Mercurial > hg > xemacs-beta
annotate src/file-coding.h @ 4614:afbfad080ddd
The URLs in our current config.guess and config.sub files are obsolete.
Update to the latest upstream release to get correct URLs, as well as fixes
and enhancements to those scripts.
| author | Jerry James <james@xemacs.org> |
|---|---|
| date | Wed, 11 Feb 2009 11:09:35 -0700 |
| parents | 80e0588fb42f |
| children | 257b468bf2ca |
| rev | line source |
|---|---|
| 771 | 1 /* Header for encoding conversion functions; coding-system object. |
| 2 #### rename me to coding-system.h | |
| 428 | 3 Copyright (C) 1991, 1995 Free Software Foundation, Inc. |
| 4 Copyright (C) 1995 Sun Microsystems, Inc. | |
| 793 | 5 Copyright (C) 2000, 2001, 2002 Ben Wing. |
| 428 | 6 |
| 7 This file is part of XEmacs. | |
| 8 | |
| 9 XEmacs is free software; you can redistribute it and/or modify it | |
| 10 under the terms of the GNU General Public License as published by the | |
| 11 Free Software Foundation; either version 2, or (at your option) any | |
| 12 later version. | |
| 13 | |
| 14 XEmacs is distributed in the hope that it will be useful, but WITHOUT | |
| 15 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or | |
| 16 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License | |
| 17 for more details. | |
| 18 | |
| 19 You should have received a copy of the GNU General Public License | |
| 20 along with XEmacs; see the file COPYING. If not, write to | |
| 21 the Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
| 22 Boston, MA 02111-1307, USA. */ | |
| 23 | |
| 24 /* Synched up with: Mule 2.3. Not in FSF. */ | |
| 25 | |
| 771 | 26 /* Authorship: |
| 27 | |
| 28 Current primary author: Ben Wing <ben@xemacs.org> | |
| 29 | |
| 30 Written by Ben Wing <ben@xemacs.org> for XEmacs, 1995, loosely based | |
| 31 on code written 91.10.09 by K.Handa <handa@etl.go.jp>. | |
| 32 Rewritten again 2000-2001 by Ben Wing to support properly | |
| 33 abstracted coding systems. | |
| 34 September 2001: Finished last part of abstraction, the detection | |
| 35 mechanism. | |
| 36 */ | |
| 428 | 37 |
| 440 | 38 #ifndef INCLUDED_file_coding_h_ |
| 39 #define INCLUDED_file_coding_h_ | |
| 428 | 40 |
| 771 | 41 /* Capsule description of the different structures, what their purpose is, |
| 42 how they fit together, and where various bits of data are stored. | |
| 43 | |
| 2297 | 44 A "coding system" is an algorithm for converting stream data in one format |
| 45 into stream data in another format. Currently most of the coding systems | |
| 46 we have created concern internationalized text, and convert between the | |
| 47 XEmacs internal format for multilingual text, and various external | |
| 771 | 48 representations of such text. However, any such conversion is possible, |
| 49 for example, compressing or uncompressing text using the gzip algorithm. | |
| 50 All coding systems provide both encode and decode routines, so that the | |
| 2297 | 51 conversion can go both ways. Unfortunately encoding and decoding may not |
| 52 be exact inverses, even for a specific instance of a coding system. Care | |
| 53 must be taken when this is not the case. | |
| 771 | 54 |
| 55 The way we handle this is by dividing the various potential coding | |
| 56 systems into types, analogous to classes in C++. Each coding system | |
| 57 type encompasses a series of related coding systems that it can | |
| 58 implement, and it has properties which control how exactly the encoding | |
| 59 works. A particular set of values for each of the properties makes up a | |
| 60 "coding system", and specifies one particular encoding. A `struct | |
| 61 Lisp_Coding_System' object encapsulates those settings -- its type, the | |
| 62 values chosen for all properties of that type, a name for the coding | |
| 63 system, some documentation. | |
| 64 | |
| 65 In addition, there are of course methods associated with a coding system | |
| 66 type, implementing the encoding, decoding, etc. These are stored in a | |
| 67 `struct coding_system_methods' object, one per coding-system type, which | |
| 68 contains mostly function pointers. This is retrievable from the | |
| 69 coding-system object (i.e. the struct Lisp_Coding_System), which has a | |
| 70 pointer to it. | |
| 71 | |
| 72 In order to actually use a coding system to do an encoding or decoding | |
| 73 operation, you need to use a coding Lstream. | |
| 74 | |
| 75 Now let's look more at attached data. All coding systems have certain | |
| 76 common data fields -- name, type, documentation, etc. -- as well as a | |
| 77 bunch more that are defined by the coding system type. To handle this | |
| 78 cleanly, each coding system type defines a structure that holds just the | |
| 79 fields of data particular to it, and calls it e.g. `struct | |
| 80 iso2022_coding_system' for coding system type `iso2022'. When the | |
| 81 memory block holding the coding system object is created, it is sized | |
| 82 such that it can hold both the struct Lisp_Coding_System and the struct | |
| 83 iso2022_coding_system (or whatever) directly following it. (This is a | |
| 84 common trick; another possibility is to have a void * pointer in the | |
| 85 struct Lisp_Coding_System, which points to another memory block holding | |
| 86 the struct iso2022_coding_system.) A macro is provided | |
| 87 (CODING_SYSTEM_TYPE_DATA) to retrieve a pointer of the right type to the | |
| 88 type-specific data contained within the overall `struct | |
| 89 Lisp_Coding_System' block. | |
| 90 | |
| 91 Lstreams, similarly, are objects of type `struct lstream' holding data | |
| 92 about the stream operation (how much data has been read or written, any | |
| 93 buffered data, any error conditions, etc.), and like coding systems have | |
| 94 different types. They have a structure called `Lstream_implementation', | |
| 95 one per lstream type, exactly analogous to `struct | |
| 96 coding_system_methods'. In addition, they have type-specific data | |
| 97 (specifying, e.g., the file number, FILE *, memory location, other | |
| 98 lstream, etc. to read the data from or write it to, and for conversion | |
| 99 processes, the current state of the process -- are we decoding ASCII or | |
