src/coding.h

   1 /* Header for coding system handler.
   2    Copyright (C) 2001, 2002, 2003, 2004, 2005,
   3                  2006, 2007, 2008, 2009  Free Software Foundation, Inc.
   4    Copyright (C) 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004,
   5      2005, 2006, 2007, 2008, 2009
   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_utf8_arg_index
  60   {
  61     coding_arg_utf8_bom = coding_arg_max,
  62     coding_arg_utf8_max
  63   };
  64
  65 enum define_coding_utf16_arg_index
  66   {
  67     coding_arg_utf16_bom = coding_arg_max,
  68     coding_arg_utf16_endian,
  69     coding_arg_utf16_max
  70   };
  71
  72 enum define_coding_ccl_arg_index
  73   {
  74     coding_arg_ccl_decoder = coding_arg_max,
  75     coding_arg_ccl_encoder,
  76     coding_arg_ccl_valids,
  77     coding_arg_ccl_max
  78   };
  79
  80 /* Hash table for all coding systems.  Keys are coding system symbols
  81    and values are spec vectors of the corresponding coding system.  A
  82    spec vector has the form [ ATTRS ALIASES EOL-TYPE ].  ATTRS is a
  83    vector of attribute of the coding system.  ALIASES is a list of
  84    aliases (symbols) of the coding system.  EOL-TYPE is `unix', `dos',
  85    `mac' or a vector of coding systems (symbols).  */
  86
  87 extern Lisp_Object Vcoding_system_hash_table;
  88
  89
  90 /* Enumeration of coding system type.  */
  91
  92 enum coding_system_type
  93   {
  94     coding_type_charset,
  95     coding_type_utf_8,
  96     coding_type_utf_16,
  97     coding_type_iso_2022,
  98     coding_type_emacs_mule,
  99     coding_type_sjis,
 100     coding_type_ccl,
 101     coding_type_raw_text,
 102     coding_type_undecided,
 103     coding_type_max
 104   };
 105
 106
 107 /* Enumeration of end-of-line format type.  */
 108
 109 enum end_of_line_type
 110   {
 111     eol_lf,             /* Line-feed only, same as Emacs' internal
 112                            format.  */
 113     eol_crlf,           /* Sequence of carriage-return and
 114                            line-feed.  */
 115     eol_cr,             /* Carriage-return only.  */
 116     eol_any,            /* Accept any of above.  Produce line-feed
 117                            only.  */
 118     eol_undecided,      /* This value is used to denote that the
 119                            eol-type is not yet undecided.  */
 120     eol_type_max
 121   };
 122
 123 /* Enumeration of index to an attribute vector of a coding system.  */
 124
 125 enum coding_attr_index
 126   {
 127     coding_attr_base_name,
 128     coding_attr_docstring,
 129     coding_attr_mnemonic,
 130     coding_attr_type,
 131     coding_attr_charset_list,
 132     coding_attr_ascii_compat,
 133     coding_attr_decode_tbl,
 134     coding_attr_encode_tbl,
 135     coding_attr_trans_tbl,
 136     coding_attr_post_read,
 137     coding_attr_pre_write,
 138     coding_attr_default_char,
 139     coding_attr_for_unibyte,
 140     coding_attr_plist,
 141
 142     coding_attr_category,
 143     coding_attr_safe_charsets,
 144
 145     /* The followings are extra attributes for each type.  */
 146     coding_attr_charset_valids,
 147
 148     coding_attr_ccl_decoder,
 149     coding_attr_ccl_encoder,
 150     coding_attr_ccl_valids,
 151
 152     coding_attr_iso_initial,
 153     coding_attr_iso_usage,
 154     coding_attr_iso_request,
 155     coding_attr_iso_flags,
 156
 157     coding_attr_utf_bom,
 158     coding_attr_utf_16_endian,
 159
 160     coding_attr_emacs_mule_full,
 161
 162     coding_attr_last_index
 163   };
 164
 165
 166 /* Macros to access an element of an attribute vector.  */
