src/coding.h

   1 /* Header for coding system handler.
   2    Copyright (C) 2001, 2002, 2003, 2004, 2005,
   3                  2006, 2007, 2008  Free Software Foundation, Inc.
   4    Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   5      2005, 2006, 2007, 2008
   6      National Institute of Advanced Industrial Science and Technology (AIST)
   7      Registration Number H14PRO021
   8    Copyright (C) 2003
   9      National Institute of Advanced Industrial Science and Technology (AIST)
  10      Registration Number H13PRO009
  11
  12 This file is part of GNU Emacs.
  13
  14 GNU Emacs is free software: you can redistribute it and/or modify
  15 it under the terms of the GNU General Public License as published by
  16 the Free Software Foundation, either version 3 of the License, or
  17 (at your option) any later version.
  18
  19 GNU Emacs is distributed in the hope that it will be useful,
  20 but WITHOUT ANY WARRANTY; without even the implied warranty of
  21 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  22 GNU General Public License for more details.
  23
  24 You should have received a copy of the GNU General Public License
  25 along with GNU Emacs.  If not, see <http://www.gnu.org/licenses/>.  */
  26
  27 #ifndef EMACS_CODING_H
  28 #define EMACS_CODING_H
  29
  30 /* Index to arguments of Fdefine_coding_system_internal.  */
  31
  32 enum define_coding_system_arg_index
  33   {
  34     coding_arg_name,
  35     coding_arg_mnemonic,
  36     coding_arg_coding_type,
  37     coding_arg_charset_list,
  38     coding_arg_ascii_compatible_p,
  39     coding_arg_decode_translation_table,
  40     coding_arg_encode_translation_table,
  41     coding_arg_post_read_conversion,
  42     coding_arg_pre_write_conversion,
  43     coding_arg_default_char,
  44     coding_arg_for_unibyte,
  45     coding_arg_plist,
  46     coding_arg_eol_type,
  47     coding_arg_max
  48   };
  49
  50 enum define_coding_iso2022_arg_index
  51   {
  52     coding_arg_iso2022_initial = coding_arg_max,
  53     coding_arg_iso2022_reg_usage,
  54     coding_arg_iso2022_request,
  55     coding_arg_iso2022_flags,
  56     coding_arg_iso2022_max
  57   };
  58
  59 enum define_coding_utf16_arg_index
  60   {
  61     coding_arg_utf16_bom = coding_arg_max,
  62     coding_arg_utf16_endian,
  63     coding_arg_utf16_max
  64   };
  65
  66 enum define_coding_ccl_arg_index
  67   {
  68     coding_arg_ccl_decoder = coding_arg_max,
  69     coding_arg_ccl_encoder,
  70     coding_arg_ccl_valids,
  71     coding_arg_ccl_max
  72   };
  73
  74 /* Hash table for all coding systems.  Keys are coding system symbols
  75    and values are spec vectors of the corresponding coding system.  A
  76    spec vector has the form [ ATTRS ALIASES EOL-TYPE ].  ATTRS is a
  77    vector of attribute of the coding system.  ALIASES is a list of
  78    aliases (symbols) of the coding system.  EOL-TYPE is `unix', `dos',
  79    `mac' or a vector of coding systems (symbols).  */
  80
  81 extern Lisp_Object Vcoding_system_hash_table;
  82
  83
  84 /* Enumeration of coding system type.  */
  85
  86 enum coding_system_type
  87   {
  88     coding_type_charset,
  89     coding_type_utf_8,
  90     coding_type_utf_16,
  91     coding_type_iso_2022,
  92     coding_type_emacs_mule,
  93     coding_type_sjis,
  94     coding_type_ccl,
  95     coding_type_raw_text,
  96     coding_type_undecided,
  97     coding_type_max
  98   };
  99
 100
 101 /* Enumeration of end-of-line format type.  */
 102
 103 enum end_of_line_type
 104   {
 105     eol_lf,             /* Line-feed only, same as Emacs' internal
 106                            format.  */
 107     eol_crlf,           /* Sequence of carriage-return and
 108                            line-feed.  */
 109     eol_cr,             /* Carriage-return only.  */
 110     eol_any,            /* Accept any of above.  Produce line-feed
 111                            only.  */
 112     eol_undecided,      /* This value is used to denote that the