| 100 Kanji characters? are we in the middle of a processing an escape | |
| 101 sequence? etc.). This type-specific data is stored in a structure | |
| 102 named `struct coding_stream'. Just like for coding systems, the | |
| 103 type-independent data in the `struct lstream' and the type-dependent | |
| 104 data in the `struct coding_stream' are stored together in the same | |
| 105 memory block. | |
| 428 | 106 |
| 771 | 107 Now things get a bit tricky. The `struct coding_stream' is |
| 108 type-specific from the point of view of an lstream, but not from the | |
| 109 point of view of a coding system. It contains only general data about | |
| 110 the conversion process, e.g. the name of the coding system used for | |
| 111 conversion, the lstream that we take data from or write it to (depending | |
| 112 on whether this was created as a read stream or a write stream), a | |
| 113 buffer to hold extra data we retrieved but can't send on yet, some | |
| 114 flags, etc. It also needs some data specific to the particular coding | |
| 115 system and thus to the particular operation going on. This data is held | |
| 116 in a structure named (e.g.) `struct iso2022_coding_stream', and it's | |
| 117 held in a separate memory block and pointed to by the generic `struct | |
| 118 coding_stream'. It's not glommed into a single memory block both | |
| 119 because that would require making changes to the generic lstream code | |
| 120 and more importantly because the coding system used in a particular | |
| 121 coding lstream can be changed at any point during the lifetime of the | |
| 122 lstream, and possibly multiple times. (For example, it can be set using | |
| 123 the Lisp primitives `set-process-input-coding-system' and | |
| 124 `set-console-tty-input-coding-system', as well as getting set when a | |
| 125 conversion operation was started with coding system `undecided' and the | |
| 2297 | 126 correct coding system was then detected.) #### This suggests implementing |
| 127 compound text extended segments by saving the state of the ctext stream, | |
| 128 and installing an appropriate for the duration of the segment. | |
| 428 | 129 |
| 771 | 130 IMPORTANT NOTE: There are at least two ancillary data structures |
| 131 associated with a coding system type. (There may also be detection data; | |
| 132 see elsewhere.) It's important, when writing a coding system type, to | |
| 133 keep straight which type of data goes where. In particular, `struct | |
| 134 foo_coding_system' is attached to the coding system object itself. This | |
| 135 is a permanent object and there's only one per coding system. It's | |
| 136 created once, usually at init time, and never destroyed. So, `struct | |
| 137 foo_coding_system' should in general not contain dynamic data! (Just | |
| 138 data describing the properties of the coding system.) In particular, | |
| 139 *NO* data about any conversion in progress. There may be many | |
| 140 conversions going on simultaneously using a particular coding system, | |
| 141 and by storing conversion data in the coding system, these conversions | |
| 142 will overwrite each other's data. | |
| 143 | |
| 144 Instead, use the lstream object, whose purpose is to encapsulate a | |
| 145 particular conversion and all associated data. From the lstream object, | |
| 146 you can get the struct coding_stream using something like | |
| 147 | |
| 148 struct coding_stream *str = LSTREAM_TYPE_DATA (lstr, coding); | |
| 149 | |
| 150 But usually this structure is already passed to you as one of the | |
| 151 parameters of the method being invoked. | |
| 152 | |
| 153 From the struct coding_stream, you can retrieve the | |
| 154 coding-system-type-specific data using something like | |
| 155 | |
| 156 struct foo_coding_stream *data = CODING_STREAM_TYPE_DATA (str, foo); | |
| 157 | |
| 158 Then, use this structure to hold all data relevant to the particular | |
| 159 conversion being done. | |
| 160 | |
| 161 Initialize this structure whenever init_coding_stream_method is called | |
| 162 (this may happen more than once), and finalize it (free resources, etc.) | |
| 163 when finalize_coding_stream_method is called. | |
| 164 */ | |
| 165 | |
| 166 struct coding_stream; | |
| 167 struct detection_state; | |
| 168 | |
| 1204 | 169 extern const struct sized_memory_description coding_system_methods_description; |
| 771 | 170 |
| 171 struct coding_system_methods; | |
| 172 | |
| 173 enum source_sink_type | |
| 428 | 174 { |
| 771 | 175 DECODES_CHARACTER_TO_BYTE, |
| 176 DECODES_BYTE_TO_BYTE, | |
| 177 DECODES_BYTE_TO_CHARACTER, | |
| 178 DECODES_CHARACTER_TO_CHARACTER | |
| 428 | 179 }; |
| 180 | |
| 181 enum eol_type | |
| 182 { | |
| 183 EOL_LF, | |
| 184 EOL_CRLF, | |
| 771 | 185 EOL_CR, |
| 1429 | 186 EOL_AUTODETECT |
| 428 | 187 }; |
| 188 | |
| 189 struct Lisp_Coding_System | |
| 190 { | |
| 3017 | 191 struct LCRECORD_HEADER header; |
| 771 | 192 struct coding_system_methods *methods; |
| 428 | 193 |
| 1204 | 194 #define CODING_SYSTEM_SLOT_DECLARATION |
| 195 #define MARKED_SLOT(x) Lisp_Object x; | |
| 196 #include "coding-system-slots.h" | |
| 771 | 197 |
| 1204 | 198 /* Eol type requested by user. See comment about EOL junk in |
| 199 coding-system-slots.h. */ | |
| 771 | 200 enum eol_type eol_type; |
| 428 | 201 |
| 2132 | 202 /* If true, this is an internal coding system, which will not show up in |
| 203 coding-system-list unless a special parameter is given to it. */ | |
| 204 int internal_p; | |
| 205 | |
| 771 | 206 /* type-specific extra data attached to a coding_system */ |
| 207 char data[1]; | |
| 428 | 208 }; |
| 209 typedef struct Lisp_Coding_System Lisp_Coding_System; | |
| 210 | |
| 440 | 211 DECLARE_LRECORD (coding_system, Lisp_Coding_System); |
| 212 #define XCODING_SYSTEM(x) XRECORD (x, coding_system, Lisp_Coding_System) | |
| 617 | 213 #define wrap_coding_system(p) wrap_record (p, coding_system) |
| 428 | 214 #define CODING_SYSTEMP(x) RECORDP (x, coding_system) |
| 215 #define CHECK_CODING_SYSTEM(x) CHECK_RECORD (x, coding_system) | |
| 216 #define CONCHECK_CODING_SYSTEM(x) CONCHECK_RECORD (x, coding_system) | |
| 217 | |
| 1204 | 218 enum coding_system_variant |
| 219 { | |