 167
 168 #define CODING_ATTR_BASE_NAME(attrs)    AREF (attrs, coding_attr_base_name)
 169 #define CODING_ATTR_TYPE(attrs)         AREF (attrs, coding_attr_type)
 170 #define CODING_ATTR_CHARSET_LIST(attrs) AREF (attrs, coding_attr_charset_list)
 171 #define CODING_ATTR_MNEMONIC(attrs)     AREF (attrs, coding_attr_mnemonic)
 172 #define CODING_ATTR_DOCSTRING(attrs)    AREF (attrs, coding_attr_docstring)
 173 #define CODING_ATTR_ASCII_COMPAT(attrs) AREF (attrs, coding_attr_ascii_compat)
 174 #define CODING_ATTR_DECODE_TBL(attrs)   AREF (attrs, coding_attr_decode_tbl)
 175 #define CODING_ATTR_ENCODE_TBL(attrs)   AREF (attrs, coding_attr_encode_tbl)
 176 #define CODING_ATTR_TRANS_TBL(attrs)    AREF (attrs, coding_attr_trans_tbl)
 177 #define CODING_ATTR_POST_READ(attrs)    AREF (attrs, coding_attr_post_read)
 178 #define CODING_ATTR_PRE_WRITE(attrs)    AREF (attrs, coding_attr_pre_write)
 179 #define CODING_ATTR_DEFAULT_CHAR(attrs) AREF (attrs, coding_attr_default_char)
 180 #define CODING_ATTR_FOR_UNIBYTE(attrs)  AREF (attrs, coding_attr_for_unibyte)
 181 #define CODING_ATTR_FLUSHING(attrs)     AREF (attrs, coding_attr_flushing)
 182 #define CODING_ATTR_PLIST(attrs)        AREF (attrs, coding_attr_plist)
 183 #define CODING_ATTR_CATEGORY(attrs)     AREF (attrs, coding_attr_category)
 184 #define CODING_ATTR_SAFE_CHARSETS(attrs)AREF (attrs, coding_attr_safe_charsets)
 185
 186
 187 /* Return the name of a coding system specified by ID.  */
 188 #define CODING_ID_NAME(id) \
 189   (HASH_KEY (XHASH_TABLE (Vcoding_system_hash_table), id))
 190
 191 /* Return the attribute vector of a coding system specified by ID.  */
 192
 193 #define CODING_ID_ATTRS(id)     \
 194   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 0))
 195
 196 /* Return the list of aliases of a coding system specified by ID.  */
 197
 198 #define CODING_ID_ALIASES(id)   \
 199   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 1))
 200
 201 /* Return the eol-type of a coding system specified by ID.  */
 202
 203 #define CODING_ID_EOL_TYPE(id)  \
 204   (AREF (HASH_VALUE (XHASH_TABLE (Vcoding_system_hash_table), id), 2))
 205
 206
 207 /* Return the spec vector of CODING_SYSTEM_SYMBOL.  */
 208
 209 #define CODING_SYSTEM_SPEC(coding_system_symbol)        \
 210   (Fgethash (coding_system_symbol, Vcoding_system_hash_table, Qnil))
 211
 212
 213 /* Return the ID of CODING_SYSTEM_SYMBOL.  */
 214
 215 #define CODING_SYSTEM_ID(coding_system_symbol)                  \
 216   hash_lookup (XHASH_TABLE (Vcoding_system_hash_table),         \
 217                coding_system_symbol, NULL)
 218
 219 /* Return 1 if CODING_SYSTEM_SYMBOL is a coding system.  */
 220
 221 #define CODING_SYSTEM_P(coding_system_symbol)           \
 222   (CODING_SYSTEM_ID (coding_system_symbol) >= 0         \
 223    || (! NILP (coding_system_symbol)                    \
 224        && ! NILP (Fcoding_system_p (coding_system_symbol))))
 225
 226 /* Check if X is a coding system or not.  */
 227
 228 #define CHECK_CODING_SYSTEM(x)                          \
 229   do {                                                  \
 230     if (CODING_SYSTEM_ID (x) < 0                        \
 231         && NILP (Fcheck_coding_system (x)))             \
 232       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 SEPC to the
 237    spec vector of the coding system.  */
 238
 239 #define CHECK_CODING_SYSTEM_GET_SPEC(x, spec)           \
 240   do {                                                  \
 241     spec = CODING_SYSTEM_SPEC (x);                      \
 242     if (NILP (spec))                                    \
