 \
 638       }                                                 \
 639     (code) = (s1 << 8) | s2;                            \
 640   } while (false)
 641
 642 /* Encode the file name NAME using the specified coding system
 643    for file names, if any.  */
 644 #define ENCODE_FILE(NAME)  encode_file_name (NAME)
 645
 646 /* Decode the file name NAME using the specified coding system
 647    for file names, if any.  */
 648 #define DECODE_FILE(NAME)  decode_file_name (NAME)
 649
 650 /* Encode the string STR using the specified coding system
 651    for system functions, if any.  */
 652 #define ENCODE_SYSTEM(str)                                                 \
 653   (! NILP (Vlocale_coding_system)                                          \
 654    ? code_convert_string_norecord (str, Vlocale_coding_system, true)       \
 655    : str)
 656
 657 /* Decode the string STR using the specified coding system
 658    for system functions, if any.  */
 659 #define DECODE_SYSTEM(str)                                                 \
 660   (! NILP (Vlocale_coding_system)                                          \
 661    ? code_convert_string_norecord (str, Vlocale_coding_system, false)      \
 662    : str)
 663
 664 /* Note that this encodes utf-8, not utf-8-emacs, so it's not a no-op.  */
 665 #define ENCODE_UTF_8(str) code_convert_string_norecord (str, Qutf_8, true)
 666
 667 /* Return true if VAL is a high surrogate.  VAL must be a 16-bit code
 668    unit.  */
 669
 670 #define UTF_16_HIGH_SURROGATE_P(val) \
 671   (((val) & 0xFC00) == 0xD800)
 672
 673 /* Return true if VAL is a low surrogate.  VAL must be a 16-bit code
 674    unit.  */
 675
 676 #define UTF_16_LOW_SURROGATE_P(val) \
 677   (((val) & 0xFC00) == 0xDC00)
 678
 679 /* Return the Unicode code point for the given UTF-16 surrogates.  */
 680
 681 INLINE int
 682 surrogates_to_codepoint (int low, int high)
 683 {
 684   eassert (0 <= low && low <= 0xFFFF);
 685   eassert (0 <= high && high <= 0xFFFF);
 686   eassert (UTF_16_LOW_SURROGATE_P (low));
 687   eassert (UTF_16_HIGH_SURROGATE_P (high));
 688   return 0x10000 + (low - 0xDC00) + ((high - 0xD800) * 0x400);
 689 }
 690
 691 /* Extern declarations.  */
 692 extern Lisp_Object code_conversion_save (bool, bool);
 693 extern bool encode_coding_utf_8 (struct coding_system *);
 694 extern void setup_coding_system (Lisp_Object, struct coding_system *);
 695 extern Lisp_Object coding_charset_list (struct coding_system *);
 696 extern Lisp_Object coding_system_charset_list (Lisp_Object);
 697 extern Lisp_Object code_convert_string (Lisp_Object, Lisp_Object,
 698                                         Lisp_Object, bool, bool, bool);
 699 extern Lisp_Object code_convert_string_norecord (Lisp_Object, Lisp_Object,
 700                                                  bool);
 701 extern Lisp_Object encode_file_name (Lisp_Object);
 702 extern Lisp_Object decode_file_name (Lisp_Object);
 703 extern Lisp_Object raw_text_coding_system (Lisp_Object);
 704 extern bool raw_text_coding_system_p (struct coding_system *);
 705 extern Lisp_Object coding_inherit_eol_type (Lisp_Object, Lisp_Object);
 706 extern Lisp_Object complement_process_encoding_system (Lisp_Object);
 707
 708 extern void decode_coding_gap (struct coding_system *,
 709                                ptrdiff_t, ptrdiff_t);
 710 extern void decode_coding_object (struct coding_system *,
 711                                   Lisp_Object, ptrdiff_t, ptrdiff_t,
 712                                   ptrdiff_t, ptrdiff_t, Lisp_Object);
 713 extern void encode_coding_object (struct coding_system *,
 714                                   Lisp_Object, ptrdiff_t, ptrdiff_t,
 715                                   ptrdiff_t, ptrdiff_t, Lisp_Object);
 716
 717 #if defined (WINDOWSNT) || defined (CYGWIN)
 718
 719 /* These functions use Lisp string objects to store the UTF-16LE
 720    strings that modern versions of Windows expect.  These strings are
 721    not particularly useful to Lisp, and all Lisp strings should be
 722    native Emacs multibyte.  */
 723
 724 /* Access the wide-character string stored in a Lisp string object.  */
 725 #define WCSDATA(x) ((wchar_t *) SDATA (x))
 726
 727 /* Convert the multi-byte string in STR to UTF-16LE encoded unibyte
 728    string, and store it in *BUF.  BUF may safely point to STR on entry.  */
 729 extern wchar_t *to_unicode (Lisp_Object str, Lisp_Object *buf);
 730
 731 /* Convert STR, a UTF-16LE encoded string embedded in a unibyte string
 732    object, to a multi-byte Emacs string and return it.  This function
 733    calls code_convert_string_norecord internally and has all its
 734    failure modes.  STR itself is not modified.  */
 735 extern Lisp_Object from_unicode (Lisp_Object str);
 736
 737 /* Convert WSTR to an Emacs string.  */
 738 extern Lisp_Object from_unicode_buffer (const wchar_t *wstr);
 739
 740 #endif /* WINDOWSNT || CYGWIN */
 741
 742 /* Macros for backward compatibility.  */
 743
 744 #define encode_coding_string(coding, string, nocopy)                    \
 745   (STRING_MULTIBYTE(string) ?                                           \
 746     (encode_coding_object (coding, string, 0, 0, SCHARS (string),       \
 747                            SBYTES (string), Qt),                        \
 748      (coding)->dst_object) : (string))
 749
 750
 751 #define decode_coding_c_string(coding, src, bytes, dst_object)          \
 752   do {                                                                  \
 753     (coding)->source = (src);                                           \
 754     (coding)->src_chars = (coding)->src_bytes = (bytes);                \
 755     decode_coding_object ((coding), Qnil, 0, 0, (bytes), (bytes),       \
 756                           (dst_object));                                \
 757   } while (false)
 758
 759
 760 extern Lisp_Object preferred_coding_system (void);
 761
 762
 763 #ifdef emacs
 764
 765 /* Coding system to be used to encode text for terminal display when
 766    terminal coding system is nil.  */
 767 extern struct coding_system safe_terminal_coding;
 768
 769 #endif
 770
 771 extern char emacs_mule_bytes[256];
 772
 773 INLINE_HEADER_END
 774
 775 #endif /* EMACS_CODING_H */