| 220 no_conversion_coding_system, | |
| 221 convert_eol_coding_system, | |
| 222 undecided_coding_system, | |
| 223 chain_coding_system, | |
| 224 text_file_wrapper_coding_system, | |
| 225 internal_coding_system, | |
| 226 gzip_coding_system, | |
| 227 mswindows_multibyte_to_unicode_coding_system, | |
| 228 mswindows_multibyte_coding_system, | |
| 229 iso2022_coding_system, | |
| 230 ccl_coding_system, | |
| 231 shift_jis_coding_system, | |
| 232 big5_coding_system, | |
| 1429 | 233 unicode_coding_system |
| 1204 | 234 }; |
| 235 | |
| 771 | 236 struct coding_system_methods |
| 237 { | |
| 238 Lisp_Object type; | |
| 239 Lisp_Object predicate_symbol; | |
| 240 | |
| 1204 | 241 /* Type expressed as an enum, needed for KKCC marking of the |
| 242 type-specific lstream data; copied into the struct coding_stream. */ | |
| 243 | |
| 244 enum coding_system_variant enumtype; | |
| 245 | |
| 771 | 246 /* Implementation specific methods: */ |
| 247 | |
| 248 /* Init method: Initialize coding-system data. Optional. */ | |
| 249 void (*init_method) (Lisp_Object coding_system); | |
| 250 | |
| 251 /* Mark method: Mark any Lisp objects in the type-specific data | |
| 252 attached to the coding-system object. Optional. */ | |
| 253 void (*mark_method) (Lisp_Object coding_system); | |
| 254 | |
| 255 /* Print method: Print the type-specific properties of this coding | |
| 256 system, as part of `print'-ing the object. If this method is defined | |
| 257 and prints anything, it should print a space as the first thing it | |
| 258 does. Optional. */ | |
| 259 void (*print_method) (Lisp_Object cs, Lisp_Object printcharfun, | |
| 260 int escapeflag); | |
| 261 | |
| 262 /* Canonicalize method: Convert this coding system to another one; called | |
| 263 once, at creation time, after all properties have been parsed. The | |
| 264 returned value should be a coding system created with | |
| 265 make_internal_coding_system() (passing the existing coding system as the | |
| 266 first argument), and will become the coding system returned by | |
| 267 `make-coding-system'. Optional. | |
| 268 | |
| 269 NOTE: There are *three* different uses of "canonical" or "canonicalize" | |
| 270 w.r.t. coding systems, and it's important to keep them straight. | |
| 271 | |
| 272 1. The canonicalize method. Used to specify a different coding | |
| 273 system, used when doing conversions, in place of the actual coding | |
| 274 system itself. Stored in the CANONICAL field of a coding system. | |
| 275 | |
| 276 2. The canonicalize-after-coding method. Used to return the encoding | |
| 277 that was "actually" used to decode some text, such that this | |
| 278 particular encoding can be used to encode the text again with the | |
| 279 expectation that the result will be the same as the original encoding. | |
| 280 Particularly important with auto-detecting coding systems. | |
| 281 | |
| 282 3. From the perspective of aliases, a "canonical" coding system is one | |
| 283 that's not an alias to some other coding system, and "canonicalization" | |
| 284 is the process of traversing the alias pointers to find the canonical | |
| 285 coding system that's equivalent to the alias. | |
| 286 */ | |
| 287 Lisp_Object (*canonicalize_method) (Lisp_Object coding_system); | |
| 288 | |
| 289 /* Canonicalize after coding method: Convert this coding system to | |
| 290 another one, after coding (usually decoding) has finished. This is | |
| 291 meant to be used by auto-detecting coding systems, which should return | |
| 292 the actually detected coding system. Optional. */ | |
| 293 Lisp_Object (*canonicalize_after_coding_method) | |
| 294 (struct coding_stream *str); | |
| 295 | |
| 296 /* Convert method: Decode or encode the data in SRC of size N, writing | |
| 297 the results into the Dynarr DST. If the conversion_end_type method | |
| 298 indicates that the source is characters (as opposed to bytes), you are | |
| 299 guaranteed to get only whole characters in the data in SRC/N. STR, a | |
| 300 struct coding_stream, stores all necessary state and other info about | |
| 301 the conversion. Coding-specific state (struct TYPE_coding_stream) can | |
| 302 be retrieved from STR using CODING_STREAM_TYPE_DATA(). Return value | |
| 303 indicates the number of bytes of the *INPUT* that were converted (not | |
| 304 the number of bytes written to the Dynarr!). This can be less than | |
| 305 the total amount of input passed in; if so, the remainder is | |
| 306 considered "rejected" and will appear again at the beginning of the | |
| 307 data passed in the next time the convert method is called. When EOF | |
| 308 is returned on the other end and there's no more data, the convert | |
| 309 method will be called one last time, STR->eof set and the passed-in | |
| 310 data will consist only of any rejected data from the previous | |
| 311 call. (At this point, file handles and similar resources can be | |
| 312 closed, but do NOT arbitrarily free data structures in the | |
| 313 type-specific data, because there are operations that can be done on | |
| 314 closed streams to query the results of the processing -- specifically, | |
| 315 for coding streams, there's the canonicalize_after_coding() method.) | |
| 316 Required. */ | |
| 317 Bytecount (*convert_method) (struct coding_stream *str, | |
| 318 const unsigned char *src, | |
| 319 unsigned_char_dynarr *dst, Bytecount n); | |
| 320 | |
| 321 /* Coding mark method: Mark any Lisp objects in the type-specific data | |
| 322 attached to `struct coding_stream'. Optional. */ | |
| 323 void (*mark_coding_stream_method) (struct coding_stream *str); | |
| 324 | |
| 325 /* Init coding stream method: Initialize the type-specific data attached | |
| 326 to the coding stream (i.e. in struct TYPE_coding_stream), when the | |
| 327 coding stream is opened. The type-specific data will be zeroed out. | |
| 328 Optional. */ | |
| 329 void (*init_coding_stream_method) (struct coding_stream *str); | |
| 330 | |
| 331 /* Rewind coding stream method: Reset any necessary type-specific data as | |
| 332 a result of the stream being rewound. Optional. */ | |
| 333 void (*rewind_coding_stream_method) (struct coding_stream *str); | |
| 334 | |