 113                            eol-type is not yet undecided.  */
 114     eol_type_max
 115   };
 116
 117 /* Enumeration of index to an attribute vector of a coding system.  */
 118
 119 enum coding_attr_index
 120   {
 121     coding_attr_base_name,
 122     coding_attr_docstring,
 123     coding_attr_mnemonic,
 124     coding_attr_type,
 125     coding_attr_charset_list,
 126     coding_attr_ascii_compat,
 127     coding_attr_decode_tbl,
 128     coding_attr_encode_tbl,
 129     coding_attr_trans_tbl,
 130     coding_attr_post_read,
 131     coding_attr_pre_write,
 132     coding_attr_default_char,
 133     coding_attr_for_unibyte,
 134     coding_attr_plist,
 135
 136     coding_attr_category,
 137     coding_attr_safe_charsets,
 138
 139     /* The followings are extra attributes for each type.  */
 140     coding_attr_charset_valids,
 141
 142     coding_attr_ccl_decoder,
 143     coding_attr_ccl_encoder,
 144     coding_attr_ccl_valids,
 145
 146     coding_attr_iso_initial,
 147     coding_attr_iso_usage,
 148     coding_attr_iso_request,
 149     coding_attr_iso_flags,
 150
 151     coding_attr_utf_16_bom,
 152     coding_attr_utf_16_endian,
 153
 154     coding_attr_emacs_mule_full,
 155
 156     coding_attr_last_index
 157   };
 158
 159
 160 /* Macros to access an element of an attribute vector.  */
 161
 162 #define CODING_ATTR_BASE_NAME(attrs)    AREF (attrs, coding_attr_base_name)
 163 #define CODING_ATTR_TYPE(attrs)         AREF (attrs, coding_attr_type)
 164 #define CODING_ATTR_CHARSET_LIST(attrs) AREF (attrs, coding_attr_charset_list)
 165 #define CODING_ATTR_MNEMONIC(attrs)     AREF (attrs, coding_attr_mnemonic)
 166 #define CODING_ATTR_DOCSTRING(attrs)    AREF (attrs, coding_attr_docstring)
 167 #define CODING_ATTR_ASCII_COMPAT(attrs) AREF (attrs, coding_attr_ascii_compat)
 168 #define CODING_ATTR_DECODE_TBL(attrs)   AREF (attrs, coding_attr_decode_tbl)
 169 #define CODING_ATTR_ENCODE_TBL(attrs)   AREF (attrs, coding_attr_encode_tbl)
 170 #define CODING_ATTR_TRANS_TBL(attrs)    AREF (attrs, coding_attr_trans_tbl)
 171 #define CODING_ATTR_POST_READ(attrs)    AREF (attrs, coding_attr_post_read)
 172 #define CODING_ATTR_PRE_WRITE(attrs)    AREF (attrs, coding_attr_pre_write)
 173 #define CODING_ATTR_DEFAULT_CHAR(attrs) AREF (attrs, coding_attr_default_char)
 174 #define CODING_ATTR_FOR_UNIBYTE(attrs)  AREF (attrs, coding_attr_for_unibyte)
 175 #define CODING_ATTR_FLUSHING(attrs)     AREF (attrs, coding_attr_flushing)
 176 #define CODING_ATTR_PLIST(attrs)        AREF (attrs, coding_attr_plist)
 177 #define CODING_ATTR_CATEGORY(attrs)     AREF (attrs, coding_attr_category)
 178 #define CODING_ATTR_SAFE_CHARSETS(attrs)AREF (attrs, coding_attr_safe_charsets)
 179
 180
 181 /* Return the name of a coding system specified by ID.  */
 182 #define CODING_ID_NAME(id) \
 183   (HASH_KEY (XHASH_TABLE (Vcoding_system_hash_table), id))
 184
 185 /* Return the attribute vector of a coding system specified by ID.  */
 186
 187 #define CODING_ID_ATTRS(id)     \
 188   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 0))
 189
 190 /* Return the list of aliases of a coding system specified by ID.  */
 191
 192 #define CODING_ID_ALIASES(id)   \
 193   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 1))
 194
 195 /* Return the eol-type of a coding system specified by ID.  */
 196
 197 #define CODING_ID_EOL_TYPE(id)  \
 198   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 2))
 199
 200
 201 /* Return the spec vector of CODING_SYSTEM_SYMBOL.  */
 202
 203 #define CODING_SYSTEM_SPEC(coding_system_symbol)        \
 204   (Fgethash (coding_system_symbol, Vcoding_system_hash_table, Qnil))
 205
 206
 207 /* Return the ID of CODING_SYSTEM_SYMBOL.  */
 208
 209 #define CODING_SYSTEM_ID(coding_system_symbol)                  \
 210   hash_lookup (XHASH_TABLE (Vcoding_system_hash_table),         \
 211                coding_system_symbol, NULL)
 212
 213 /* Return 1 if CODING_SYSTEM_SYMBOL is a coding system.  */
 214
 215 #define CODING_SYSTEM_P(coding_system_symbol)           \
 216   (CODING_SYSTEM_ID (coding_system_symbol) >= 0         \