 243       {                                                 \
 244         Fcheck_coding_system (x);                       \
 245         spec = CODING_SYSTEM_SPEC (x);                  \
 246       }                                                 \
 247     if (NILP (spec))                                    \
 248       wrong_type_argument (Qcoding_system_p, (x));      \
 249   } while (0)
 250
 251
 252 /* Check if X is a coding system or not.  If it is, set ID to the
 253    ID of the coding system.  */
 254
 255 #define CHECK_CODING_SYSTEM_GET_ID(x, id)                       \
 256   do                                                            \
 257     {                                                           \
 258       id = CODING_SYSTEM_ID (x);                                \
 259       if (id < 0)                                               \
 260         {                                                       \
 261           Fcheck_coding_system (x);                             \
 262           id = CODING_SYSTEM_ID (x);                            \
 263         }                                                       \
 264       if (id < 0)                                               \
 265         wrong_type_argument (Qcoding_system_p, (x));    \
 266     } while (0)
 267
 268
 269 /*** GENERAL section ***/
 270
 271 /* Enumeration of result code of code conversion.  */
 272 enum coding_result_code
 273   {
 274     CODING_RESULT_SUCCESS,
 275     CODING_RESULT_INSUFFICIENT_SRC,
 276     CODING_RESULT_INSUFFICIENT_DST,
 277     CODING_RESULT_INCONSISTENT_EOL,
 278     CODING_RESULT_INVALID_SRC,
 279     CODING_RESULT_INTERRUPT,
 280     CODING_RESULT_INSUFFICIENT_MEM
 281   };
 282
 283
 284 /* Macros used for the member `mode' of the struct coding_system.  */
 285
 286 /* If set, recover the original CR or LF of the already decoded text
 287    when the decoding routine encounters an inconsistent eol format.  */
 288 #define CODING_MODE_INHIBIT_INCONSISTENT_EOL    0x01
 289
 290 /* If set, the decoding/encoding routines treat the current data as
 291    the last block of the whole text to be converted, and do the
 292    appropriate finishing job.  */
 293 #define CODING_MODE_LAST_BLOCK                  0x02
 294
 295 /* If set, it means that the current source text is in a buffer which
 296    enables selective display.  */
 297 #define CODING_MODE_SELECTIVE_DISPLAY           0x04
 298
 299 /* This flag is used by the decoding/encoding routines on the fly.  If
 300    set, it means that right-to-left text is being processed.  */
 301 #define CODING_MODE_DIRECTION                   0x08
 302
 303 #define CODING_MODE_FIXED_DESTINATION           0x10
 304
 305 /* If set, it means that the encoding routines produces some safe
 306    ASCII characters (usually '?') for unsupported characters.  */
 307 #define CODING_MODE_SAFE_ENCODING               0x20
 308
 309 /* Structure of the field `spec.iso_2022' in the structure
 310    `coding_system'.  */
 311 struct iso_2022_spec
 312 {
 313   /* Bit-wise-or of CODING_ISO_FLAG_XXX.  */
 314   unsigned flags;
 315
 316   /* The current graphic register invoked to each graphic plane.  */
 317   int current_invocation[2];
 318
 319   /* The current charset designated to each graphic register.  The
 320      value -1 means that not charset is designated, -2 means that
 321      there was an invalid designation previously.  */
 322   int current_designation[4];
 323
 324   /* Set to 1 temporarily only when graphic register 2 or 3 is invoked
 325      by single-shift while encoding.  */