| 335 /* Finalize coding stream method: Clean up the type-specific data | |
| 336 attached to the coding stream (i.e. in struct TYPE_coding_stream). | |
| 337 Happens when the Lstream is deleted using Lstream_delete() or is | |
| 338 garbage-collected. Most streams are deleted after they've been used, | |
| 339 so it's less likely (but still possible) that allocated data will | |
| 340 stick around until GC time. (File handles can also be closed when EOF | |
| 341 is signalled; but some data must stick around after this point, for | |
| 342 the benefit of canonicalize_after_coding. See the convert method.) | |
| 343 Called only once (NOT called at disksave time). Optional. */ | |
| 344 void (*finalize_coding_stream_method) (struct coding_stream *str); | |
| 345 | |
| 346 /* Finalize method: Clean up type-specific data (e.g. free allocated | |
| 347 data) attached to the coding system (i.e. in struct | |
| 348 TYPE_coding_system), when the coding system is about to be garbage | |
| 349 collected. (Currently not called.) Called only once (NOT called at | |
| 350 disksave time). Optional. */ | |
| 351 void (*finalize_method) (Lisp_Object codesys); | |
| 352 | |
| 353 /* Conversion end type method: Does this coding system encode bytes -> | |
| 354 characters, characters -> characters, bytes -> bytes, or | |
| 355 characters -> bytes?. Default is characters -> bytes. Optional. */ | |
| 356 enum source_sink_type (*conversion_end_type_method) (Lisp_Object codesys); | |
| 357 | |
| 358 /* Putprop method: Set the value of a type-specific property. If | |
| 359 the property name is unrecognized, return 0. If the value is disallowed | |
| 360 or erroneous, signal an error. Currently called only at creation time. | |
| 361 Optional. */ | |
| 362 int (*putprop_method) (Lisp_Object codesys, | |
| 363 Lisp_Object key, | |
| 364 Lisp_Object value); | |
| 365 | |
| 366 /* Getprop method: Return the value of a type-specific property. If | |
| 367 the property name is unrecognized, return Qunbound. Optional. | |
| 368 */ | |
| 369 Lisp_Object (*getprop_method) (Lisp_Object coding_system, | |
| 370 Lisp_Object prop); | |
| 371 | |
| 372 /* These next three are set as part of the call to | |
| 373 INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA. */ | |
| 374 | |
| 375 /* Description of the extra data (struct foo_coding_system) attached to a | |
| 1204 | 376 coding system, for pdump purposes. */ |
| 377 const struct sized_memory_description *extra_description; | |
| 771 | 378 /* size of struct foo_coding_system -- extra data associated with |
| 379 the coding system */ | |
| 380 int extra_data_size; | |
| 381 /* size of struct foo_coding_stream -- extra data associated with the | |
| 382 struct coding_stream, needed for each active coding process | |
| 383 using this coding system. note that we can have more than one | |
| 384 process active at once (simply by creating more than one coding | |
| 385 lstream using this coding system), so we can't store this data in | |
| 386 the coding system object. */ | |
| 387 int coding_data_size; | |
| 388 }; | |
| 389 | |
| 390 /***** Calling a coding-system method *****/ | |
| 391 | |
| 392 #define RAW_CODESYSMETH(cs, m) ((cs)->methods->m##_method) | |
| 393 #define HAS_CODESYSMETH_P(cs, m) (!!RAW_CODESYSMETH (cs, m)) | |
| 394 #define CODESYSMETH(cs, m, args) (((cs)->methods->m##_method) args) | |
| 395 | |
| 396 /* Call a void-returning coding-system method, if it exists. */ | |
| 397 #define MAYBE_CODESYSMETH(cs, m, args) do { \ | |
| 398 Lisp_Coding_System *maybe_codesysmeth_cs = (cs); \ | |
| 399 if (HAS_CODESYSMETH_P (maybe_codesysmeth_cs, m)) \ | |
| 400 CODESYSMETH (maybe_codesysmeth_cs, m, args); \ | |
| 401 } while (0) | |
| 402 | |
| 403 /* Call a coding-system method, if it exists, or return GIVEN. | |
| 404 NOTE: Multiply-evaluates CS. */ | |
| 405 #define CODESYSMETH_OR_GIVEN(cs, m, args, given) \ | |
| 406 (HAS_CODESYSMETH_P (cs, m) ? \ | |
| 407 CODESYSMETH (cs, m, args) : (given)) | |
| 408 | |
| 409 #define XCODESYSMETH(cs, m, args) \ | |
| 410 CODESYSMETH (XCODING_SYSTEM (cs), m, args) | |
| 411 #define MAYBE_XCODESYSMETH(cs, m, args) \ | |
| 412 MAYBE_CODESYSMETH (XCODING_SYSTEM (cs), m, args) | |
| 413 #define XCODESYSMETH_OR_GIVEN(cs, m, args, given) \ | |
| 414 CODESYSMETH_OR_GIVEN (XCODING_SYSTEM (cs), m, args, given) | |
| 415 | |
| 416 | |
| 417 /***** Defining new coding-system types *****/ | |
| 418 | |
| 1204 | 419 extern const struct sized_memory_description coding_system_empty_extra_description; |
| 771 | 420 |
| 800 | 421 #ifdef ERROR_CHECK_TYPES |
| 771 | 422 #define DECLARE_CODING_SYSTEM_TYPE(type) \ |
| 423 \ | |
| 424 extern struct coding_system_methods * type##_coding_system_methods; \ | |
| 826 | 425 DECLARE_INLINE_HEADER ( \ |
| 426 struct type##_coding_system * \ | |
| 771 | 427 error_check_##type##_coding_system_data (Lisp_Coding_System *cs) \ |
| 826 | 428 ) \ |
| 771 | 429 { \ |
| 430 assert (CODING_SYSTEM_TYPE_P (cs, type)); \ | |
| 431 /* Catch accidental use of INITIALIZE_CODING_SYSTEM_TYPE in place \ | |
| 432 of INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA. */ \ | |
| 433 assert (cs->methods->extra_data_size > 0); \ | |
| 434 return (struct type##_coding_system *) cs->data; \ | |
| 435 } \ | |
| 436 \ | |
| 826 | 437 DECLARE_INLINE_HEADER ( \ |
| 438 struct type##_coding_stream * \ | |
| 771 | 439 error_check_##type##_coding_stream_data (struct coding_stream *s) \ |
| 826 | 440 ) \ |
| 771 | 441 { \ |
| 442 assert (XCODING_SYSTEM_TYPE_P (s->codesys, type)); \ | |
| 443 return (struct type##_coding_stream *) s->data; \ | |
| 444 } \ | |
| 445 \ | |
| 826 | 446 DECLARE_INLINE_HEADER ( \ |
| 447 Lisp_Coding_System * \ | |
| 771 | 448 error_check_##type##_coding_system_type (Lisp_Object obj) \ |
| 826 | 449 ) \ |
| 771 | 450 { \ |
| 451 Lisp_Coding_System *cs = XCODING_SYSTEM (obj); \ | |
| 452 assert (CODING_SYSTEM_TYPE_P (cs, type)); \ | |
| 453 return cs; \ | |
| 454 } \ | |
| 455 \ | |
| 456 DECLARE_NOTHING | |
| 457 #else | |
| 458 #define DECLARE_CODING_SYSTEM_TYPE(type) \ | |
| 459 extern struct coding_system_methods * type##_coding_system_methods | |
| 800 | 460 #endif /* ERROR_CHECK_TYPES */ |
| 771 | 461 |
| 462 #define DEFINE_CODING_SYSTEM_TYPE(type) \ | |
