src/coding.h

   1 /* Header for coding system handler.
   2    Copyright (C) 2001-2018 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 /* Index to arguments of Fdefine_coding_system_internal.  */
  32
  33 enum define_coding_system_arg_index
  34   {
  35     coding_arg_name,
  36     coding_arg_mnemonic,
  37     coding_arg_coding_type,
  38     coding_arg_charset_list,
  39     coding_arg_ascii_compatible_p,
  40     coding_arg_decode_translation_table,
  41     coding_arg_encode_translation_table,
  42     coding_arg_post_read_conversion,
  43     coding_arg_pre_write_conversion,
  44     coding_arg_default_char,
  45     coding_arg_for_unibyte,
  46     coding_arg_plist,
  47     coding_arg_eol_type,
  48     coding_arg_max
  49   };
  50
  51 enum define_coding_iso2022_arg_index
  52   {
  53     coding_arg_iso2022_initial = coding_arg_max,
  54     coding_arg_iso2022_reg_usage,
  55     coding_arg_iso2022_request,
  56     coding_arg_iso2022_flags,
  57     coding_arg_iso2022_max
  58   };
  59
  60 enum define_coding_utf8_arg_index
  61   {
  62     coding_arg_utf8_bom = coding_arg_max,
  63     coding_arg_utf8_max
  64   };
  65
  66 enum define_coding_utf16_arg_index
  67   {
  68     coding_arg_utf16_bom = coding_arg_max,
  69     coding_arg_utf16_endian,
  70     coding_arg_utf16_max
  71   };
  72
  73 enum define_coding_ccl_arg_index
  74   {
  75     coding_arg_ccl_decoder = coding_arg_max,
  76     coding_arg_ccl_encoder,
  77     coding_arg_ccl_valids,
  78     coding_arg_ccl_max
  79   };
  80
  81 enum define_coding_undecided_arg_index
  82   {
  83     coding_arg_undecided_inhibit_null_byte_detection = coding_arg_max,
  84     coding_arg_undecided_inhibit_iso_escape_detection,
  85     coding_arg_undecided_prefer_utf_8,
  86     coding_arg_undecided_max
  87   };
  88
  89 /* Hash table for all coding systems.  Keys are coding system symbols
  90    and values are spec vectors of the corresponding coding system.  A
  91    spec vector has the form [ ATTRS ALIASES EOL-TYPE ].  ATTRS is a
  92    vector of attribute of the coding system.  ALIASES is a list of
  93    aliases (symbols) of the coding system.  EOL-TYPE is `unix', `dos',
  94    `mac' or a vector of coding systems (symbols).  */
  95
  96 extern Lisp_Object Vcoding_system_hash_table;
  97
  98
  99 /* Enumeration of index to an attribute vector of a coding system.  */
 100
 101 enum coding_attr_index
 102   {
 103     coding_attr_base_name,
 104     coding_attr_docstring,
 105     coding_attr_mnemonic,
 106     coding_attr_type,
 107     coding_attr_charset_list,
 108     coding_attr_ascii_compat,
 109     coding_attr_decode_tbl,
 110     coding_attr_encode_tbl,
 111     coding_attr_trans_tbl,
 112     coding_attr_post_read,
 113     coding_attr_pre_write,
 114     coding_attr_default_char,
 115     coding_attr_for_unibyte,
 116     coding_attr_plist,
 117
 118     coding_attr_category,
 119     coding_attr_safe_charsets,
 120
 121     /* The followings are extra attributes for each type.  */
 122     coding_attr_charset_valids,
 123
 124     coding_attr_ccl_decoder,
 125     coding_attr_ccl_encoder,
 126     coding_attr_ccl_valids,
 127
 128     coding_attr_iso_initial,
 129     coding_attr_iso_usage,
 130     coding_attr_iso_request,
 131     coding_attr_iso_flags,
 132
 133     coding_attr_utf_bom,
 134     coding_attr_utf_16_endian,
 135
 136     coding_attr_emacs_mule_full,
 137
 138     coding_attr_undecided_inhibit_null_byte_detection,
 139     coding_attr_undecided_inhibit_iso_escape_detection,
 140     coding_attr_undecided_prefer_utf_8,
