code.delx.au - gnu-emacs/blob - src/coding.h

   1 /* Header for coding system handler.
   2    Copyright (C) 1995, 1997 Electrotechnical Laboratory, JAPAN.
   3      Licensed to the Free Software Foundation.
   4    Copyright (C) 2003
   5      National Institute of Advanced Industrial Science and Technology (AIST)
   6      Registration Number H13PRO009
   7
   8 This file is part of GNU Emacs.
   9
  10 GNU Emacs is free software; you can redistribute it and/or modify
  11 it under the terms of the GNU General Public License as published by
  12 the Free Software Foundation; either version 2, or (at your option)
  13 any later version.
  14
  15 GNU Emacs is distributed in the hope that it will be useful,
  16 but WITHOUT ANY WARRANTY; without even the implied warranty of
  17 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  18 GNU General Public License for more details.
  19
  20 You should have received a copy of the GNU General Public License
  21 along with GNU Emacs; see the file COPYING.  If not, write to
  22 the Free Software Foundation, Inc., 59 Temple Place - Suite 330,
  23 Boston, MA 02111-1307, USA.  */
  24
  25 #ifndef EMACS_CODING_H
  26 #define EMACS_CODING_H
  27
  28 /* Index to arguments of Fdefine_coding_system_internal.  */
  29
  30 enum define_coding_system_arg_index
  31   {
  32     coding_arg_name,
  33     coding_arg_mnemonic,
  34     coding_arg_coding_type,
  35     coding_arg_charset_list,
  36     coding_arg_ascii_compatible_p,
  37     coding_arg_decode_translation_table,
  38     coding_arg_encode_translation_table,
  39     coding_arg_post_read_conversion,
  40     coding_arg_pre_write_conversion,
  41     coding_arg_default_char,
  42     coding_arg_for_unibyte,
  43     coding_arg_plist,
  44     coding_arg_eol_type,
  45     coding_arg_max
  46   };
  47
  48 enum define_coding_iso2022_arg_index
  49   {
  50     coding_arg_iso2022_initial = coding_arg_max,
  51     coding_arg_iso2022_reg_usage,
  52     coding_arg_iso2022_request,
  53     coding_arg_iso2022_flags,
  54     coding_arg_iso2022_max
  55   };
  56
  57 enum define_coding_utf16_arg_index
  58   {
  59     coding_arg_utf16_bom = coding_arg_max,
  60     coding_arg_utf16_endian,
  61     coding_arg_utf16_max
  62   };
  63
  64 enum define_coding_ccl_arg_index
  65   {
  66     coding_arg_ccl_decoder = coding_arg_max,
  67     coding_arg_ccl_encoder,
  68     coding_arg_ccl_valids,
  69     coding_arg_ccl_max
  70   };
  71
  72 /* Hash table for all coding systems.  Keys are coding system symbols
  73    and values are spec vectors of the corresponding coding system.  A
  74    spec vector has the form [ ATTRS ALIASES EOL-TYPE ].  ATTRS is a
  75    vector of attribute of the coding system.  ALIASES is a list of
  76    aliases (symbols) of the coding system.  EOL-TYPE is `unix', `dos',
  77    `mac' or a vector of coding systems (symbols).  */
  78
  79 extern Lisp_Object Vcoding_system_hash_table;
  80
  81
  82 /* Enumeration of coding system type.  */
  83
  84 enum coding_system_type
  85   {
  86     coding_type_charset,
  87     coding_type_utf_8,
  88     coding_type_utf_16,
  89     coding_type_iso_2022,
  90     coding_type_emacs_mule,
  91     coding_type_sjis,
  92     coding_type_ccl,
  93     coding_type_raw_text,
  94     coding_type_undecided,
  95     coding_type_max
  96   };
  97
  98
  99 /* Enumeration of end-of-line format type.  */
 100
 101 enum end_of_line_type
 102   {
 103     eol_lf,             /* Line-feed only, same as Emacs' internal
 104                            format.  */
 105     eol_crlf,           /* Sequence of carriage-return and
 106                            line-feed.  */
 107     eol_cr,             /* Carriage-return only.  */