 326   int single_shifting;
 327
 328   /* Set to 1 temporarily only when processing at beginning of line.  */
 329   int bol;
 330 };
 331
 332 struct ccl_spec;
 333
 334 enum utf_bom_type
 335   {
 336     utf_detect_bom,
 337     utf_without_bom,
 338     utf_with_bom
 339   };
 340
 341 enum utf_16_endian_type
 342   {
 343     utf_16_big_endian,
 344     utf_16_little_endian
 345   };
 346
 347 struct utf_16_spec
 348 {
 349   enum utf_bom_type bom;
 350   enum utf_16_endian_type endian;
 351   int surrogate;
 352 };
 353
 354 struct coding_detection_info
 355 {
 356   /* Values of these members are bitwise-OR of CATEGORY_MASK_XXXs.  */
 357   /* Which categories are already checked.  */
 358   int checked;
 359   /* Which categories are strongly found.  */
 360   int found;
 361   /* Which categories are rejected.  */
 362   int rejected;
 363 };
 364
 365
 366 struct coding_system
 367 {
 368   /* ID number of the coding system.  This is an index to
 369      Vcoding_system_hash_table.  This value is set by
 370      setup_coding_system.  At the early stage of building time, this
 371      value is -1 in the array coding_categories to indicate that no
 372      coding-system of that category is yet defined.  */
 373   int id;
 374
 375   /* Flag bits of the coding system.  The meaning of each bit is common
 376      to all types of coding systems.  */
 377   int common_flags;
 378
 379   /* Mode bits of the coding system.  See the comments of the macros
 380      CODING_MODE_XXX.  */
 381   unsigned int mode;
 382
 383   /* Detailed information specific to each type of coding system.  */
 384   union
 385     {
 386       struct iso_2022_spec iso_2022;
 387       struct ccl_spec *ccl;     /* Defined in ccl.h.  */
 388       struct utf_16_spec utf_16;
 389       enum utf_bom_type utf_8_bom;
 390       int emacs_mule_full_support;
 391     } spec;
 392
 393   int max_charset_id;
 394   unsigned char *safe_charsets;
 395
 396   /* The following two members specify how binary 8-bit code 128..255
 397      are represented in source and destination text respectively.  1
 398      means they are represented by 2-byte sequence, 0 means they are
 399      represented by 1-byte as is (see the comment in character.h).  */
 400   unsigned src_multibyte : 1;
 401   unsigned dst_multibyte : 1;
 402
 403   /* How may heading bytes we can skip for decoding.  This is set to
 404      -1 in setup_coding_system, and updated by detect_coding.  So,
 405      when this is equal to the byte length of the text being
 406      converted, we can skip the actual conversion process.  */
 407   int head_ascii;
 408
 409   /* The following members are set by encoding/decoding routine.  */
 410   EMACS_INT produced, produced_char, consumed, consumed_char;
 411
 412   /* Number of error source data found in a decoding routine.  */
 413   int errors;
 414
 415   /* Store the positions of error source data. */
 416   EMACS_INT *error_positions;
 417
 418   /* Finish status of code conversion.  */
 419   enum coding_result_code result;
 420
 421   EMACS_INT src_pos, src_pos_byte, src_chars, src_bytes;
 422   Lisp_Object src_object;
 423   const unsigned char *source;
 424
 425   EMACS_INT dst_pos, dst_pos_byte, dst_bytes;
 426   Lisp_Object dst_object;
 427   unsigned char *destination;
 428
 429   /* Set to 1 if the source of conversion is not in the member
 430      `charbuf', but at `src_object'.  */
 431   int chars_at_source;
 432
 433   /* If an element is non-negative, it is a character code.
 434
 435      If it is in the range -128..-1, it is a 8-bit character code
 436      minus 256.