 141
 142     coding_attr_last_index
 143   };
 144
 145
 146 /* Macros to access an element of an attribute vector.  */
 147
 148 #define CODING_ATTR_BASE_NAME(attrs)    AREF (attrs, coding_attr_base_name)
 149 #define CODING_ATTR_TYPE(attrs)         AREF (attrs, coding_attr_type)
 150 #define CODING_ATTR_CHARSET_LIST(attrs) AREF (attrs, coding_attr_charset_list)
 151 #define CODING_ATTR_MNEMONIC(attrs)     AREF (attrs, coding_attr_mnemonic)
 152 #define CODING_ATTR_DOCSTRING(attrs)    AREF (attrs, coding_attr_docstring)
 153 #define CODING_ATTR_ASCII_COMPAT(attrs) AREF (attrs, coding_attr_ascii_compat)
 154 #define CODING_ATTR_DECODE_TBL(attrs)   AREF (attrs, coding_attr_decode_tbl)
 155 #define CODING_ATTR_ENCODE_TBL(attrs)   AREF (attrs, coding_attr_encode_tbl)
 156 #define CODING_ATTR_TRANS_TBL(attrs)    AREF (attrs, coding_attr_trans_tbl)
 157 #define CODING_ATTR_POST_READ(attrs)    AREF (attrs, coding_attr_post_read)
 158 #define CODING_ATTR_PRE_WRITE(attrs)    AREF (attrs, coding_attr_pre_write)
 159 #define CODING_ATTR_DEFAULT_CHAR(attrs) AREF (attrs, coding_attr_default_char)
 160 #define CODING_ATTR_FOR_UNIBYTE(attrs)  AREF (attrs, coding_attr_for_unibyte)
 161 #define CODING_ATTR_PLIST(attrs)        AREF (attrs, coding_attr_plist)
 162 #define CODING_ATTR_CATEGORY(attrs)     AREF (attrs, coding_attr_category)
 163 #define CODING_ATTR_SAFE_CHARSETS(attrs)AREF (attrs, coding_attr_safe_charsets)
 164
 165
 166 /* Return the name of a coding system specified by ID.  */
 167 #define CODING_ID_NAME(id) \
 168   (HASH_KEY (XHASH_TABLE (Vcoding_system_hash_table), id))
 169
 170 /* Return the attribute vector of a coding system specified by ID.  */
 171
 172 #define CODING_ID_ATTRS(id)     \
 173   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 0))
 174
 175 /* Return the list of aliases of a coding system specified by ID.  */
 176
 177 #define CODING_ID_ALIASES(id)   \
 178   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 1))
 179
 180 /* Return the eol-type of a coding system specified by ID.  */
 181
 182 #define CODING_ID_EOL_TYPE(id)  \
 183   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 2))
 184
 185
 186 /* Return the spec vector of CODING_SYSTEM_SYMBOL.  */
 187
 188 #define CODING_SYSTEM_SPEC(coding_system_symbol)        \
 189   (Fgethash (coding_system_symbol, Vcoding_system_hash_table, Qnil))
 190
 191
 192 /* Return the ID of CODING_SYSTEM_SYMBOL.  */
 193
 194 #define CODING_SYSTEM_ID(coding_system_symbol)                  \
 195   hash_lookup (XHASH_TABLE (Vcoding_system_hash_table),         \
 196                coding_system_symbol, NULL)
 197
 198 /* Return true if CODING_SYSTEM_SYMBOL is a coding system.  */
 199
 200 #define CODING_SYSTEM_P(coding_system_symbol)           \
 201   (CODING_SYSTEM_ID (coding_system_symbol) >= 0         \
 202    || (! NILP (coding_system_symbol)                    \
 203        && ! NILP (Fcoding_system_p (coding_system_symbol))))
 204
 205 /* Check if X is a coding system or not.  */
 206
 207 #define CHECK_CODING_SYSTEM(x)                          \
 208   do {                                                  \
 209     if (CODING_SYSTEM_ID (x) < 0                        \
 210         && NILP (Fcheck_coding_system (x)))             \
 211       wrong_type_argument (Qcoding_system_p, (x));      \
 212   } while (false)
 213
 214
 215 /* Check if X is a coding system or not.  If it is, set SEPC to the
 216    spec vector of the coding system.  */
 217
 218 #define CHECK_CODING_SYSTEM_GET_SPEC(x, spec)           \
 219   do {                                                  \