 108     eol_any,            /* Accept any of above.  Produce line-feed
 109                            only.  */
 110     eol_undecided,      /* This value is used to denote that the
 111                            eol-type is not yet undecided.  */
 112     eol_type_max
 113   };
 114
 115 /* Enumeration of index to an attribute vector of a coding system.  */
 116
 117 enum coding_attr_index
 118   {
 119     coding_attr_base_name,
 120     coding_attr_docstring,
 121     coding_attr_mnemonic,
 122     coding_attr_type,
 123     coding_attr_charset_list,
 124     coding_attr_ascii_compat,
 125     coding_attr_decode_tbl,
 126     coding_attr_encode_tbl,
 127     coding_attr_trans_tbl,
 128     coding_attr_post_read,
 129     coding_attr_pre_write,
 130     coding_attr_default_char,
 131     coding_attr_for_unibyte,
 132     coding_attr_plist,
 133
 134     coding_attr_category,
 135     coding_attr_safe_charsets,
 136
 137     /* The followings are extra attributes for each type.  */
 138     coding_attr_charset_valids,
 139
 140     coding_attr_ccl_decoder,
 141     coding_attr_ccl_encoder,
 142     coding_attr_ccl_valids,
 143
 144     coding_attr_iso_initial,
 145     coding_attr_iso_usage,
 146     coding_attr_iso_request,
 147     coding_attr_iso_flags,
 148
 149     coding_attr_utf_16_bom,
 150     coding_attr_utf_16_endian,
 151
 152     coding_attr_emacs_mule_full,
 153
 154     coding_attr_last_index
 155   };
 156
 157
 158 /* Macros to access an element of an attribute vector.  */
 159
 160 #define CODING_ATTR_BASE_NAME(attrs)    AREF (attrs, coding_attr_base_name)
 161 #define CODING_ATTR_TYPE(attrs)         AREF (attrs, coding_attr_type)
 162 #define CODING_ATTR_CHARSET_LIST(attrs) AREF (attrs, coding_attr_charset_list)
 163 #define CODING_ATTR_MNEMONIC(attrs)     AREF (attrs, coding_attr_mnemonic)
 164 #define CODING_ATTR_DOCSTRING(attrs)    AREF (attrs, coding_attr_docstring)
 165 #define CODING_ATTR_ASCII_COMPAT(attrs) AREF (attrs, coding_attr_ascii_compat)
 166 #define CODING_ATTR_DECODE_TBL(attrs)   AREF (attrs, coding_attr_decode_tbl)
 167 #define CODING_ATTR_ENCODE_TBL(attrs)   AREF (attrs, coding_attr_encode_tbl)
 168 #define CODING_ATTR_TRANS_TBL(attrs)    AREF (attrs, coding_attr_trans_tbl)
 169 #define CODING_ATTR_POST_READ(attrs)    AREF (attrs, coding_attr_post_read)
 170 #define CODING_ATTR_PRE_WRITE(attrs)    AREF (attrs, coding_attr_pre_write)
 171 #define CODING_ATTR_DEFAULT_CHAR(attrs) AREF (attrs, coding_attr_default_char)
 172 #define CODING_ATTR_FOR_UNIBYTE(attrs)  AREF (attrs, coding_attr_for_unibyte)
 173 #define CODING_ATTR_FLUSHING(attrs)     AREF (attrs, coding_attr_flushing)
 174 #define CODING_ATTR_PLIST(attrs)        AREF (attrs, coding_attr_plist)
 175 #define CODING_ATTR_CATEGORY(attrs)     AREF (attrs, coding_attr_category)
 176 #define CODING_ATTR_SAFE_CHARSETS(attrs)AREF (attrs, coding_attr_safe_charsets)
 177
 178
 179 /* Return the name of a coding system specified by ID.  */
 180 #define CODING_ID_NAME(id) \
 181   (HASH_KEY (XHASH_TABLE (Vcoding_system_hash_table), id))
 182
 183 /* Return the attribute vector of a coding system specified by ID.  */
 184
 185 #define CODING_ID_ATTRS(id)     \
 186   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 0))
 187
 188 /* Return the list of aliases of a coding system specified by ID.  */
 189
 190 #define CODING_ID_ALIASES(id)   \
 191   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 1))
 192
 193 /* Return the eol-type of a coding system specified by ID.  */
 194
 195 #define CODING_ID_EOL_TYPE(id)  \
 196   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 2))
 197
 198
 199 /* Return the spec vector of CODING_SYSTEM_SYMBOL.  */
 200
 201 #define CODING_SYSTEM_SPEC(coding_system_symbol)        \