 217    || (! NILP (coding_system_symbol)                    \
 218        && ! NILP (Fcoding_system_p (coding_system_symbol))))
 219
 220 /* Check if X is a coding system or not.  */
 221
 222 #define CHECK_CODING_SYSTEM(x)                          \
 223   do {                                                  \
 224     if (CODING_SYSTEM_ID (x) < 0                        \
 225         && NILP (Fcheck_coding_system (x)))             \
 226       wrong_type_argument (Qcoding_system_p, (x));      \
 227   } while (0)
 228
 229
 230 /* Check if X is a coding system or not.  If it is, set SEPC to the
 231    spec vector of the coding system.  */
 232
 233 #define CHECK_CODING_SYSTEM_GET_SPEC(x, spec)           \
 234   do {                                                  \
 235     spec = CODING_SYSTEM_SPEC (x);                      \
 236     if (NILP (spec))                                    \
 237       {                                                 \
 238         Fcheck_coding_system (x);                       \
 239         spec = CODING_SYSTEM_SPEC (x);                  \
 240       }                                                 \
 241     if (NILP (spec))                                    \
 242       wrong_type_argument (Qcoding_system_p, (x));      \
 243   } while (0)
 244
 245
 246 /* Check if X is a coding system or not.  If it is, set ID to the
 247    ID of the coding system.  */
 248
 249 #define CHECK_CODING_SYSTEM_GET_ID(x, id)                       \
 250   do                                                            \
 251     {                                                           \
 252       id = CODING_SYSTEM_ID (x);                                \
 253       if (id < 0)                                               \
 254         {                                                       \
 255           Fcheck_coding_system (x);                             \
 256           id = CODING_SYSTEM_ID (x);                            \
 257         }                                                       \
 258       if (id < 0)                                               \
 259         wrong_type_argument (Qcoding_system_p, (x));    \
 260     } while (0)
 261
 262
 263 /*** GENERAL section ***/
 264
 265 /* Enumeration of result code of code conversion.  */
 266 enum coding_result_code
 267   {
 268     CODING_RESULT_SUCCESS,
 269     CODING_RESULT_INSUFFICIENT_SRC,
 270     CODING_RESULT_INSUFFICIENT_DST,
 271     CODING_RESULT_INCONSISTENT_EOL,
 272     CODING_RESULT_INVALID_SRC,
 273     CODING_RESULT_INTERRUPT,
 274     CODING_RESULT_INSUFFICIENT_MEM
 275   };
 276
 277
 278 /* Macros used for the member `mode' of the struct coding_system.  */
 279
 280 /* If set, recover the original CR or LF of the already decoded text
 281    when the decoding routine encounters an inconsistent eol format.  */
 282 #define CODING_MODE_INHIBIT_INCONSISTENT_EOL    0x01
 283
 284 /* If set, the decoding/encoding routines treat the current data as
 285    the last block of the whole text to be converted, and do the
 286    appropriate finishing job.  */
 287 #define CODING_MODE_LAST_BLOCK                  0x02
 288
 289 /* If set, it means that the current source text is in a buffer which
 290    enables selective display.  */
 291 #define CODING_MODE_SELECTIVE_DISPLAY           0x04
 292
 293 /* This flag is used by the decoding/encoding routines on the fly.  If
 294    set, it means that right-to-left text is being processed.  */
 295 #define CODING_MODE_DIRECTION                   0x08
 296
 297 #define CODING_MODE_FIXED_DESTINATION           0x10
 298
 299 /* If set, it means that the encoding routines produces some safe
 300    ASCII characters (usually '?') for unsupported characters.  */
 301 #define CODING_MODE_SAFE_ENCODING               0x20
 302
 303 /* Structure of the field `spec.iso_2022' in the structure
 304    `coding_system'.  */
 305 struct iso_2022_spec
 306 {
 307   /* Bit-wise-or of CODING_ISO_FLAG_XXX.  */
 308   unsigned flags;
 309