 220     spec = CODING_SYSTEM_SPEC (x);                      \
 221     if (NILP (spec))                                    \
 222       {                                                 \
 223         Fcheck_coding_system (x);                       \
 224         spec = CODING_SYSTEM_SPEC (x);                  \
 225       }                                                 \
 226     if (NILP (spec))                                    \
 227       wrong_type_argument (Qcoding_system_p, (x));      \
 228   } while (false)
 229
 230
 231 /* Check if X is a coding system or not.  If it is, set ID to the
 232    ID of the coding system.  */
 233
 234 #define CHECK_CODING_SYSTEM_GET_ID(x, id)                       \
 235   do                                                            \
 236     {                                                           \
 237       id = CODING_SYSTEM_ID (x);                                \
 238       if (id < 0)                                               \
 239         {                                                       \
 240           Fcheck_coding_system (x);                             \
 241           id = CODING_SYSTEM_ID (x);                            \
 242         }                                                       \
 243       if (id < 0)                                               \
 244         wrong_type_argument (Qcoding_system_p, (x));    \
 245     } while (false)
 246
 247
 248 /*** GENERAL section ***/
 249
 250 /* Enumeration of result code of code conversion.  */
 251 enum coding_result_code
 252   {
 253     CODING_RESULT_SUCCESS,
 254     CODING_RESULT_INSUFFICIENT_SRC,
 255     CODING_RESULT_INSUFFICIENT_DST,
 256     CODING_RESULT_INVALID_SRC,
 257     CODING_RESULT_INTERRUPT
 258   };
 259
 260
 261 /* Macros used for the member `mode' of the struct coding_system.  */
 262
 263 /* If set, the decoding/encoding routines treat the current data as
 264    the last block of the whole text to be converted, and do the
 265    appropriate finishing job.  */
 266 #define CODING_MODE_LAST_BLOCK                  0x01
 267
 268 /* If set, it means that the current source text is in a buffer which
 269    enables selective display.  */
 270 #define CODING_MODE_SELECTIVE_DISPLAY           0x02
 271
 272 /* This flag is used by the decoding/encoding routines on the fly.  If
 273    set, it means that right-to-left text is being processed.  */
 274 #define CODING_MODE_DIRECTION                   0x04
 275
 276 #define CODING_MODE_FIXED_DESTINATION           0x08
 277
 278 /* If set, it means that the encoding routines produces some safe
 279    ASCII characters (usually '?') for unsupported characters.  */
 280 #define CODING_MODE_SAFE_ENCODING               0x10
 281
 282   /* For handling composition sequence.  */
 283 #include "composite.h"
 284
 285 enum composition_state
 286   {
 287     COMPOSING_NO,
 288     COMPOSING_CHAR,
 289     COMPOSING_RULE,
 290     COMPOSING_COMPONENT_CHAR,
 291     COMPOSING_COMPONENT_RULE
 292   };
 293
 294 /* Structure for the current composition status.  */
 295 struct composition_status
 296 {
 297   enum composition_state state;
 298   enum composition_method method;
 299   bool old_form;          /* true if pre-21 form */
 300   int length;             /* number of elements produced in charbuf */
 301   int nchars;             /* number of characters composed */
 302   int ncomps;             /* number of composition components */
 303   /* Maximum carryover is for the case of COMPOSITION_WITH_RULE_ALTCHARS.