 437
 438      If it is less than -128, it specifies the start of an annotation
 439      chunk.  The length of the chunk is -128 minus the value of the
 440      element.  The following elements are OFFSET, ANNOTATION-TYPE, and
 441      a sequence of actual data for the annotation.  OFFSET is a
 442      character position offset from dst_pos or src_pos,
 443      ANNOTATION-TYPE specfies the meaning of the annotation and how to
 444      handle the following data..  */
 445   int *charbuf;
 446   int charbuf_size, charbuf_used;
 447
 448   /* Set to 1 if charbuf contains an annotation.  */
 449   int annotated;
 450
 451   unsigned char carryover[64];
 452   int carryover_bytes;
 453
 454   int default_char;
 455
 456   int (*detector) P_ ((struct coding_system *,
 457                        struct coding_detection_info *));
 458   void (*decoder) P_ ((struct coding_system *));
 459   int (*encoder) P_ ((struct coding_system *));
 460 };
 461
 462 /* Meanings of bits in the member `common_flags' of the structure
 463    coding_system.  The lowest 8 bits are reserved for various kind of
 464    annotations (currently two of them are used).  */
 465 #define CODING_ANNOTATION_MASK                  0x00FF
 466 #define CODING_ANNOTATE_COMPOSITION_MASK        0x0001
 467 #define CODING_ANNOTATE_DIRECTION_MASK          0x0002
 468 #define CODING_ANNOTATE_CHARSET_MASK            0x0003
 469 #define CODING_FOR_UNIBYTE_MASK                 0x0100
 470 #define CODING_REQUIRE_FLUSHING_MASK            0x0200
 471 #define CODING_REQUIRE_DECODING_MASK            0x0400
 472 #define CODING_REQUIRE_ENCODING_MASK            0x0800
 473 #define CODING_REQUIRE_DETECTION_MASK           0x1000
 474 #define CODING_RESET_AT_BOL_MASK                0x2000
 475
 476 /* Return 1 if the coding context CODING requires annotaion
 477    handling.  */
 478 #define CODING_REQUIRE_ANNOTATION(coding) \
 479   ((coding)->common_flags & CODING_ANNOTATION_MASK)
 480
 481 /* Return 1 if the coding context CODING prefers decoding into unibyte.  */
 482 #define CODING_FOR_UNIBYTE(coding) \
 483   ((coding)->common_flags & CODING_FOR_UNIBYTE_MASK)
 484
 485 /* Return 1 if the coding context CODING requires specific code to be
 486    attached at the tail of converted text.  */
 487 #define CODING_REQUIRE_FLUSHING(coding) \
 488   ((coding)->common_flags & CODING_REQUIRE_FLUSHING_MASK)
 489
 490 /* Return 1 if the coding context CODING requires code conversion on
 491    decoding.  */
 492 #define CODING_REQUIRE_DECODING(coding) \
 493   ((coding)->dst_multibyte              \
 494    || (coding)->common_flags & CODING_REQUIRE_DECODING_MASK)
 495
 496
 497 /* Return 1 if the coding context CODING requires code conversion on
 498    encoding.
 499    The non-multibyte part of the condition is to support encoding of
 500    unibyte strings/buffers generated by string-as-unibyte or
 501    (set-buffer-multibyte nil) from multibyte strings/buffers.  */
 502 #define CODING_REQUIRE_ENCODING(coding)                         \
 503   ((coding)->src_multibyte                                      \
 504    || (coding)->common_flags & CODING_REQUIRE_ENCODING_MASK     \
 505    || (coding)->mode & CODING_MODE_SELECTIVE_DISPLAY)
 506
 507
 508 /* Return 1 if the coding context CODING requires some kind of code
 509    detection.  */
 510 #define CODING_REQUIRE_DETECTION(coding) \
 511   ((coding)->common_flags & CODING_REQUIRE_DETECTION_MASK)
 512
 513 /* Return 1 if the coding context CODING requires code conversion on
 514    decoding or some kind of code detection.  */
 515 #define CODING_MAY_REQUIRE_DECODING(coding)     \
 516   (CODING_REQUIRE_DECODING (coding)             \
 517    || CODING_REQUIRE_DETECTION (coding))
 518
 519 /* Macros to decode or encode a character of JISX0208 in SJIS.  S1 and
 520    S2 are the 1st and 2nd position-codes of JISX0208 in SJIS coding
 521    system.  C1 and C2 are the 1st and 2nd position codes of Emacs'
 522    internal format.  */
 523
 524 #define SJIS_TO_JIS(code)                               \
 525   do {                                                  \
 526     int s1, s2, j1, j2;                                 \
 527                                                         \
 528     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 529                                                         \
 530     if (s2 >= 0x9F)                                     \