 310   /* The current graphic register invoked to each graphic plane.  */
 311   int current_invocation[2];
 312
 313   /* The current charset designated to each graphic register.  The
 314      value -1 means that not charset is designated, -2 means that
 315      there was an invalid designation previously.  */
 316   int current_designation[4];
 317
 318   /* Set to 1 temporarily only when graphic register 2 or 3 is invoked
 319      by single-shift while encoding.  */
 320   int single_shifting;
 321
 322   /* Set to 1 temporarily only when processing at beginning of line.  */
 323   int bol;
 324 };
 325
 326 struct ccl_spec;
 327
 328 enum utf_16_bom_type
 329   {
 330     utf_16_detect_bom,
 331     utf_16_without_bom,
 332     utf_16_with_bom
 333   };
 334
 335 enum utf_16_endian_type
 336   {
 337     utf_16_big_endian,
 338     utf_16_little_endian
 339   };
 340
 341 struct utf_16_spec
 342 {
 343   enum utf_16_bom_type bom;
 344   enum utf_16_endian_type endian;
 345   int surrogate;
 346 };
 347
 348 struct coding_detection_info
 349 {
 350   /* Values of these members are bitwise-OR of CATEGORY_MASK_XXXs.  */
 351   /* Which categories are already checked.  */
 352   int checked;
 353   /* Which categories are strongly found.  */
 354   int found;
 355   /* Which categories are rejected.  */
 356   int rejected;
 357 };
 358
 359
 360 struct coding_system
 361 {
 362   /* ID number of the coding system.  This is an index to
 363      Vcoding_system_hash_table.  This value is set by
 364      setup_coding_system.  At the early stage of building time, this
 365      value is -1 in the array coding_categories to indicate that no
 366      coding-system of that category is yet defined.  */
 367   int id;
 368
 369   /* Flag bits of the coding system.  The meaning of each bit is common
 370      to all types of coding systems.  */
 371   int common_flags;
 372
 373   /* Mode bits of the coding system.  See the comments of the macros
 374      CODING_MODE_XXX.  */
 375   unsigned int mode;
 376
 377   /* Detailed information specific to each type of coding system.  */
 378   union
 379     {
 380       struct iso_2022_spec iso_2022;
 381       struct ccl_spec *ccl;     /* Defined in ccl.h.  */
 382       struct utf_16_spec utf_16;
 383       int emacs_mule_full_support;
 384     } spec;
 385
 386   int max_charset_id;
 387   char *safe_charsets;
 388
 389   /* The following two members specify how binary 8-bit code 128..255
 390      are represented in source and destination text respectively.  1
 391      means they are represented by 2-byte sequence, 0 means they are
 392      represented by 1-byte as is (see the comment in character.h).  */
 393   unsigned src_multibyte : 1;
 394   unsigned dst_multibyte : 1;
 395
 396   /* How may heading bytes we can skip for decoding.  This is set to
 397      -1 in setup_coding_system, and updated by detect_coding.  So,
 398      when this is equal to the byte length of the text being
 399      converted, we can skip the actual conversion process.  */
 400   int head_ascii;
 401
 402   /* The following members are set by encoding/decoding routine.  */
 403   EMACS_INT produced, produced_char, consumed, consumed_char;
 404
 405   /* Number of error source data found in a decoding routine.  */
 406   int errors;
 407
 408   /* Store the positions of error source data. */
 409   EMACS_INT *error_positions;
 410
 411   /* Finish status of code conversion.  */
 412   enum coding_result_code result;
 413
 414   EMACS_INT src_pos, src_pos_byte, src_chars, src_bytes;
 415   Lisp_Object src_object;
 416   const unsigned char *source;
 417
 418   EMACS_INT dst_pos, dst_pos_byte, dst_bytes;
 419   Lisp_Object dst_object;
 420   unsigned char *destination;
 421
 422   /* Set to 1 if the source of conversion is not in the member
 423      `charbuf', but at `src_object'.  */
 424   int chars_at_source;
 425
 426   /* If an element is non-negative, it is a character code.
 427
 428      If it is in the range -128..-1, it is a 8-bit character code
 429      minus 256.