 304      See the comment in coding.c.  */
 305   int carryover[4               /* annotation header */
 306                 + MAX_COMPOSITION_COMPONENTS * 3 - 2 /* ALTs and RULEs */
 307                 + 2                                  /* intermediate -1 -1 */
 308                 + MAX_COMPOSITION_COMPONENTS         /* CHARs */
 309                 ];
 310 };
 311
 312
 313 /* Structure of the field `spec.iso_2022' in the structure
 314    `coding_system'.  */
 315 struct iso_2022_spec
 316 {
 317   /* Bit-wise-or of CODING_ISO_FLAG_XXX.  */
 318   unsigned flags;
 319
 320   /* The current graphic register invoked to each graphic plane.  */
 321   int current_invocation[2];
 322
 323   /* The current charset designated to each graphic register.  The
 324      value -1 means that not charset is designated, -2 means that
 325      there was an invalid designation previously.  */
 326   int current_designation[4];
 327
 328   /* If positive, we are now scanning CTEXT extended segment.  */
 329   int ctext_extended_segment_len;
 330
 331   /* True temporarily only when graphic register 2 or 3 is invoked by
 332      single-shift while encoding.  */
 333   bool_bf single_shifting : 1;
 334
 335   /* True temporarily only when processing at beginning of line.  */
 336   bool_bf bol : 1;
 337
 338   /* If true, we are now scanning embedded UTF-8 sequence.  */
 339   bool_bf embedded_utf_8 : 1;
 340
 341   /* The current composition.  */
 342   struct composition_status cmp_status;
 343 };
 344
 345 struct emacs_mule_spec
 346 {
 347   struct composition_status cmp_status;
 348 };
 349
 350 struct undecided_spec
 351 {
 352   /* Inhibit null byte detection.  1 means always inhibit,
 353      -1 means do not inhibit, 0 means rely on user variable.  */
 354   int inhibit_nbd;
 355
 356   /* Inhibit ISO escape detection.  -1, 0, 1 as above.  */
 357   int inhibit_ied;
 358
 359   /* Prefer UTF-8 when the input could be other encodings.  */
 360   bool prefer_utf_8;
 361 };
 362
 363 enum utf_bom_type
 364   {
 365     utf_detect_bom,
 366     utf_without_bom,
 367     utf_with_bom
 368   };
 369
 370 enum utf_16_endian_type
 371   {
 372     utf_16_big_endian,
 373     utf_16_little_endian
 374   };
 375
 376 struct utf_16_spec
 377 {
 378   enum utf_bom_type bom;
 379   enum utf_16_endian_type endian;
 380   int surrogate;
 381 };
 382
 383 struct coding_detection_info
 384 {
 385   /* Values of these members are bitwise-OR of CATEGORY_MASK_XXXs.  */
 386   /* Which categories are already checked.  */
 387   int checked;
 388   /* Which categories are strongly found.  */
 389   int found;
 390   /* Which categories are rejected.  */
 391   int rejected;
 392 };
 393
 394
 395 struct coding_system
 396 {
 397   /* ID number of the coding system.  This is an index to
 398      Vcoding_system_hash_table.  This value is set by
 399      setup_coding_system.  At the early stage of building time, this
 400      value is -1 in the array coding_categories to indicate that no
 401      coding-system of that category is yet defined.  */
 402   ptrdiff_t id;
 403
 404   /* Flag bits of the coding system.  The meaning of each bit is common
 405      to all types of coding systems.  */
 406   unsigned common_flags : 14;
 407
 408   /* Mode bits of the coding system.  See the comments of the macros
 409      CODING_MODE_XXX.  */
 410   unsigned mode : 5;
 411
 412   /* The following two members specify how binary 8-bit code 128..255
 413      are represented in source and destination text respectively.  True
 414      means they are represented by 2-byte sequence, false means they are
 415      represented by 1-byte as is (see the comment in character.h).  */
 416   bool_bf src_multibyte : 1;
 417   bool_bf dst_multibyte : 1;
 418
 419   /* True if the source of conversion is not in the member
 420      `charbuf', but at `src_object'.  */
 421   bool_bf chars_at_source : 1;
 422
 423   /* Nonzero if the result of conversion is in `destination'
 424      buffer rather than in `dst_object'.  */
 425   bool_bf raw_destination : 1;
 426
 427   /* Set to true if charbuf contains an annotation.  */
 428   bool_bf annotated : 1;
 429
 430   /* Used internally in coding.c.  See the comment of detect_ascii.  */
 431   unsigned eol_seen : 3;
 432