 531       (j1 = s1 * 2 - (s1 >= 0xE0 ? 0x160 : 0xE0),       \
 532        j2 = s2 - 0x7E);                                 \
 533     else                                                \
 534       (j1 = s1 * 2 - ((s1 >= 0xE0) ? 0x161 : 0xE1),     \
 535        j2 = s2 - ((s2 >= 0x7F) ? 0x20 : 0x1F));         \
 536     (code) = (j1 << 8) | j2;                            \
 537   } while (0)
 538
 539 #define SJIS_TO_JIS2(code)                              \
 540   do {                                                  \
 541     int s1, s2, j1, j2;                                 \
 542                                                         \
 543     s1 = (code) >> 8, s2 = (code) & 0xFF;               \
 544                                                         \
 545     if (s2 >= 0x9F)                                     \
 546       {                                                 \
 547         j1 = (s1 == 0xF0 ? 0x28                         \
 548               : s1 == 0xF1 ? 0x24                       \
 549               : s1 == 0xF2 ? 0x2C                       \
 550               : s1 == 0xF3 ? 0x2E                       \
 551               : 0x6E + (s1 - 0xF4) * 2);                \
 552         j2 = s2 - 0x7E;                                 \
 553       }                                                 \
 554     else                                                \
 555       {                                                 \
 556         j1 = (s1 <= 0xF2 ? 0x21 + (s1 - 0xF0) * 2       \
 557               : s1 <= 0xF4 ? 0x2D + (s1 - 0xF3) * 2     \
 558               : 0x6F + (s1 - 0xF5) * 2);                \
 559         j2 = s2 - ((s2 >= 0x7F ? 0x20 : 0x1F));         \
 560       }                                                 \
 561     (code) = (j1 << 8) | j2;                            \
 562   } while (0)
 563
 564
 565 #define JIS_TO_SJIS(code)                               \
 566   do {                                                  \
 567     int s1, s2, j1, j2;                                 \
 568                                                         \
 569     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 570     if (j1 & 1)                                         \
 571       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x71 : 0xB1),       \
 572        s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F));         \
 573     else                                                \
 574       (s1 = j1 / 2 + ((j1 < 0x5F) ? 0x70 : 0xB0),       \
 575        s2 = j2 + 0x7E);                                 \
 576     (code) = (s1 << 8) | s2;                            \
 577   } while (0)
 578
 579 #define JIS_TO_SJIS2(code)                              \
 580   do {                                                  \
 581     int s1, s2, j1, j2;                                 \
 582                                                         \
 583     j1 = (code) >> 8, j2 = (code) & 0xFF;               \
 584     if (j1 & 1)                                         \
 585       {                                                 \
 586         s1 = (j1 <= 0x25 ? 0xF0 + (j1 - 0x21) / 2       \
 587               : j1 <= 0x27 ? 0xF3 + (j1 - 0x2D) / 2     \
 588               : 0xF5 + (j1 - 0x6F) / 2);                \
 589         s2 = j2 + ((j2 >= 0x60) ? 0x20 : 0x1F);         \
 590       }                                                 \
 591     else                                                \
 592       {                                                 \
 593         s1 = (j1 == 0x28 ? 0xF0                         \
 594               : j1 == 0x24 ? 0xF1                       \
 595               : j1 == 0x2C ? 0xF2                       \
 596               : j1 == 0x2E ? 0xF3                       \
 597               : 0xF4 + (j1 - 0x6E) / 2);                \
 598         s2 = j2 + 0x7E;                                 \
 599       }                                                 \
 600     (code) = (s1 << 8) | s2;                            \
 601   } while (0)
 602
 603 /* Encode the file name NAME using the specified coding system
 604    for file names, if any.  */
 605 #define ENCODE_FILE(name)                                                  \
 606   (! NILP (Vfile_name_coding_system)                                       \