 430
 431      If it is less than -128, it specifies the start of an annotation
 432      chunk.  The length of the chunk is -128 minus the value of the
 433      element.  The following elements are OFFSET, ANNOTATION-TYPE, and
 434      a sequence of actual data for the annotation.  OFFSET is a
 435      character position offset from dst_pos or src_pos,
 436      ANNOTATION-TYPE specfies the meaning of the annotation and how to
 437      handle the following data..  */
 438   int *charbuf;
 439   int charbuf_size, charbuf_used;
 440
 441   /* Set to 1 if charbuf contains an annotation.  */
 442   int annotated;
 443
 444   unsigned char carryover[64];
 445   int carryover_bytes;
 446
 447   int default_char;
 448
 449   int (*detector) P_ ((struct coding_system *,
 450                        struct coding_detection_info *));
 451   void (*decoder) P_ ((struct coding_system *));
 452   int (*encoder) P_ ((struct coding_system *));
 453 };
 454
 455 /* Meanings of bits in the member `common_flags' of the structure
 456    coding_system.  The lowest 8 bits are reserved for various kind of
 457    annotations (currently two of them are used).  */
 458 #define CODING_ANNOTATION_MASK                  0x00FF
 459 #define CODING_ANNOTATE_COMPOSITION_MASK        0x0001
 460 #define CODING_ANNOTATE_DIRECTION_MASK          0x0002
 461 #define CODING_ANNOTATE_CHARSET_MASK            0x0003
 462 #define CODING_FOR_UNIBYTE_MASK                 0x0100
 463 #define CODING_REQUIRE_FLUSHING_MASK            0x0200
 464 #define CODING_REQUIRE_DECODING_MASK            0x0400
 465 #define CODING_REQUIRE_ENCODING_MASK            0x0800
 466 #define CODING_REQUIRE_DETECTION_MASK           0x1000
 467 #define CODING_RESET_AT_BOL_MASK                0x2000
 468
 469 /* Return 1 if the coding context CODING requires annotaion
 470    handling.  */
 471 #define CODING_REQUIRE_ANNOTATION(coding) \
 472   ((coding)->common_flags & CODING_ANNOTATION_MASK)
 473
 474 /* Return 1 if the coding context CODING prefers decoding into unibyte.  */
 475 #define CODING_FOR_UNIBYTE(coding) \
 476   ((coding)->common_flags & CODING_FOR_UNIBYTE_MASK)
 477
 478 /* Return 1 if the coding context CODING requires specific code to be
 479    attached at the tail of converted text.  */
 480 #define CODING_REQUIRE_FLUSHING(coding) \
 481   ((coding)->common_flags & CODING_REQUIRE_FLUSHING_MASK)
 482
 483 /* Return 1 if the coding context CODING requires code conversion on
 484    decoding.  */
 485 #define CODING_REQUIRE_DECODING(coding) \
 486   ((coding)->dst_multibyte              \
 487    || (coding)->common_flags & CODING_REQUIRE_DECODING_MASK)
 488
 489
 490 /* Return 1 if the coding context CODING requires code conversion on
 491    encoding.