| 463 struct coding_system_methods * type##_coding_system_methods | |
| 464 | |
| 1204 | 465 #define DEFINE_CODING_SYSTEM_TYPE_WITH_DATA(type) \ |
| 466 struct coding_system_methods * type##_coding_system_methods; \ | |
| 467 static const struct sized_memory_description \ | |
| 468 type##_coding_system_description_0 = { \ | |
| 469 sizeof (struct type##_coding_system), \ | |
| 470 type##_coding_system_description \ | |
| 471 } | |
| 472 | |
| 771 | 473 #define INITIALIZE_CODING_SYSTEM_TYPE(ty, pred_sym) do { \ |
| 474 ty##_coding_system_methods = \ | |
| 475 xnew_and_zero (struct coding_system_methods); \ | |
| 476 ty##_coding_system_methods->type = Q##ty; \ | |
| 477 ty##_coding_system_methods->extra_description = \ | |
| 1204 | 478 &coding_system_empty_extra_description; \ |
| 479 ty##_coding_system_methods->enumtype = ty##_coding_system; \ | |
| 771 | 480 defsymbol_nodump (&ty##_coding_system_methods->predicate_symbol, \ |
| 481 pred_sym); \ | |
| 482 add_entry_to_coding_system_type_list (ty##_coding_system_methods); \ | |
| 2367 | 483 dump_add_root_block_ptr (&ty##_coding_system_methods, \ |
| 771 | 484 &coding_system_methods_description); \ |
| 485 } while (0) | |
| 486 | |
| 487 #define REINITIALIZE_CODING_SYSTEM_TYPE(type) do { \ | |
| 488 staticpro_nodump (&type##_coding_system_methods->predicate_symbol); \ | |
| 489 } while (0) | |
| 490 | |
| 491 /* This assumes the existence of two structures: | |
| 492 | |
| 493 struct foo_coding_system (attached to the coding system) | |
| 494 struct foo_coding_stream (per coding process, attached to the | |
| 495 struct coding_stream) | |
| 1204 | 496 const struct memory_description foo_coding_system_description[] |
| 497 (data description of struct foo_coding_system) | |
| 771 | 498 |
| 1204 | 499 For an example of how to do the description, see |
| 771 | 500 chain_coding_system_description. |
| 501 */ | |
| 502 #define INITIALIZE_CODING_SYSTEM_TYPE_WITH_DATA(type, pred_sym) \ | |
| 503 do { \ | |
| 504 INITIALIZE_CODING_SYSTEM_TYPE (type, pred_sym); \ | |
| 505 type##_coding_system_methods->extra_data_size = \ | |
| 506 sizeof (struct type##_coding_system); \ | |
| 507 type##_coding_system_methods->extra_description = \ | |
| 1204 | 508 &type##_coding_system_description_0; \ |
| 771 | 509 type##_coding_system_methods->coding_data_size = \ |
| 510 sizeof (struct type##_coding_stream); \ | |
| 511 } while (0) | |
| 512 | |
| 513 /* Declare that coding-system-type TYPE has method METH; used in | |
| 514 initialization routines */ | |
| 515 #define CODING_SYSTEM_HAS_METHOD(type, meth) \ | |
| 516 (type##_coding_system_methods->meth##_method = type##_##meth) | |
| 517 | |
| 518 /***** Macros for accessing coding-system types *****/ | |
| 519 | |
| 520 #define CODING_SYSTEM_TYPE_P(cs, type) \ | |
| 521 ((cs)->methods == type##_coding_system_methods) | |
| 522 #define XCODING_SYSTEM_TYPE_P(cs, type) \ | |
| 523 CODING_SYSTEM_TYPE_P (XCODING_SYSTEM (cs), type) | |
| 524 | |
| 800 | 525 #ifdef ERROR_CHECK_TYPES |
| 771 | 526 # define CODING_SYSTEM_TYPE_DATA(cs, type) \ |
| 527 error_check_##type##_coding_system_data (cs) | |
| 528 #else | |
| 529 # define CODING_SYSTEM_TYPE_DATA(cs, type) \ | |
| 530 ((struct type##_coding_system *) \ | |
| 531 (cs)->data) | |
| 532 #endif | |
| 533 | |
| 534 #define XCODING_SYSTEM_TYPE_DATA(cs, type) \ | |
| 535 CODING_SYSTEM_TYPE_DATA (XCODING_SYSTEM_OF_TYPE (cs, type), type) | |
| 536 | |
| 800 | 537 #ifdef ERROR_CHECK_TYPES |
| 771 | 538 # define XCODING_SYSTEM_OF_TYPE(x, type) \ |
| 539 error_check_##type##_coding_system_type (x) | |
| 540 # define XSETCODING_SYSTEM_OF_TYPE(x, p, type) do \ | |
| 541 { \ | |
| 793 | 542 x = wrap_coding_system (p); \ |
| 543 assert (CODING_SYSTEM_TYPEP (XCODING_SYSTEM (x), type)); \ | |
| 771 | 544 } while (0) |
| 545 #else | |
| 546 # define XCODING_SYSTEM_OF_TYPE(x, type) XCODING_SYSTEM (x) | |
| 793 | 547 # define XSETCODING_SYSTEM_OF_TYPE(x, p, type) do \ |
| 548 { \ | |
| 549 x = wrap_coding_system (p); \ | |
| 550 } while (0) | |
| 771 | 551 #endif /* ERROR_CHECK_TYPE_CHECK */ |
| 552 | |
| 553 #define CODING_SYSTEM_TYPEP(x, type) \ | |
| 554 (CODING_SYSTEMP (x) && CODING_SYSTEM_TYPE_P (XCODING_SYSTEM (x), type)) | |
| 555 #define CHECK_CODING_SYSTEM_OF_TYPE(x, type) do { \ | |
| 556 CHECK_CODING_SYSTEM (x); \ | |
| 557 if (!CODING_SYSTEM_TYPE_P (XCODING_SYSTEM (x), type)) \ | |
| 558 dead_wrong_type_argument \ | |
| 559 (type##_coding_system_methods->predicate_symbol, x); \ | |
| 560 } while (0) | |
| 561 #define CONCHECK_CODING_SYSTEM_OF_TYPE(x, type) do { \ | |
| 562 CONCHECK_CODING_SYSTEM (x); \ | |
| 563 if (!(CODING_SYSTEM_TYPEP (x, type))) \ | |
| 564 x = wrong_type_argument \ | |
| 565 (type##_coding_system_methods->predicate_symbol, x); \ | |
| 566 } while (0) | |
| 567 | |
| 568 #define CODING_SYSTEM_METHODS(codesys) ((codesys)->methods) | |
| 428 | 569 #define CODING_SYSTEM_NAME(codesys) ((codesys)->name) |
| 771 | 570 #define CODING_SYSTEM_DESCRIPTION(codesys) ((codesys)->description) |
| 571 #define CODING_SYSTEM_TYPE(codesys) ((codesys)->methods->type) | |
| 428 | 572 #define CODING_SYSTEM_MNEMONIC(codesys) ((codesys)->mnemonic) |
| 771 | 573 #define CODING_SYSTEM_DOCUMENTATION(codesys) ((codesys)->documentation) |
| 428 | 574 #define CODING_SYSTEM_POST_READ_CONVERSION(codesys) \ |
| 575 ((codesys)->post_read_conversion) | |
| 576 #define CODING_SYSTEM_PRE_WRITE_CONVERSION(codesys) \ | |
| 577 ((codesys)->pre_write_conversion) | |
| 578 #define CODING_SYSTEM_EOL_TYPE(codesys) ((codesys)->eol_type) | |
| 771 | 579 #define CODING_SYSTEM_EOL_LF(codesys) ((codesys)->eol[EOL_LF]) |
| 580 #define CODING_SYSTEM_EOL_CRLF(codesys) ((codesys)->eol[EOL_CRLF]) | |
| 581 #define CODING_SYSTEM_EOL_CR(codesys) ((codesys)->eol[EOL_CR]) | |
| 582 #define CODING_SYSTEM_TEXT_FILE_WRAPPER(codesys) ((codesys)->text_file_wrapper) | |
| 583 #define CODING_SYSTEM_AUTO_EOL_WRAPPER(codesys) ((codesys)->auto_eol_wrapper) | |
| 584 #define CODING_SYSTEM_SUBSIDIARY_PARENT(codesys) ((codesys)->subsidiary_parent) | |
| 585 #define CODING_SYSTEM_CANONICAL(codesys) ((codesys)->canonical) | |
|
4568
1d74a1d115ee
Add #'query-coding-region tests; do the work necessary to get them running.