 607    && !EQ (Vfile_name_coding_system, make_number (0))                      \
 608    ? code_convert_string_norecord (name, Vfile_name_coding_system, 1)      \
 609    : (! NILP (Vdefault_file_name_coding_system)                            \
 610       && !EQ (Vdefault_file_name_coding_system, make_number (0))           \
 611       ? code_convert_string_norecord (name, Vdefault_file_name_coding_system, 1) \
 612       : name))
 613
 614
 615 /* Decode the file name NAME using the specified coding system
 616    for file names, if any.  */
 617 #define DECODE_FILE(name)                                                  \
 618   (! NILP (Vfile_name_coding_system)                                       \
 619    && !EQ (Vfile_name_coding_system, make_number (0))                      \
 620    ? code_convert_string_norecord (name, Vfile_name_coding_system, 0)      \
 621    : (! NILP (Vdefault_file_name_coding_system)                            \
 622       && !EQ (Vdefault_file_name_coding_system, make_number (0))           \
 623       ? code_convert_string_norecord (name, Vdefault_file_name_coding_system, 0) \
 624       : name))
 625
 626
 627 /* Encode the string STR using the specified coding system
 628    for system functions, if any.  */
 629 #define ENCODE_SYSTEM(str)                                                 \
 630   (! NILP (Vlocale_coding_system)                                          \
 631    && !EQ (Vlocale_coding_system, make_number (0))                         \
 632    ? code_convert_string_norecord (str, Vlocale_coding_system, 1)          \
 633    : str)
 634
 635 /* Decode the string STR using the specified coding system
 636    for system functions, if any.  */
 637 #define DECODE_SYSTEM(str)                                                 \
 638   (! NILP (Vlocale_coding_system)                                          \
 639    && !EQ (Vlocale_coding_system, make_number (0))                         \
 640    ? code_convert_string_norecord (str, Vlocale_coding_system, 0)          \
 641    : str)
 642
 643 /* Used by the gtk menu code.  Note that this encodes utf-8, not
 644    utf-8-emacs, so it's not a no-op.  */
 645 #define ENCODE_UTF_8(str) code_convert_string_norecord (str, Qutf_8, 1)
 646
 647 /* Extern declarations.  */
 648 extern Lisp_Object code_conversion_save P_ ((int, int));
 649 extern int decoding_buffer_size P_ ((struct coding_system *, int));
 650 extern int encoding_buffer_size P_ ((struct coding_system *, int));
 651 extern void setup_coding_system P_ ((Lisp_Object, struct coding_system *));
 652 extern Lisp_Object coding_charset_list P_ ((struct coding_system *));
 653 extern Lisp_Object coding_system_charset_list P_ ((Lisp_Object));
 654 extern void detect_coding P_ ((struct coding_system *));
 655 extern Lisp_Object code_convert_region P_ ((Lisp_Object, Lisp_Object,
 656                                             Lisp_Object, Lisp_Object,
 657                                             int, int));
 658 extern Lisp_Object code_convert_string P_ ((Lisp_Object, Lisp_Object,
 659                                             Lisp_Object, int, int, int));
 660 extern Lisp_Object code_convert_string_norecord P_ ((Lisp_Object, Lisp_Object,
 661                                                      int));
 662 extern Lisp_Object raw_text_coding_system P_ ((Lisp_Object));
 663 extern Lisp_Object coding_inherit_eol_type P_ ((Lisp_Object, Lisp_Object));
 664
 665 extern int decode_coding_gap P_ ((struct coding_system *,
 666                                   EMACS_INT, EMACS_INT));
 667 extern int encode_coding_gap P_ ((struct coding_system *,
 668                                   EMACS_INT, EMACS_INT));
 669 extern void decode_coding_object P_ ((struct coding_system *,
 670                                       Lisp_Object, EMACS_INT, EMACS_INT,
 671                                       EMACS_INT, EMACS_INT, Lisp_Object));
 672 extern void encode_coding_object P_ ((struct coding_system *,
 673                                       Lisp_Object, EMACS_INT, EMACS_INT,
 674                                       EMACS_INT, EMACS_INT, Lisp_Object));
 675
 676 /* Macros for backward compatibility.  */
 677
 678 #define decode_coding_region(coding, from, to)          \
 679   decode_coding_object (coding, Fcurrent_buffer (),     \
 680                         from, CHAR_TO_BYTE (from),      \