 202   (Fgethash (coding_system_symbol, Vcoding_system_hash_table, Qnil))
 203
 204
 205 /* Return the ID of CODING_SYSTEM_SYMBOL.  */
 206
 207 #define CODING_SYSTEM_ID(coding_system_symbol)                  \
 208   hash_lookup (XHASH_TABLE (Vcoding_system_hash_table),         \
 209                coding_system_symbol, NULL)
 210
 211 /* Return 1 iff CODING_SYSTEM_SYMBOL is a coding system.  */
 212
 213 #define CODING_SYSTEM_P(coding_system_symbol)   \
 214   (! NILP (CODING_SYSTEM_SPEC (coding_system_symbol)))
 215
 216 /* Check if X is a coding system or not.  */
 217
 218 #define CHECK_CODING_SYSTEM(x)                          \
 219   do {                                                  \
 220     if (!CODING_SYSTEM_P (x))                           \
 221       wrong_type_argument (Qcoding_system_p, (x));      \
 222   } while (0)
 223
 224
 225 /* Check if X is a coding system or not.  If it is, set SEPC to the
 226    spec vector of the coding system.  */
 227
 228 #define CHECK_CODING_SYSTEM_GET_SPEC(x, spec)           \
 229   do {                                                  \
 230     spec = CODING_SYSTEM_SPEC (x);                      \
 231     if (NILP (spec))                                    \
 232       x = wrong_type_argument (Qcoding_system_p, (x));  \
 233   } while (0)
 234
 235
 236 /* Check if X is a coding system or not.  If it is, set ID to the
 237    ID of the coding system.  */
 238
 239 #define CHECK_CODING_SYSTEM_GET_ID(x, id)                       \
 240   do                                                            \
 241     {                                                           \
 242       id = CODING_SYSTEM_ID (x);                                \
 243       if (id < 0)                                               \
 244         x = wrong_type_argument (Qcoding_system_p, (x));        \
 245     } while (0)
 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_INCONSISTENT_EOL,
 257     CODING_RESULT_INVALID_SRC,
 258     CODING_RESULT_INTERRUPT,
 259     CODING_RESULT_INSUFFICIENT_MEM
 260   };
 261
 262
 263 /* Macros used for the member `mode' of the struct coding_system.  */
 264
 265 /* If set, recover the original CR or LF of the already decoded text
 266    when the decoding routine encounters an inconsistent eol format.  */
 267 #define CODING_MODE_INHIBIT_INCONSISTENT_EOL    0x01
 268
 269 /* If set, the decoding/encoding routines treat the current data as
 270    the last block of the whole text to be converted, and do the
 271    appropriate finishing job.  */
 272 #define CODING_MODE_LAST_BLOCK                  0x02
 273
 274 /* If set, it means that the current source text is in a buffer which
 275    enables selective display.  */
 276 #define CODING_MODE_SELECTIVE_DISPLAY           0x04
 277
 278 /* This flag is used by the decoding/encoding routines on the fly.  If
 279    set, it means that right-to-left text is being processed.  */
 280 #define CODING_MODE_DIRECTION                   0x08
 281
 282 #define CODING_MODE_FIXED_DESTINATION           0x10
 283
 284 /* If set, it means that the encoding routines produces some safe
 285    ASCII characters (usually '?') for unsupported characters.  */
 286 #define CODING_MODE_SAFE_ENCODING               0x20
 287
 288 /* Structure of the field `spec.iso_2022' in the structure
 289    `coding_system'.  */
 290 struct iso_2022_spec
 291 {
 292   /* Bit-wise-or of CODING_ISO_FLAG_XXX.  */
 293   unsigned flags;
 294
 295   /* The current graphic register invoked to each graphic plane.  */
 296   int current_invocation[2];
 297
 298   /* The current charset designated to each graphic register.  The
 299      value -1 means that not charset is designated, -2 means that
 300      there was an invalid designation previously.  */