 492    The non-multibyte part of the condition is to support encoding of
 493    unibyte strings/buffers generated by string-as-unibyte or
 494    (set-buffer-multibyte nil) from multibyte strings/buffers.  */
 495 #define CODING_REQUIRE_ENCODING(coding)                         \
 496   ((coding)->src_multibyte                                      \
 497    || (coding)->common_flags & CODING_REQUIRE_ENCODING_MASK     \
 498    || (coding)->mode & CODING_MODE_SELECTIVE_DISPLAY)
 499
 500
 501 /* Return 1 if the coding context CODING requires some kind of code
 502    detection.  */
 503 #define CODING_REQUIRE_DETECTION(coding) \
 504   ((coding)->common_flags & CODING_REQUIRE_DETECTION_MASK)
 505
 506 /* Return 1 if the coding context CODING requires code conversion on
 507    decoding or some kind of code detection.  */
 508 #define CODING_MAY_REQUIRE_DECODING(coding)     \
 509   (CODING_REQUIRE_DECODING (coding)             \
 510    || CODING_REQUIRE_DETECTION (coding))
 511
 512 /* Macros to decode or encode a character of JISX0208 in SJIS.  S1 and
 513    S2 are the 1st and 2nd position-codes of JISX0208 in SJIS coding
 514    system.  C1 and C2 are the 1st and 2nd position codes of Emacs'
 515    internal format.  */
 516
 517 #define SJIS_TO_JIS(code)                               \
 518   do {                                                  \
 519     int s1, s2, j1, j2;                                 \
 520                                                         \
 521     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 522                                                         \
 523     if (s2 >= 0x9F)                                     \
 524       (j1 = s1 * 2 - (s1 >= 0xE0 ? 0x160 : 0xE0),       \
 525        j2 = s2 - 0x7E);                                 \
 526     else                                                \
 527       (j1 = s1 * 2 - ((s1 >= 0xE0) ? 0x161 : 0xE1),     \
 528        j2 = s2 - ((s2 >= 0x7F) ? 0x20 : 0x1F));         \
 529     (code) = (j1 << 8) | j2;                            \
 530   } while (0)
 531
 532 #define SJIS_TO_JIS2(code)                              \
 533   do {                                                  \
 534     int s1, s2, j1, j2;                                 \
 535                                                         \
 536     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 537                                                         \
 538     if (s2 >= 0x9F)                                     \
 539       {                                                 \
 540         j1 = (s1 == 0xF0 ? 0x28                         \
 541               : s1 == 0xF1 ? 0x24                       \
 542               : s1 == 0xF2 ? 0x2C                       \
 543               : s1 == 0xF3 ? 0x2E                       \
 544               : 0x6E + (s1 - 0xF4) * 2);                \
 545         j2 = s2 - 0x7E;                                 \
 546       }                                                 \
 547     else                                                \
 548       {                                                 \
 549         j1 = (s1 <= 0xF2 ? 0x21 + (s1 - 0xF0) * 2       \
 550               : s1 <= 0xF4 ? 0x2D + (s1 - 0xF3) * 2     \
 551               : 0x6F + (s1 - 0xF5) * 2);                \
 552         j2 = s2 - ((s2 >= 0x7F ? 0x20 : 0x1F));         \
 553       }                                                 \
 554     (code) = (j1 << 8) | j2;                            \
 555   } while (0)
 556
 557
 558 #define JIS_TO_SJIS(code)                               \
 559   do {                                                  \
 560     int s1, s2, j1, j2;                                 \
 561                                                         \
 562     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 563     if (j1 & 1)                                         \
 564       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x71 : 0xB1),       \
 565        s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F));         \
 566     else                                                \
 567       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x70 : 0xB0),       \
 568        s2 = j2 + 0x7E);                                 \
 569     (code) = (s1 << 8) | s2;                            \
 570   } while (0)
 571
 572 #define JIS_TO_SJIS2(code)                              \
 573   do {                                                  \
 574     int s1, s2, j1, j2;                                 \
 575                                                         \
 576     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 577     if (j1 & 1)                                         \
 578       {                                                 \
 579         s1 = (j1 <= 0x25 ? 0xF0 + (j1 - 0x21) / 2       \
 580               : j1 <= 0x27 ? 0xF3 + (j1 - 0x2D) / 2     \