 433   /* Finish status of code conversion.  */
 434   ENUM_BF (coding_result_code) result : 3;
 435
 436   int max_charset_id;
 437
 438   /* Detailed information specific to each type of coding system.  */
 439   union
 440     {
 441       struct iso_2022_spec iso_2022;
 442       struct ccl_spec *ccl;     /* Defined in ccl.h.  */
 443       struct utf_16_spec utf_16;
 444       enum utf_bom_type utf_8_bom;
 445       struct emacs_mule_spec emacs_mule;
 446       struct undecided_spec undecided;
 447     } spec;
 448
 449   unsigned char *safe_charsets;
 450
 451   /* How may heading bytes we can skip for decoding.  This is set to
 452      -1 in setup_coding_system, and updated by detect_coding.  So,
 453      when this is equal to the byte length of the text being
 454      converted, we can skip the actual conversion process except for
 455      the eol format.  */
 456   ptrdiff_t head_ascii;
 457
 458   /* How many bytes/chars at the source are detected as valid utf-8
 459      sequence.  Set by detect_coding_utf_8.  */
 460   ptrdiff_t detected_utf8_bytes, detected_utf8_chars;
 461
 462   /* The following members are set by encoding/decoding routine.  */
 463   ptrdiff_t produced, produced_char, consumed, consumed_char;
 464
 465   ptrdiff_t src_pos, src_pos_byte, src_chars, src_bytes;
 466   Lisp_Object src_object;
 467   const unsigned char *source;
 468
 469   ptrdiff_t dst_pos, dst_pos_byte, dst_bytes;
 470   Lisp_Object dst_object;
 471   unsigned char *destination;
 472
 473   /* If an element is non-negative, it is a character code.
 474
 475      If it is in the range -128..-1, it is a 8-bit character code
 476      minus 256.
 477
 478      If it is less than -128, it specifies the start of an annotation
 479      chunk.  The length of the chunk is -128 minus the value of the
 480      element.  The following elements are OFFSET, ANNOTATION-TYPE, and
 481      a sequence of actual data for the annotation.  OFFSET is a
 482      character position offset from dst_pos or src_pos,
 483      ANNOTATION-TYPE specifies the meaning of the annotation and how to
 484      handle the following data..  */
 485   int *charbuf;
 486   int charbuf_size, charbuf_used;
 487
 488   unsigned char carryover[64];
 489   int carryover_bytes;
 490
 491   int default_char;
 492
 493   bool (*detector) (struct coding_system *, struct coding_detection_info *);
 494   void (*decoder) (struct coding_system *);
 495   bool (*encoder) (struct coding_system *);
 496 };
 497
 498 /* Meanings of bits in the member `common_flags' of the structure
 499    coding_system.  The lowest 8 bits are reserved for various kind of
 500    annotations (currently two of them are used).  */
 501 #define CODING_ANNOTATION_MASK                  0x00FF
 502 #define CODING_ANNOTATE_COMPOSITION_MASK        0x0001
 503 #define CODING_ANNOTATE_DIRECTION_MASK          0x0002
 504 #define CODING_ANNOTATE_CHARSET_MASK            0x0003
 505 #define CODING_FOR_UNIBYTE_MASK                 0x0100
 506 #define CODING_REQUIRE_FLUSHING_MASK            0x0200
 507 #define CODING_REQUIRE_DECODING_MASK            0x0400
 508 #define CODING_REQUIRE_ENCODING_MASK            0x0800
 509 #define CODING_REQUIRE_DETECTION_MASK           0x1000
 510 #define CODING_RESET_AT_BOL_MASK                0x2000
 511
 512 /* Return nonzero if the coding context CODING requires annotation
 513    handling.  */
 514 #define CODING_REQUIRE_ANNOTATION(coding) \
 515   ((coding)->common_flags & CODING_ANNOTATION_MASK)
 516
 517 /* Return nonzero if the coding context CODING prefers decoding into
 518    unibyte.  */
 519 #define CODING_FOR_UNIBYTE(coding) \
 520   ((coding)->common_flags & CODING_FOR_UNIBYTE_MASK)
 521
 522 /* Return nonzero if the coding context CODING requires specific code to be
 523    attached at the tail of converted text.  */
 524 #define CODING_REQUIRE_FLUSHING(coding) \
 525   ((coding)->common_flags & CODING_REQUIRE_FLUSHING_MASK)
 526
 527 /* Return nonzero if the coding context CODING requires code conversion on
 528    decoding.  */
 529 #define CODING_REQUIRE_DECODING(coding) \
 530   ((coding)->dst_multibyte              \
 531    || (coding)->common_flags & CODING_REQUIRE_DECODING_MASK)
 532
 533
 534 /* Return nonzero if the coding context CODING requires code conversion on
 535    encoding.