 301   int current_designation[4];
 302
 303   /* Set to 1 temporarily only when graphic register 2 or 3 is invoked
 304      by single-shift while encoding.  */
 305   int single_shifting;
 306
 307   /* Set to 1 temporarily only when processing at beginning of line.  */
 308   int bol;
 309 };
 310
 311 struct ccl_spec;
 312
 313 enum utf_16_bom_type
 314   {
 315     utf_16_detect_bom,
 316     utf_16_without_bom,
 317     utf_16_with_bom
 318   };
 319
 320 enum utf_16_endian_type
 321   {
 322     utf_16_big_endian,
 323     utf_16_little_endian
 324   };
 325
 326 struct utf_16_spec
 327 {
 328   enum utf_16_bom_type bom;
 329   enum utf_16_endian_type endian;
 330   int surrogate;
 331 };
 332
 333 struct coding_detection_info
 334 {
 335   /* Values of these members are bitwise-OR of CATEGORY_MASK_XXXs.  */
 336   /* Which categories are already checked.  */
 337   int checked;
 338   /* Which categories are strongly found.  */
 339   int found;
 340   /* Which categories are rejected.  */
 341   int rejected;
 342 };
 343
 344
 345 struct coding_system
 346 {
 347   /* ID number of the coding system.  This is an index to
 348      Vcoding_system_hash_table.  This value is set by
 349      setup_coding_system.  At the early stage of building time, this
 350      value is -1 in the array coding_categories to indicate that no
 351      coding-system of that category is yet defined.  */
 352   int id;
 353
 354   /* Flag bits of the coding system.  The meaning of each bit is common
 355      to all types of coding systems.  */
 356   int common_flags;
 357
 358   /* Mode bits of the coding system.  See the comments of the macros
 359      CODING_MODE_XXX.  */
 360   unsigned int mode;
 361
 362   /* Detailed information specific to each type of coding system.  */
 363   union
 364     {
 365       struct iso_2022_spec iso_2022;
 366       struct ccl_spec *ccl;     /* Defined in ccl.h.  */
 367       struct utf_16_spec utf_16;
 368       int emacs_mule_full_support;
 369     } spec;
 370
 371   int max_charset_id;
 372   char *safe_charsets;
 373
 374   /* The following two members specify how binary 8-bit code 128..255
 375      are represented in source and destination text respectively.  1
 376      means they are represented by 2-byte sequence, 0 means they are
 377      represented by 1-byte as is (see the comment in character.h).  */
 378   unsigned src_multibyte : 1;
 379   unsigned dst_multibyte : 1;
 380
 381   /* How may heading bytes we can skip for decoding.  This is set to
 382      -1 in setup_coding_system, and updated by detect_coding.  So,
 383      when this is equal to the byte length of the text being
 384      converted, we can skip the actual conversion process.  */
 385   int head_ascii;
 386
 387   /* The following members are set by encoding/decoding routine.  */
 388   EMACS_INT produced, produced_char, consumed, consumed_char;
 389
 390   /* Number of error source data found in a decoding routine.  */
 391   int errors;
 392
 393   /* Store the positions of error source data. */
 394   EMACS_INT *error_positions;
 395
 396   /* Finish status of code conversion.  */
 397   enum coding_result_code result;
 398
 399   EMACS_INT src_pos, src_pos_byte, src_chars, src_bytes;
 400   Lisp_Object src_object;
 401   const unsigned char *source;
 402
 403   EMACS_INT dst_pos, dst_pos_byte, dst_bytes;
 404   Lisp_Object dst_object;
 405   unsigned char *destination;
 406
 407   /* Set to 1 iff the source of conversion is not in the member
 408      `charbuf', but at `src_object'.  */
 409   int chars_at_source;
 410
 411   /* If an element is non-negative, it is a character code.
 412
 413      If it is in the range -128..-1, it is a 8-bit character code
 414      minus 256.