 681                         to, CHAR_TO_BYTE (to), Fcurrent_buffer ())
 682
 683
 684 #define encode_coding_region(coding, from, to)          \
 685   encode_coding_object (coding, Fcurrent_buffer (),     \
 686                         from, CHAR_TO_BYTE (from),      \
 687                         to, CHAR_TO_BYTE (to), Fcurrent_buffer ())
 688
 689
 690 #define decode_coding_string(coding, string, nocopy)                    \
 691   decode_coding_object (coding, string, 0, 0, XSTRING (string)->size,   \
 692                         STRING_BYTES (XSTRING (string)), Qt)
 693
 694 #define encode_coding_string(coding, string, nocopy)                    \
 695   (encode_coding_object (coding, string, 0, 0, XSTRING (string)->size,  \
 696                          STRING_BYTES (XSTRING (string)), Qt),          \
 697    (coding)->dst_object)
 698
 699
 700 #define decode_coding_c_string(coding, src, bytes, dst_object)          \
 701   do {                                                                  \
 702     (coding)->source = (src);                                           \
 703     (coding)->src_chars = (coding)->src_bytes = (bytes);                \
 704     decode_coding_object ((coding), Qnil, 0, 0, (bytes), (bytes),       \
 705                           (dst_object));                                \
 706   } while (0)
 707
 708
 709 extern Lisp_Object preferred_coding_system P_ (());
 710
 711
 712 extern Lisp_Object Qutf_8, Qutf_8_emacs;
 713
 714 extern Lisp_Object Qcoding_system, Qeol_type, Qcoding_category_index;
 715 extern Lisp_Object Qcoding_system_p;
 716 extern Lisp_Object Qraw_text, Qemacs_mule, Qno_conversion, Qundecided;
 717 extern Lisp_Object Qiso_2022;
 718 extern Lisp_Object Qbuffer_file_coding_system;
 719
 720 extern Lisp_Object Qunix, Qdos, Qmac;
 721
 722 extern Lisp_Object Qtranslation_table;
 723 extern Lisp_Object Qtranslation_table_id;
 724
 725 /* Mnemonic strings to indicate each type of end-of-line.  */
 726 extern Lisp_Object eol_mnemonic_unix, eol_mnemonic_dos, eol_mnemonic_mac;
 727 /* Mnemonic string to indicate type of end-of-line is not yet decided.  */
 728 extern Lisp_Object eol_mnemonic_undecided;
 729
 730 #ifdef emacs
 731 extern Lisp_Object Qfile_coding_system;
 732 extern Lisp_Object Qcall_process, Qcall_process_region;
 733 extern Lisp_Object Qstart_process, Qopen_network_stream;
 734 extern Lisp_Object Qwrite_region;
 735
 736 extern char *emacs_strerror P_ ((int));
 737
 738 /* Coding-system for reading files and receiving data from process.  */
 739 extern Lisp_Object Vcoding_system_for_read;
 740 /* Coding-system for writing files and sending data to process.  */
 741 extern Lisp_Object Vcoding_system_for_write;
 742 /* Coding-system actually used in the latest I/O.  */
 743 extern Lisp_Object Vlast_coding_system_used;
 744 /* Coding-system to use with system messages (e.g. strerror).  */
 745 extern Lisp_Object Vlocale_coding_system;
 746
 747 /* If non-zero, process buffer inherits the coding system used to decode
 748    the subprocess output.  */
 749 extern int inherit_process_coding_system;
 750
 751 /* Coding system to be used to encode text for terminal display when
 752    terminal coding system is nil.  */
 753 extern struct coding_system safe_terminal_coding;
 754
 755 /* Default coding systems used for process I/O.  */
 756 extern Lisp_Object Vdefault_process_coding_system;
 757
 758 /* Function to call to force a user to force select a propert coding
 759    system.  */
 760 extern Lisp_Object Vselect_safe_coding_system_function;
 761
 762 /* If nonzero, on writing a file, Vselect_safe_coding_system_function
 763    is called even if Vcoding_system_for_write is non-nil.  */
 764 extern int coding_system_require_warning;
 765
 766 /* Coding system for file names, or nil if none.  */
 767 extern Lisp_Object Vfile_name_coding_system;
 768
 769 /* Coding system for file names used only when
 770    Vfile_name_coding_system is nil.  */
 771 extern Lisp_Object Vdefault_file_name_coding_system;
 772
 773 #endif
 774
 775 /* Error signaled when there's a problem with detecting coding system */
 776 extern Lisp_Object Qcoding_system_error;
 777
 778 extern char emacs_mule_bytes[256];
 779 extern int emacs_mule_string_char P_ ((unsigned char *));
 780
 781 #endif /* EMACS_CODING_H */
 782
 783 /* arch-tag: 2bc3b4fa-6870-4f64-8135-b962b2d290e4
 784    (do not change this comment) */