Aidan Kehoe <kehoea@parhasard.net>
parents:
3017
diff
changeset
|
586 #define CODING_SYSTEM_SAFE_CHARSETS(codesys) ((codesys)->safe_charsets) |
|
1d74a1d115ee
Add #'query-coding-region tests; do the work necessary to get them running.
Aidan Kehoe <kehoea@parhasard.net>
parents:
3017
diff
changeset
|
587 #define CODING_SYSTEM_SAFE_CHARS(codesys) ((codesys)->safe_chars) |
| 428 | 588 |
| 771 | 589 #define CODING_SYSTEM_CHAIN_CHAIN(codesys) \ |
| 590 (CODING_SYSTEM_TYPE_DATA (codesys, chain)->chain) | |
| 591 #define CODING_SYSTEM_CHAIN_COUNT(codesys) \ | |
| 592 (CODING_SYSTEM_TYPE_DATA (codesys, chain)->count) | |
| 593 #define CODING_SYSTEM_CHAIN_CANONICALIZE_AFTER_CODING(codesys) \ | |
| 594 (CODING_SYSTEM_TYPE_DATA (codesys, chain)->canonicalize_after_coding) | |
| 428 | 595 |
| 771 | 596 #define XCODING_SYSTEM_METHODS(codesys) \ |
| 597 CODING_SYSTEM_METHODS (XCODING_SYSTEM (codesys)) | |
| 428 | 598 #define XCODING_SYSTEM_NAME(codesys) \ |
| 599 CODING_SYSTEM_NAME (XCODING_SYSTEM (codesys)) | |
| 771 | 600 #define XCODING_SYSTEM_DESCRIPTION(codesys) \ |
| 601 CODING_SYSTEM_DESCRIPTION (XCODING_SYSTEM (codesys)) | |
| 428 | 602 #define XCODING_SYSTEM_TYPE(codesys) \ |
| 603 CODING_SYSTEM_TYPE (XCODING_SYSTEM (codesys)) | |
| 604 #define XCODING_SYSTEM_MNEMONIC(codesys) \ | |
| 605 CODING_SYSTEM_MNEMONIC (XCODING_SYSTEM (codesys)) | |
| 771 | 606 #define XCODING_SYSTEM_DOCUMENTATION(codesys) \ |
| 607 CODING_SYSTEM_DOCUMENTATION (XCODING_SYSTEM (codesys)) | |
| 428 | 608 #define XCODING_SYSTEM_POST_READ_CONVERSION(codesys) \ |
| 609 CODING_SYSTEM_POST_READ_CONVERSION (XCODING_SYSTEM (codesys)) | |
| 610 #define XCODING_SYSTEM_PRE_WRITE_CONVERSION(codesys) \ | |
| 611 CODING_SYSTEM_PRE_WRITE_CONVERSION (XCODING_SYSTEM (codesys)) | |
| 612 #define XCODING_SYSTEM_EOL_TYPE(codesys) \ | |
| 613 CODING_SYSTEM_EOL_TYPE (XCODING_SYSTEM (codesys)) | |
| 614 #define XCODING_SYSTEM_EOL_LF(codesys) \ | |
| 615 CODING_SYSTEM_EOL_LF (XCODING_SYSTEM (codesys)) | |
| 616 #define XCODING_SYSTEM_EOL_CRLF(codesys) \ | |
| 617 CODING_SYSTEM_EOL_CRLF (XCODING_SYSTEM (codesys)) | |
| 618 #define XCODING_SYSTEM_EOL_CR(codesys) \ | |
| 619 CODING_SYSTEM_EOL_CR (XCODING_SYSTEM (codesys)) | |
| 771 | 620 #define XCODING_SYSTEM_TEXT_FILE_WRAPPER(codesys) \ |
| 621 CODING_SYSTEM_TEXT_FILE_WRAPPER (XCODING_SYSTEM (codesys)) | |
| 622 #define XCODING_SYSTEM_AUTO_EOL_WRAPPER(codesys) \ | |
| 623 CODING_SYSTEM_AUTO_EOL_WRAPPER (XCODING_SYSTEM (codesys)) | |
| 624 #define XCODING_SYSTEM_SUBSIDIARY_PARENT(codesys) \ | |
| 625 CODING_SYSTEM_SUBSIDIARY_PARENT (XCODING_SYSTEM (codesys)) | |
| 626 #define XCODING_SYSTEM_CANONICAL(codesys) \ | |
| 627 CODING_SYSTEM_CANONICAL (XCODING_SYSTEM (codesys)) | |
|
4568
1d74a1d115ee
Add #'query-coding-region tests; do the work necessary to get them running.
Aidan Kehoe <kehoea@parhasard.net>
parents:
3017
diff
changeset
|
628 #define XCODING_SYSTEM_SAFE_CHARSETS(codesys) \ |
|
1d74a1d115ee
Add #'query-coding-region tests; do the work necessary to get them running.
Aidan Kehoe <kehoea@parhasard.net>
parents:
3017
diff
changeset
|
629 CODING_SYSTEM_SAFE_CHARSETS (XCODING_SYSTEM (codesys)) |
|
1d74a1d115ee
Add #'query-coding-region tests; do the work necessary to get them running.
Aidan Kehoe <kehoea@parhasard.net>
parents:
3017
diff
changeset
|
630 #define XCODING_SYSTEM_SAFE_CHARS(codesys) \ |
|
1d74a1d115ee
Add #'query-coding-region tests; do the work necessary to get them running.