 415
 416      If it is less than -128, it specifies the start of an annotation
 417      chunk.  The length of the chunk is -128 minus the value of the
 418      element.  The following elements are OFFSET, ANNOTATION-TYPE, and
 419      a sequence of actual data for the annotation.  OFFSET is a
 420      character position offset from dst_pos or src_pos,
 421      ANNOTATION-TYPE specfies the meaning of the annotation and how to
 422      handle the following data..  */
 423   int *charbuf;
 424   int charbuf_size, charbuf_used;
 425
 426   /* Set to 1 if charbuf contains an annotation.  */
 427   int annotated;
 428
 429   unsigned char carryover[64];
 430   int carryover_bytes;
 431
 432   int default_char;
 433
 434   int (*detector) P_ ((struct coding_system *,
 435                        struct coding_detection_info *));
 436   void (*decoder) P_ ((struct coding_system *));
 437   int (*encoder) P_ ((struct coding_system *));
 438 };
 439
 440 /* Meanings of bits in the member `common_flags' of the structure
 441    coding_system.  The lowest 8 bits are reserved for various kind of
 442    annotations (currently two of them are used).  */
 443 #define CODING_ANNOTATION_MASK                  0x00FF
 444 #define CODING_ANNOTATE_COMPOSITION_MASK        0x0001
 445 #define CODING_ANNOTATE_DIRECTION_MASK          0x0002
 446 #define CODING_ANNOTATE_CHARSET_MASK            0x0003
 447 #define CODING_FOR_UNIBYTE_MASK                 0x0100
 448 #define CODING_REQUIRE_FLUSHING_MASK            0x0200
 449 #define CODING_REQUIRE_DECODING_MASK            0x0400
 450 #define CODING_REQUIRE_ENCODING_MASK            0x0800
 451 #define CODING_REQUIRE_DETECTION_MASK           0x1000
 452 #define CODING_RESET_AT_BOL_MASK                0x2000
 453
 454 /* Return 1 if the coding context CODING requires annotaion
 455    handling.  */
 456 #define CODING_REQUIRE_ANNOTATION(coding) \
 457   ((coding)->common_flags & CODING_ANNOTATION_MASK)
 458
 459 /* Return 1 if the coding context CODING prefers decoding into unibyte.  */
 460 #define CODING_FOR_UNIBYTE(coding) \
 461   ((coding)->common_flags & CODING_FOR_UNIBYTE_MASK)
 462
 463 /* Return 1 if the coding context CODING requires specific code to be
 464    attached at the tail of converted text.  */
 465 #define CODING_REQUIRE_FLUSHING(coding) \
 466   ((coding)->common_flags & CODING_REQUIRE_FLUSHING_MASK)
 467
 468 /* Return 1 if the coding context CODING requires code conversion on
 469    decoding.  */
 470 #define CODING_REQUIRE_DECODING(coding) \
 471   ((coding)->dst_multibyte              \
 472    || (coding)->common_flags & CODING_REQUIRE_DECODING_MASK)
 473
 474
 475 /* Return 1 if the coding context CODING requires code conversion on
 476    encoding.  */
 477 #define CODING_REQUIRE_ENCODING(coding)                         \
 478   ((coding)->src_multibyte                                      \
 479    || (coding)->common_flags & CODING_REQUIRE_ENCODING_MASK     \
 480    || (coding)->mode & CODING_MODE_SELECTIVE_DISPLAY)
 481
 482
 483 /* Return 1 if the coding context CODING requires some kind of code
 484    detection.  */
 485 #define CODING_REQUIRE_DETECTION(coding) \
 486   ((coding)->common_flags & CODING_REQUIRE_DETECTION_MASK)
 487
 488 /* Return 1 if the coding context CODING requires code conversion on
 489    decoding or some kind of code detection.  */
 490 #define CODING_MAY_REQUIRE_DECODING(coding)     \
 491   (CODING_REQUIRE_DECODING (coding)             \
 492    || CODING_REQUIRE_DETECTION (coding))
 493
 494 /* Macros to decode or encode a character of JISX0208 in SJIS.  S1 and
 495    S2 are the 1st and 2nd position-codes of JISX0208 in SJIS coding
 496    system.  C1 and C2 are the 1st and 2nd position codes of Emacs'
 497    internal format.  */
 498
 499 #define SJIS_TO_JIS(code)                               \
 500   do {                                                  \
 501     int s1, s2, j1, j2;                                 \
 502                                                         \
 503     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 504                                                         \
 505     if (s2 >= 0x9F)                                     \
 506       (j1 = s1 * 2 - (s1 >= 0xE0 ? 0x160 : 0xE0),       \
 507        j2 = s2 - 0x7E);                                 \
 508     else                                                \
 509       (j1 = s1 * 2 - ((s1 >= 0xE0) ? 0x161 : 0xE1),     \