 581               : 0xF5 + (j1 - 0x6F) / 2);                \
 582         s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F);         \
 583       }                                                 \
 584     else                                                \
 585       {                                                 \
 586         s1 = (j1 == 0x28 ? 0xF0                         \
 587               : j1 == 0x24 ? 0xF1                       \
 588               : j1 == 0x2C ? 0xF2                       \
 589               : j1 == 0x2E ? 0xF3                       \
 590               : 0xF4 + (j1 - 0x6E) / 2);                \
 591         s2 = j2 + 0x7E;                                 \
 592       }                                                 \
 593     (code) = (s1 << 8) | s2;                            \
 594   } while (0)
 595
 596 /* Encode the file name NAME using the specified coding system
 597    for file names, if any.  */
 598 #define ENCODE_FILE(name)                                                  \
 599   (! NILP (Vfile_name_coding_system)                                       \
 600    && !EQ (Vfile_name_coding_system, make_number (0))                      \
 601    ? code_convert_string_norecord (name, Vfile_name_coding_system, 1)      \
 602    : (! NILP (Vdefault_file_name_coding_system)                            \
 603       && !EQ (Vdefault_file_name_coding_system, make_number (0))           \
 604       ? code_convert_string_norecord (name, Vdefault_file_name_coding_system, 1) \
 605       : name))
 606
 607
 608 /* Decode the file name NAME using the specified coding system
 609    for file names, if any.  */
 610 #define DECODE_FILE(name)                                                  \
 611   (! NILP (Vfile_name_coding_system)                                       \
 612    && !EQ (Vfile_name_coding_system, make_number (0))                      \
 613    ? code_convert_string_norecord (name, Vfile_name_coding_system, 0)      \
 614    : (! NILP (Vdefault_file_name_coding_system)                            \
 615       && !EQ (Vdefault_file_name_coding_system, make_number (0))           \
 616       ? code_convert_string_norecord (name, Vdefault_file_name_coding_system, 0) \
 617       : name))
 618
 619
 620 /* Encode the string STR using the specified coding system
 621    for system functions, if any.  */
 622 #define ENCODE_SYSTEM(str)                                                 \
 623   (! NILP (Vlocale_coding_system)                                          \
 624    && !EQ (Vlocale_coding_system, make_number (0))                         \
 625    ? code_convert_string_norecord (str, Vlocale_coding_system, 1)          \
 626    : str)
 627
 628 /* Decode the string STR using the specified coding system
 629    for system functions, if any.  */
 630 #define DECODE_SYSTEM(str)                                                 \
 631   (! NILP (Vlocale_coding_system)                                          \
 632    && !EQ (Vlocale_coding_system, make_number (0))                         \
 633    ? code_convert_string_norecord (str, Vlocale_coding_system, 0)          \
 634    : str)
 635
 636 /* Used by the gtk menu code.  Note that this encodes utf-8, not
 637    utf-8-emacs, so it's not a no-op.  */
 638 #define ENCODE_UTF_8(str) code_convert_string_norecord (str, Qutf_8, 1)
 639
 640 /* Extern declarations.  */
 641 extern Lisp_Object code_conversion_save P_ ((int, int));
 642 extern int decoding_buffer_size P_ ((struct coding_system *, int));
 643 extern int encoding_buffer_size P_ ((struct coding_system *, int));
 644 extern void setup_coding_system P_ ((Lisp_Object, struct coding_system *));
 645 extern Lisp_Object coding_charset_list P_ ((struct coding_system *));
 646 extern void detect_coding P_ ((struct coding_system *));
 647 extern Lisp_Object code_convert_region P_ ((Lisp_Object, Lisp_Object,
 648                                             Lisp_Object, Lisp_Object,
 649                                             int, int));
 650 extern Lisp_Object code_convert_string P_ ((Lisp_Object, Lisp_Object,
 651                                             Lisp_Object, int, int, int));
 652 extern Lisp_Object code_convert_string_norecord P_ ((Lisp_Object, Lisp_Object,
 653                                                      int));
 654 extern Lisp_Object raw_text_coding_system P_ ((Lisp_Object));
 655 extern Lisp_Object coding_inherit_eol_type P_ ((Lisp_Object, Lisp_Object));
 656
 657 extern int decode_coding_gap P_ ((struct coding_system *,
 658                                   EMACS_INT, EMACS_INT));
 659 extern int encode_coding_gap P_ ((struct coding_system *,
 660                                   EMACS_INT, EMACS_INT));
 661 extern void decode_coding_object P_ ((struct coding_system *,
 662                                       Lisp_Object, EMACS_INT, EMACS_INT,