Aidan Kehoe <kehoea@parhasard.net>
parents:
3017
diff
changeset
|
631 CODING_SYSTEM_SAFE_CHARS (XCODING_SYSTEM (codesys)) |
| 428 | 632 |
| 771 | 633 #define XCODING_SYSTEM_CHAIN_CHAIN(codesys) \ |
| 634 CODING_SYSTEM_CHAIN_CHAIN (XCODING_SYSTEM (codesys)) | |
| 635 #define XCODING_SYSTEM_CHAIN_COUNT(codesys) \ | |
| 636 CODING_SYSTEM_CHAIN_COUNT (XCODING_SYSTEM (codesys)) | |
| 637 #define XCODING_SYSTEM_CHAIN_CANONICALIZE_AFTER_CODING(codesys) \ | |
| 638 CODING_SYSTEM_CHAIN_CANONICALIZE_AFTER_CODING (XCODING_SYSTEM (codesys)) | |
| 428 | 639 |
| 771 | 640 /**************************************************/ |
| 641 /* Detection */ | |
| 642 /**************************************************/ | |
| 428 | 643 |
| 771 | 644 #define MAX_DETECTOR_CATEGORIES 256 |
| 645 #define MAX_DETECTORS 64 | |
| 428 | 646 |
| 771 | 647 #define MAX_BYTES_PROCESSED_FOR_DETECTION 65536 |
| 428 | 648 |
| 771 | 649 struct detection_state |
| 428 | 650 { |
| 771 | 651 int seen_non_ascii; |
| 652 Bytecount bytes_seen; | |
| 428 | 653 |
| 771 | 654 char categories[MAX_DETECTOR_CATEGORIES]; |
| 655 Bytecount data_offset[MAX_DETECTORS]; | |
| 656 /* ... more data follows; data_offset[detector_##TYPE] points to | |
| 657 the data for that type */ | |
| 428 | 658 }; |
| 659 | |
| 771 | 660 #define DETECTION_STATE_DATA(st, type) \ |
| 661 ((struct type##_detector *) \ | |
| 662 ((char *) (st) + (st)->data_offset[detector_##type])) | |
| 428 | 663 |
| 448 | 664 /* Distinguishable categories of encodings. |
| 665 | |
| 666 This list determines the initial priority of the categories. | |
| 667 | |
| 668 For better or worse, currently Mule files are encoded in 7-bit ISO 2022. | |
| 669 For this reason, under Mule ISO_7 gets highest priority. | |
| 670 | |
| 671 Putting NO_CONVERSION second prevents "binary corruption" in the | |
| 672 default case in all but the (presumably) extremely rare case of a | |
| 673 binary file which contains redundant escape sequences but no 8-bit | |
| 674 characters. | |
| 675 | |
| 676 The remaining priorities are based on perceived "internationalization | |
| 677 political correctness." An exception is UCS-4 at the bottom, since | |
| 678 basically everything is compatible with UCS-4, but it is likely to | |
| 679 be very rare as an external encoding. */ | |
| 680 | |
| 771 | 681 /* Macros to define code of control characters for ISO2022's functions. */ |
| 682 /* Used by the detection routines of other coding system types as well. */ | |
| 683 /* code */ /* function */ | |
| 684 #define ISO_CODE_LF 0x0A /* line-feed */ | |
| 685 #define ISO_CODE_CR 0x0D /* carriage-return */ | |
| 686 #define ISO_CODE_SO 0x0E /* shift-out */ | |
| 687 #define ISO_CODE_SI 0x0F /* shift-in */ | |
| 688 #define ISO_CODE_ESC 0x1B /* escape */ | |
| 689 #define ISO_CODE_DEL 0x7F /* delete */ | |
| 690 #define ISO_CODE_SS2 0x8E /* single-shift-2 */ | |
| 691 #define ISO_CODE_SS3 0x8F /* single-shift-3 */ | |
| 692 #define ISO_CODE_CSI 0x9B /* control-sequence-introduce */ | |
| 693 | |
| 694 enum detection_result | |
| 695 { | |
| 696 /* Basically means a magic cookie was seen indicating this type, or | |
| 697 something similar. */ | |
| 698 DET_NEAR_CERTAINTY = 4, | |
| 699 DET_HIGHEST = 4, | |
| 700 /* Characteristics seen that are unlikely to be other coding system types | |
| 701 -- e.g. ISO-2022 escape sequences, or perhaps a consistent pattern of | |
| 702 alternating zero bytes in UTF-16, along with Unicode LF or CRLF | |
| 703 sequences at regular intervals. (Zero bytes are unlikely or impossible | |
| 704 in most text encodings.) */ | |
| 705 DET_QUITE_PROBABLE = 3, | |
| 706 /* Strong or medium statistical likelihood. At least some | |
| 707 characteristics seen that match what's normally found in this encoding | |
| 708 -- e.g. in Shift-JIS, a number of two-byte Japanese character | |
| 709 sequences in the right range, and nothing out of range; or in Unicode, | |
| 710 much higher statistical variance in the odd bytes than in the even | |
| 711 bytes, or vice-versa (perhaps the presence of regular EOL sequences | |
| 712 would bump this too to DET_QUITE_PROBABLE). This is quite often a | |
| 713 statistical test. */ | |
| 714 DET_SOMEWHAT_LIKELY = 2, | |
| 715 /* Weak statistical likelihood. Pretty much any features at all that | |
| 716 characterize this encoding, and nothing that rules against it. */ | |
| 717 DET_SLIGHTLY_LIKELY = 1, | |
| 718 /* Default state. Perhaps it indicates pure ASCII or something similarly | |
| 719 vague seen in Shift-JIS, or, exactly as the level says, it might mean | |
| 720 in a statistical-based detector that the pros and cons are balanced | |
| 721 out. This is also the lowest level that will be accepted by the | |
| 722 auto-detector without asking the user: If all available detectors | |
| 723 report lower levels for all categories with attached coding systems, | |
| 724 the user will be shown the results and explicitly prompted for action. | |
| 725 The user will also be prompted if this is the highest available level | |
| 726 and more than one detector reports the level. (See below about the | |
| 727 consequent necessity of an "ASCII" detector, which will return level 1 | |
| 728 or higher for most plain text files.) */ | |
| 729 DET_AS_LIKELY_AS_UNLIKELY = 0, | |
| 730 /* Some characteristics seen that are unusual for this encoding -- | |
| 731 e.g. unusual control characters in a plain-text encoding, lots of | |
| 732 8-bit characters, or little statistical variance in the odd and even | |
| 733 bytes in UTF-16. */ | |
| 734 DET_SOMEWHAT_UNLIKELY = -1, | |
| 735 /* This indicates that there is very little chance the data is in the | |
| 736 right format; this is probably the lowest level you can get when | |
| 737 presenting random binary data to a text file, because there are no | |
| 738 "specific sequences" you can see that would totally rule out | |
| 739 recognition. */ | |
| 740 DET_QUITE_IMPROBABLE = -2, | |
| 741 /* An erroneous sequence was seen. */ | |