 510        j2 = s2 - ((s2 >= 0x7F) ? 0x20 : 0x1F));         \
 511     (code) = (j1 << 8) | j2;                            \
 512   } while (0)
 513
 514 #define SJIS_TO_JIS2(code)                              \
 515   do {                                                  \
 516     int s1, s2, j1, j2;                                 \
 517                                                         \
 518     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 519                                                         \
 520     if (s2 >= 0x9F)                                     \
 521       {                                                 \
 522         j1 = (s1 == 0xF0 ? 0x28                         \
 523               : s1 == 0xF1 ? 0x24                       \
 524               : s1 == 0xF2 ? 0x2C                       \
 525               : s1 == 0xF3 ? 0x2E                       \
 526               : 0x6E + (s1 - 0xF4) * 2);                \
 527         j2 = s2 - 0x7E;                                 \
 528       }                                                 \
 529     else                                                \
 530       {                                                 \
 531         j1 = (s1 <= 0xF2 ? 0x21 + (s1 - 0xF0) * 2       \
 532               : s1 <= 0xF4 ? 0x2D + (s1 - 0xF3) * 2     \
 533               : 0x6F + (s1 - 0xF5) * 2);                \
 534         j2 = s2 - ((s2 >= 0x7F ? 0x20 : 0x1F));         \
 535       }                                                 \
 536     (code) = (j1 << 8) | j2;                            \
 537   } while (0)
 538
 539
 540 #define JIS_TO_SJIS(code)                               \
 541   do {                                                  \
 542     int s1, s2, j1, j2;                                 \
 543                                                         \
 544     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 545     if (j1 & 1)                                         \
 546       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x71 : 0xB1),       \
 547        s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F));         \
 548     else                                                \
 549       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x70 : 0xB0),       \
 550        s2 = j2 + 0x7E);                                 \
 551     (code) = (s1 << 8) | s2;                            \
 552   } while (0)
 553
 554 #define JIS_TO_SJIS2(code)                              \
 555   do {                                                  \
 556     int s1, s2, j1, j2;                                 \
 557                                                         \
 558     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 559     if (j1 & 1)                                         \
 560       {                                                 \
 561         s1 = (j1 <= 0x25 ? 0xF0 + (j1 - 0x21) / 2       \
 562               : j1 <= 0x27 ? 0xF3 + (j1 - 0x2D) / 2     \
 563               : 0xF5 + (j1 - 0x6F) / 2);                \
 564         s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F);         \
 565       }                                                 \
 566     else                                                \
 567       {                                                 \
 568         s1 = (j1 == 0x28 ? 0xF0                         \
 569               : j1 == 0x24 ? 0xF1                       \
 570               : j1 == 0x2C ? 0xF2                       \
 571               : j1 == 0x2E ? 0xF3                       \
 572               : 0xF4 + (j1 - 0x6E) / 2);                \
 573         s2 = j2 + 0x7E;                                 \
 574       }                                                 \
 575     (code) = (s1 << 8) | s2;                            \
 576   } while (0)
 577
 578 /* Encode the file name NAME using the specified coding system
 579    for file names, if any.  */
 580 #define ENCODE_FILE(name)                                                  \
 581   (! NILP (Vfile_name_coding_system)                                       \
 582    && XFASTINT (Vfile_name_coding_system) != 0                             \
 583    ? code_convert_string_norecord (name, Vfile_name_coding_system, 1)      \
 584    : (! NILP (Vdefault_file_name_coding_system)                            \
 585       && XFASTINT (Vdefault_file_name_coding_system) != 0                  \