 536    The non-multibyte part of the condition is to support encoding of
 537    unibyte strings/buffers generated by string-as-unibyte or
 538    (set-buffer-multibyte nil) from multibyte strings/buffers.  */
 539 #define CODING_REQUIRE_ENCODING(coding)                         \
 540   ((coding)->src_multibyte                                      \
 541    || (coding)->common_flags & CODING_REQUIRE_ENCODING_MASK     \
 542    || (coding)->mode & CODING_MODE_SELECTIVE_DISPLAY)
 543
 544
 545 /* Return nonzero if the coding context CODING requires some kind of code
 546    detection.  */
 547 #define CODING_REQUIRE_DETECTION(coding) \
 548   ((coding)->common_flags & CODING_REQUIRE_DETECTION_MASK)
 549
 550 /* Return nonzero if the coding context CODING requires code conversion on
 551    decoding or some kind of code detection.  */
 552 #define CODING_MAY_REQUIRE_DECODING(coding)     \
 553   (CODING_REQUIRE_DECODING (coding)             \
 554    || CODING_REQUIRE_DETECTION (coding))
 555
 556 /* Macros to decode or encode a character of JISX0208 in SJIS.  S1 and
 557    S2 are the 1st and 2nd position-codes of JISX0208 in SJIS coding
 558    system.  C1 and C2 are the 1st and 2nd position codes of Emacs'
 559    internal format.  */
 560
 561 #define SJIS_TO_JIS(code)                               \
 562   do {                                                  \
 563     int s1, s2, j1, j2;                                 \
 564                                                         \
 565     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 566                                                         \
 567     if (s2 >= 0x9F)                                     \
 568       (j1 = s1 * 2 - (s1 >= 0xE0 ? 0x160 : 0xE0),       \
 569        j2 = s2 - 0x7E);                                 \
 570     else                                                \
 571       (j1 = s1 * 2 - ((s1 >= 0xE0) ? 0x161 : 0xE1),     \
 572        j2 = s2 - ((s2 >= 0x7F) ? 0x20 : 0x1F));         \
 573     (code) = (j1 << 8) | j2;                            \
 574   } while (false)
 575
 576 #define SJIS_TO_JIS2(code)                              \
 577   do {                                                  \
 578     int s1, s2, j1, j2;                                 \
 579                                                         \
 580     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 581                                                         \
 582     if (s2 >= 0x9F)                                     \
 583       {                                                 \
 584         j1 = (s1 == 0xF0 ? 0x28                         \
 585               : s1 == 0xF1 ? 0x24                       \
 586               : s1 == 0xF2 ? 0x2C                       \
 587               : s1 == 0xF3 ? 0x2E                       \
 588               : 0x6E + (s1 - 0xF4) * 2);                \
 589         j2 = s2 - 0x7E;                                 \
 590       }                                                 \
 591     else                                                \
 592       {                                                 \
 593         j1 = (s1 <= 0xF2 ? 0x21 + (s1 - 0xF0) * 2       \
 594               : s1 <= 0xF4 ? 0x2D + (s1 - 0xF3) * 2     \
 595               : 0x6F + (s1 - 0xF5) * 2);                \
 596         j2 = s2 - ((s2 >= 0x7F ? 0x20 : 0x1F));         \
 597       }                                                 \
 598     (code) = (j1 << 8) | j2;                            \
 599   } while (false)
 600
 601
 602 #define JIS_TO_SJIS(code)                               \
 603   do {                                                  \
 604     int s1, s2, j1, j2;                                 \
 605                                                         \
 606     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 607     if (j1 & 1)                                         \
 608       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x71 : 0xB1),       \