 663                                       EMACS_INT, EMACS_INT, Lisp_Object));
 664 extern void encode_coding_object P_ ((struct coding_system *,
 665                                       Lisp_Object, EMACS_INT, EMACS_INT,
 666                                       EMACS_INT, EMACS_INT, Lisp_Object));
 667
 668 /* Macros for backward compatibility.  */
 669
 670 #define decode_coding_region(coding, from, to)          \
 671   decode_coding_object (coding, Fcurrent_buffer (),     \
 672                         from, CHAR_TO_BYTE (from),      \
 673                         to, CHAR_TO_BYTE (to), Fcurrent_buffer ())
 674
 675
 676 #define encode_coding_region(coding, from, to)          \
 677   encode_coding_object (coding, Fcurrent_buffer (),     \
 678                         from, CHAR_TO_BYTE (from),      \
 679                         to, CHAR_TO_BYTE (to), Fcurrent_buffer ())
 680
 681
 682 #define decode_coding_string(coding, string, nocopy)                    \
 683   decode_coding_object (coding, string, 0, 0, XSTRING (string)->size,   \
 684                         STRING_BYTES (XSTRING (string)), Qt)
 685
 686 #define encode_coding_string(coding, string, nocopy)                    \
 687   (encode_coding_object (coding, string, 0, 0, XSTRING (string)->size,  \
 688                          STRING_BYTES (XSTRING (string)), Qt),          \
 689    (coding)->dst_object)
 690
 691
 692 #define decode_coding_c_string(coding, src, bytes, dst_object)          \
 693   do {                                                                  \
 694     (coding)->source = (src);                                           \
 695     (coding)->src_chars = (coding)->src_bytes = (bytes);                \
 696     decode_coding_object ((coding), Qnil, 0, 0, (bytes), (bytes),       \
 697                           (dst_object));                                \
 698   } while (0)
 699
 700
 701 extern Lisp_Object preferred_coding_system P_ (());
 702
 703
 704 extern Lisp_Object Qutf_8, Qutf_8_emacs;
 705
 706 extern Lisp_Object Qcoding_system, Qeol_type, Qcoding_category_index;
 707 extern Lisp_Object Qcoding_system_p;
 708 extern Lisp_Object Qraw_text, Qemacs_mule, Qno_conversion, Qundecided;
 709 extern Lisp_Object Qiso_2022;
 710 extern Lisp_Object Qbuffer_file_coding_system;
 711
 712 extern Lisp_Object Qunix, Qdos, Qmac;
 713
 714 extern Lisp_Object Qtranslation_table;
 715 extern Lisp_Object Qtranslation_table_id;
 716
 717 /* Mnemonic strings to indicate each type of end-of-line.  */
 718 extern Lisp_Object eol_mnemonic_unix, eol_mnemonic_dos, eol_mnemonic_mac;
 719 /* Mnemonic string to indicate type of end-of-line is not yet decided.  */
 720 extern Lisp_Object eol_mnemonic_undecided;
 721
 722 #ifdef emacs
 723 extern Lisp_Object Qfile_coding_system;
 724 extern Lisp_Object Qcall_process, Qcall_process_region;
 725 extern Lisp_Object Qstart_process, Qopen_network_stream;
 726 extern Lisp_Object Qwrite_region;
 727
 728 extern char *emacs_strerror P_ ((int));
 729
 730 /* Coding-system for reading files and receiving data from process.  */
 731 extern Lisp_Object Vcoding_system_for_read;
 732 /* Coding-system for writing files and sending data to process.  */
 733 extern Lisp_Object Vcoding_system_for_write;
 734 /* Coding-system actually used in the latest I/O.  */
 735 extern Lisp_Object Vlast_coding_system_used;
 736 /* Coding-system to use with system messages (e.g. strerror).  */
 737 extern Lisp_Object Vlocale_coding_system;
 738
 739 /* If non-zero, process buffer inherits the coding system used to decode
 740    the subprocess output.  */
 741 extern int inherit_process_coding_system;
 742
 743 /* Coding system to be used to encode text for terminal display when
 744    terminal coding system is nil.  */
 745 extern struct coding_system safe_terminal_coding;
 746
 747 /* Default coding systems used for process I/O.  */
 748 extern Lisp_Object Vdefault_process_coding_system;
 749
 750 /* Function to call to force a user to force select a propert coding
 751    system.  */
 752 extern Lisp_Object Vselect_safe_coding_system_function;
 753
 754 /* If nonzero, on writing a file, Vselect_safe_coding_system_function
 755    is called even if Vcoding_system_for_write is non-nil.  */
 756 extern int coding_system_require_warning;
 757
 758 /* Coding system for file names, or nil if none.  */
 759 extern Lisp_Object Vfile_name_coding_system;
 760
 761 /* Coding system for file names used only when
 762    Vfile_name_coding_system is nil.  */
 763 extern Lisp_Object Vdefault_file_name_coding_system;
 764
 765 #endif
 766
 767 /* Error signaled when there's a problem with detecting coding system */
 768 extern Lisp_Object Qcoding_system_error;
 769
 770 extern char emacs_mule_bytes[256];
 771 extern int emacs_mule_string_char P_ ((unsigned char *));
 772
 773 #endif /* EMACS_CODING_H */
 774
 775 /* arch-tag: 2bc3b4fa-6870-4f64-8135-b962b2d290e4
 776    (do not change this comment) */