 586       ? code_convert_string_norecord (name, Vdefault_file_name_coding_system, 1) \
 587       : name))
 588
 589
 590 /* Decode the file name NAME using the specified coding system
 591    for file names, if any.  */
 592 #define DECODE_FILE(name)                                                  \
 593   (! NILP (Vfile_name_coding_system)                                       \
 594    && XFASTINT (Vfile_name_coding_system) != 0                             \
 595    ? code_convert_string_norecord (name, Vfile_name_coding_system, 0)      \
 596    : (! NILP (Vdefault_file_name_coding_system)                            \
 597       && XFASTINT (Vdefault_file_name_coding_system) != 0                  \
 598       ? code_convert_string_norecord (name, Vdefault_file_name_coding_system, 0) \
 599       : name))
 600
 601
 602 #ifdef WINDOWSNT
 603 /* Encode the string STR using the specified coding system
 604    for w32 system functions, if any.  */
 605 #define ENCODE_SYSTEM(str)                                              \
 606   (! NILP (Vlocale_coding_system)                                       \
 607    && XFASTINT (Vlocale_coding_system) != 0                             \
 608    ? code_convert_string_norecord (str, Vlocale_coding_system, 1)       \
 609    : str)
 610
 611 /* Decode the string STR using the specified coding system
 612    for w32 system functions, if any.  */
 613 #define DECODE_SYSTEM(name)                                             \
 614   (! NILP (Vlocale_coding_system)                                       \
 615    && XFASTINT (Vlocale_coding_system) != 0                             \
 616    ? code_convert_string_norecord (str, Vlocale_coding_system, 0)       \
 617    : str)
 618
 619 #else /* WINDOWSNT */
 620
 621 #define ENCODE_SYSTEM(str) string_make_unibyte(str)
 622 #define DECODE_SYSTEM(name) name
 623
 624 #endif /* !WINDOWSNT */
 625
 626 /* Used by the gtk menu code.  Note that this encodes utf-8, not
 627    utf-8-emacs, so it's not a no-op.  */
 628 #define ENCODE_UTF_8(str) code_convert_string_norecord (str, Qutf_8, 1)
 629
 630 /* Extern declarations.  */
 631 extern Lisp_Object code_conversion_save P_ ((int, int));
 632 extern int decoding_buffer_size P_ ((struct coding_system *, int));
 633 extern int encoding_buffer_size P_ ((struct coding_system *, int));
 634 extern void setup_coding_system P_ ((Lisp_Object, struct coding_system *));
 635 extern void detect_coding P_ ((struct coding_system *));
 636 extern Lisp_Object code_convert_region P_ ((Lisp_Object, Lisp_Object,
 637                                             Lisp_Object, Lisp_Object,
 638                                             int, int));
 639 extern Lisp_Object code_convert_string P_ ((Lisp_Object, Lisp_Object,
 640                                             Lisp_Object, int, int, int));
 641 extern Lisp_Object code_convert_string_norecord P_ ((Lisp_Object, Lisp_Object,
 642                                                      int));
 643 extern Lisp_Object raw_text_coding_system P_ ((Lisp_Object));
 644 extern Lisp_Object coding_inherit_eol_type P_ ((Lisp_Object, Lisp_Object));
 645
 646 extern int decode_coding_gap P_ ((struct coding_system *,
 647                                   EMACS_INT, EMACS_INT));
 648 extern int encode_coding_gap P_ ((struct coding_system *,
 649                                   EMACS_INT, EMACS_INT));
 650 extern void decode_coding_object P_ ((struct coding_system *,
 651                                       Lisp_Object, EMACS_INT, EMACS_INT,
 652                                       EMACS_INT, EMACS_INT, Lisp_Object));
 653 extern void encode_coding_object P_ ((struct coding_system *,
 654                                       Lisp_Object, EMACS_INT, EMACS_INT,
 655                                       EMACS_INT, EMACS_INT, Lisp_Object));
 656
 657 /* Macros for backward compatibility.  */
 658
 659 #define decode_coding_region(coding, from, to)          \
 660   decode_coding_object (coding, Fcurrent_buffer (),     \
 661                         from, CHAR_TO_BYTE (from),      \
 662                         to, CHAR_TO_BYTE (to), Fcurrent_buffer ())
 663
 664
 665 #define encode_coding_region(coding, from, to)          \
 666   encode_coding_object (coding, Fcurrent_buffer (),     \
 667                         from, CHAR_TO_BYTE (from),      \
 668                         to, CHAR_TO_BYTE (to), Fcurrent_buffer ())
 669
 670
 671 #define decode_coding_string(coding, string, nocopy)                    \