 609        s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F));         \
 610     else                                                \
 611       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x70 : 0xB0),       \
 612        s2 = j2 + 0x7E);                                 \
 613     (code) = (s1 << 8) | s2;                            \
 614   } while (false)
 615
 616 #define JIS_TO_SJIS2(code)                              \
 617   do {                                                  \
 618     int s1, s2, j1, j2;                                 \
 619                                                         \
 620     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 621     if (j1 & 1)                                         \
 622       {                                                 \
 623         s1 = (j1 <= 0x25 ? 0xF0 + (j1 - 0x21) / 2       \
 624               : j1 <= 0x2F ? 0xF3 + (j1 - 0x2D) / 2     \
 625               : 0xF5 + (j1 - 0x6F) / 2);                \
 626         s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F);         \
 627       }                                                 \
 628     else                                                \
 629       {                                                 \
 630         s1 = (j1 == 0x28 ? 0xF0                         \
 631               : j1 == 0x24 ? 0xF1                       \
 632               : j1 == 0x2C ? 0xF2                       \
 633               : j1 == 0x2E ? 0xF3                       \
 634               : 0xF4 + (j1 - 0x6E) / 2);                \
 635         s2 = j2 + 0x7E;                                 \
 636       }                                                 \
 637     (code) = (s1 << 8) | s2;                            \
 638   } while (false)
 639
 640 /* Encode the file name NAME using the specified coding system
 641    for file names, if any.  */
 642 #define ENCODE_FILE(NAME)  encode_file_name (NAME)
 643
 644 /* Decode the file name NAME using the specified coding system
 645    for file names, if any.  */
 646 #define DECODE_FILE(NAME)  decode_file_name (NAME)
 647
 648 /* Encode the string STR using the specified coding system
 649    for system functions, if any.  */
 650 #define ENCODE_SYSTEM(str)                                                 \
 651   (! NILP (Vlocale_coding_system)                                          \
 652    ? code_convert_string_norecord (str, Vlocale_coding_system, true)       \
 653    : str)
 654
 655 /* Decode the string STR using the specified coding system
 656    for system functions, if any.  */
 657 #define DECODE_SYSTEM(str)                                                 \
 658   (! NILP (Vlocale_coding_system)                                          \
 659    ? code_convert_string_norecord (str, Vlocale_coding_system, false)      \
 660    : str)
 661
 662 /* Note that this encodes utf-8, not utf-8-emacs, so it's not a no-op.  */
 663 #define ENCODE_UTF_8(str) code_convert_string_norecord (str, Qutf_8, true)
 664
 665 /* Return true if VAL is a high surrogate.  VAL must be a 16-bit code
 666    unit.  */
 667
 668 #define UTF_16_HIGH_SURROGATE_P(val) \
 669   (((val) & 0xFC00) == 0xD800)
 670
 671 /* Return true if VAL is a low surrogate.  VAL must be a 16-bit code
 672    unit.  */
 673
 674 #define UTF_16_LOW_SURROGATE_P(val) \
 675   (((val) & 0xFC00) == 0xDC00)
 676
 677 /* Extern declarations.  */
 678 extern Lisp_Object code_conversion_save (bool, bool);
 679 extern bool encode_coding_utf_8 (struct coding_system *);
 680 extern bool utf8_string_p (Lisp_Object);
 681 extern void setup_coding_system (Lisp_Object, struct coding_system *);
 682 extern Lisp_Object coding_charset_list (struct coding_system *);
 683 extern Lisp_Object coding_system_charset_list (Lisp_Object);
 684 extern Lisp_Object code_convert_string (Lisp_Object, Lisp_Object,
 685                                         Lisp_Object, bool, bool, bool);