 672   decode_coding_object (coding, string, 0, 0, XSTRING (string)->size,   \
 673                         STRING_BYTES (XSTRING (string)), Qt)
 674
 675 #define encode_coding_string(coding, string, nocopy)                    \
 676   (encode_coding_object (coding, string, 0, 0, XSTRING (string)->size,  \
 677                          STRING_BYTES (XSTRING (string)), Qt),          \
 678    (coding)->dst_object)
 679
 680
 681 #define decode_coding_c_string(coding, src, bytes, dst_object)          \
 682   do {                                                                  \
 683     (coding)->source = (src);                                           \
 684     (coding)->src_chars = (coding)->src_bytes = (bytes);                \
 685     decode_coding_object ((coding), Qnil, 0, 0, (bytes), (bytes),       \
 686                           (dst_object));                                \
 687   } while (0)
 688
 689
 690 extern Lisp_Object preferred_coding_system P_ (());
 691
 692
 693 extern Lisp_Object Qutf_8, Qutf_8_emacs;
 694
 695 extern Lisp_Object Qcoding_system, Qeol_type, Qcoding_category_index;
 696 extern Lisp_Object Qcoding_system_p;
 697 extern Lisp_Object Qraw_text, Qemacs_mule, Qno_conversion, Qundecided;
 698 extern Lisp_Object Qiso_2022;
 699 extern Lisp_Object Qbuffer_file_coding_system;
 700
 701 extern Lisp_Object Qunix, Qdos, Qmac;
 702
 703 extern Lisp_Object Qtranslation_table;
 704 extern Lisp_Object Qtranslation_table_id;
 705
 706 /* Mnemonic strings to indicate each type of end-of-line.  */
 707 extern Lisp_Object eol_mnemonic_unix, eol_mnemonic_dos, eol_mnemonic_mac;
 708 /* Mnemonic string to indicate type of end-of-line is not yet decided.  */
 709 extern Lisp_Object eol_mnemonic_undecided;
 710
 711 #ifdef emacs
 712 extern Lisp_Object Qfile_coding_system;
 713 extern Lisp_Object Qcall_process, Qcall_process_region, Qprocess_argument;
 714 extern Lisp_Object Qstart_process, Qopen_network_stream;
 715 extern Lisp_Object Qwrite_region;
 716
 717 extern char *emacs_strerror P_ ((int));
 718
 719 /* Coding-system for reading files and receiving data from process.  */
 720 extern Lisp_Object Vcoding_system_for_read;
 721 /* Coding-system for writing files and sending data to process.  */
 722 extern Lisp_Object Vcoding_system_for_write;
 723 /* Coding-system actually used in the latest I/O.  */
 724 extern Lisp_Object Vlast_coding_system_used;
 725 /* Coding-system to use with system messages (e.g. strerror).  */
 726 extern Lisp_Object Vlocale_coding_system;
 727
 728 /* If non-zero, process buffer inherits the coding system used to decode
 729    the subprocess output.  */
 730 extern int inherit_process_coding_system;
 731
 732 /* Coding-system to be used for encoding terminal output.  This
 733    structure contains information of a coding-system specified by the
 734    function `set-terminal-coding-system'.  */
 735 extern struct coding_system terminal_coding;
 736
 737 /* Coding system to be used to encode text for terminal display when
 738    terminal coding system is nil.  */
 739 extern struct coding_system safe_terminal_coding;
 740
 741 /* Coding-system of what is sent from terminal keyboard.  This
 742    structure contains information of a coding-system specified by the
 743    function `set-keyboard-coding-system'.  */
 744 extern struct coding_system keyboard_coding;
 745
 746 /* Default coding systems used for process I/O.  */
 747 extern Lisp_Object Vdefault_process_coding_system;
 748
 749 /* Function to call to force a user to force select a propert coding
 750    system.  */
 751 extern Lisp_Object Vselect_safe_coding_system_function;
 752
 753 /* If nonzero, on writing a file, Vselect_safe_coding_system_function
 754    is called even if Vcoding_system_for_write is non-nil.  */
 755 extern int coding_system_require_warning;
 756
 757 /* Coding system for file names, or nil if none.  */
 758 extern Lisp_Object Vfile_name_coding_system;
 759
 760 /* Coding system for file names used only when
 761    Vfile_name_coding_system is nil.  */
 762 extern Lisp_Object Vdefault_file_name_coding_system;
 763
 764 #endif
 765
 766 /* Error signaled when there's a problem with detecting coding system */
 767 extern Lisp_Object Qcoding_system_error;
 768
 769 extern char emacs_mule_bytes[256];
 770 extern int emacs_mule_string_char P_ ((unsigned char *));
 771
 772 #endif /* EMACS_CODING_H */
 773
 774 /* arch-tag: 2bc3b4fa-6870-4f64-8135-b962b2d290e4
 775    (do not change this comment) */