 686 extern Lisp_Object code_convert_string_norecord (Lisp_Object, Lisp_Object,
 687                                                  bool);
 688 extern Lisp_Object encode_file_name (Lisp_Object);
 689 extern Lisp_Object decode_file_name (Lisp_Object);
 690 extern Lisp_Object raw_text_coding_system (Lisp_Object);
 691 extern bool raw_text_coding_system_p (struct coding_system *);
 692 extern Lisp_Object coding_inherit_eol_type (Lisp_Object, Lisp_Object);
 693 extern Lisp_Object complement_process_encoding_system (Lisp_Object);
 694
 695 extern void decode_coding_gap (struct coding_system *,
 696                                ptrdiff_t, ptrdiff_t);
 697 extern void decode_coding_object (struct coding_system *,
 698                                   Lisp_Object, ptrdiff_t, ptrdiff_t,
 699                                   ptrdiff_t, ptrdiff_t, Lisp_Object);
 700 extern void encode_coding_object (struct coding_system *,
 701                                   Lisp_Object, ptrdiff_t, ptrdiff_t,
 702                                   ptrdiff_t, ptrdiff_t, Lisp_Object);
 703 /* Defined in this file.  */
 704 INLINE int surrogates_to_codepoint (int, int);
 705
 706 #if defined (WINDOWSNT) || defined (CYGWIN)
 707
 708 /* These functions use Lisp string objects to store the UTF-16LE
 709    strings that modern versions of Windows expect.  These strings are
 710    not particularly useful to Lisp, and all Lisp strings should be
 711    native Emacs multibyte.  */
 712
 713 /* Access the wide-character string stored in a Lisp string object.  */
 714 #define WCSDATA(x) ((wchar_t *) SDATA (x))
 715
 716 /* Convert the multi-byte string in STR to UTF-16LE encoded unibyte
 717    string, and store it in *BUF.  BUF may safely point to STR on entry.  */
 718 extern wchar_t *to_unicode (Lisp_Object str, Lisp_Object *buf);
 719
 720 /* Convert STR, a UTF-16LE encoded string embedded in a unibyte string
 721    object, to a multi-byte Emacs string and return it.  This function
 722    calls code_convert_string_norecord internally and has all its
 723    failure modes.  STR itself is not modified.  */
 724 extern Lisp_Object from_unicode (Lisp_Object str);
 725
 726 /* Convert WSTR to an Emacs string.  */
 727 extern Lisp_Object from_unicode_buffer (const wchar_t *wstr);
 728
 729 #endif /* WINDOWSNT || CYGWIN */
 730
 731 /* Macros for backward compatibility.  */
 732
 733 #define encode_coding_string(coding, string, nocopy)                    \
 734   (STRING_MULTIBYTE(string) ?                                           \
 735     (encode_coding_object (coding, string, 0, 0, SCHARS (string),       \
 736                            SBYTES (string), Qt),                        \
 737      (coding)->dst_object) : (string))
 738
 739
 740 #define decode_coding_c_string(coding, src, bytes, dst_object)          \
 741   do {                                                                  \
 742     (coding)->source = (src);                                           \
 743     (coding)->src_chars = (coding)->src_bytes = (bytes);                \
 744     decode_coding_object ((coding), Qnil, 0, 0, (bytes), (bytes),       \
 745                           (dst_object));                                \
 746   } while (false)
 747
 748
 749 /* Return the Unicode code point for the given UTF-16 surrogates.  */
 750
 751 INLINE int
 752 surrogates_to_codepoint (int low, int high)
 753 {
 754   eassert (0 <= low && low <= 0xFFFF);
 755   eassert (0 <= high && high <= 0xFFFF);
 756   eassert (UTF_16_LOW_SURROGATE_P (low));
 757   eassert (UTF_16_HIGH_SURROGATE_P (high));
 758   return 0x10000 + (low - 0xDC00) + ((high - 0xD800) * 0x400);
 759 }
 760
 761 extern Lisp_Object preferred_coding_system (void);
 762
 763
 764 #ifdef emacs
 765
 766 /* Coding system to be used to encode text for terminal display when
 767    terminal coding system is nil.  */
 768 extern struct coding_system safe_terminal_coding;
 769
 770 #endif
 771
 772 extern char emacs_mule_bytes[256];
 773
 774 #endif /* EMACS_CODING_H */