| 742 DET_NEARLY_IMPOSSIBLE = -3, | |
| 1429 | 743 DET_LOWEST = -3 |
| 771 | 744 }; |
| 745 | |
| 746 extern int coding_detector_count; | |
| 747 extern int coding_detector_category_count; | |
| 748 | |
| 749 struct detector_category | |
| 428 | 750 { |
| 771 | 751 int id; |
| 752 Lisp_Object sym; | |
| 753 }; | |
| 754 | |
| 755 typedef struct | |
| 756 { | |
| 757 Dynarr_declare (struct detector_category); | |
| 758 } detector_category_dynarr; | |
| 759 | |
| 760 struct detector | |
| 761 { | |
| 762 int id; | |
| 763 detector_category_dynarr *cats; | |
| 764 Bytecount data_size; | |
| 765 /* Detect method: Required. */ | |
| 766 void (*detect_method) (struct detection_state *st, | |
| 767 const unsigned char *src, Bytecount n); | |
| 768 /* Finalize detection state method: Clean up any allocated data in the | |
| 769 detection state. Called only once (NOT called at disksave time). | |
| 770 Optional. */ | |
| 771 void (*finalize_detection_state_method) (struct detection_state *st); | |
| 428 | 772 }; |
| 773 | |
| 771 | 774 /* Lvalue for a particular detection result -- detection state ST, |
| 775 category CAT */ | |
| 776 #define DET_RESULT(st, cat) ((st)->categories[detector_category_##cat]) | |
| 777 /* In state ST, set all detection results associated with detector DET to | |
| 778 RESULT. */ | |
| 779 #define SET_DET_RESULTS(st, det, result) \ | |
| 780 set_detection_results (st, detector_##det, result) | |
| 781 | |
| 782 typedef struct | |
| 783 { | |
| 784 Dynarr_declare (struct detector); | |
| 785 } detector_dynarr; | |
| 786 | |
| 787 extern detector_dynarr *all_coding_detectors; | |
| 788 | |
| 789 #define DEFINE_DETECTOR_CATEGORY(detector, cat) \ | |
| 790 int detector_category_##cat | |
| 791 #define DECLARE_DETECTOR_CATEGORY(detector, cat) \ | |
| 792 extern int detector_category_##cat | |
| 793 #define INITIALIZE_DETECTOR_CATEGORY(detector, cat) \ | |
| 794 do { \ | |
| 795 struct detector_category dog; \ | |
| 796 xzero (dog); \ | |
| 797 detector_category_##cat = coding_detector_category_count++; \ | |
| 798 dump_add_opaque_int (&detector_category_##cat); \ | |
| 799 dog.id = detector_category_##cat; \ | |
| 800 dog.sym = Q##cat; \ | |
| 801 Dynarr_add (Dynarr_at (all_coding_detectors, detector_##detector).cats, \ | |
| 802 dog); \ | |
| 803 } while (0) | |
| 804 | |
| 805 #define DEFINE_DETECTOR(Detector) \ | |
| 806 int detector_##Detector | |
| 807 #define DECLARE_DETECTOR(Detector) \ | |
| 808 extern int detector_##Detector | |
| 809 #define INITIALIZE_DETECTOR(Detector) \ | |
| 810 do { \ | |
| 811 struct detector det; \ | |
| 812 xzero (det); \ | |
| 813 detector_##Detector = coding_detector_count++; \ | |
| 814 dump_add_opaque_int (&detector_##Detector); \ | |
| 815 det.id = detector_##Detector; \ | |
| 816 det.cats = Dynarr_new2 (detector_category_dynarr, \ | |
| 817 struct detector_category); \ | |
| 818 det.data_size = sizeof (struct Detector##_detector); \ | |
| 819 Dynarr_add (all_coding_detectors, det); \ | |
| 820 } while (0) | |
| 821 #define DETECTOR_HAS_METHOD(Detector, Meth) \ | |
| 822 Dynarr_at (all_coding_detectors, detector_##Detector).Meth##_method = \ | |
| 802 | 823 Detector##_##Meth |
| 771 | 824 |
| 825 | |
| 826 /**************************************************/ | |
| 827 /* Decoding/Encoding */ | |
| 828 /**************************************************/ | |
| 829 | |
| 830 /* Is the source (SOURCEP == 1) or sink (SOURCEP == 0) when encoding specified | |
| 831 in characters? */ | |
| 832 | |
| 833 enum source_or_sink | |
| 834 { | |
| 835 CODING_SOURCE, | |
| 836 CODING_SINK | |
| 837 }; | |
| 838 | |
| 839 enum encode_decode | |
| 840 { | |
| 841 CODING_ENCODE, | |
| 842 CODING_DECODE | |
| 843 }; | |
| 844 | |
| 845 /* Data structure attached to an lstream of type `coding', | |
| 846 containing values specific to the coding process. Additional | |
| 847 data is stored in the DATA field below; the exact form of that data | |
| 848 is controlled by the type of the coding system that governs the | |
| 849 conversion (field CODESYS). CODESYS may be set at any time | |
| 850 throughout the lifetime of the lstream and possibly more than once. | |
| 851 See long comment above for more info. */ | |
| 852 | |
| 853 struct coding_stream | |
| 854 { | |
| 1204 | 855 /* Enumerated constant listing which type of console this is (TTY, X, |
| 856 MS-Windows, etc.). This duplicates the method structure in | |
| 857 XCODING_SYSTEM (str->codesys)->methods->type, which formerly was the | |
| 858 only way to determine the coding system type. We need this constant | |
| 859 now for KKCC, so that it can be used in an XD_UNION clause to | |
| 860 determine the Lisp objects in the type-specific data. */ | |
| 861 enum coding_system_variant type; | |
| 862 | |
| 771 | 863 /* Coding system that governs the conversion. */ |
| 864 Lisp_Object codesys; | |
| 865 /* Original coding system, pre-canonicalization. */ | |
| 866 Lisp_Object orig_codesys; | |
| 867 | |
| 868 /* Back pointer to current stream. */ | |
| 869 Lstream *us; | |
| 870 | |
| 871 /* Stream that we read the unprocessed data from or write the processed | |
| 872 data to. */ | |
| 873 Lstream *other_end; | |
| 874 | |
| 875 /* In order to handle both reading to and writing from a coding stream, | |
| 876 we phrase the conversion methods like write methods -- we can | |
| 877 implement reading in terms of a write method but not vice-versa, | |
| 878 because the write method is forced to take only what it's given but | |
| 879 the read method can read more data from the other end if necessary. | |
| 880 On the other hand, the write method is free to generate all the data | |
| 2297 | 881 it wants (and just write it to the other end), but the read method |
| 771 | 882 can return only as much as was asked for, so we need to implement our |
| 883 own buffering. */ | |
| 884 | |
| 885 /* If we are reading, then we can return only a fixed amount of data, but | |
| 886 the converter is free to return as much as it wants, so we direct it | |
| 887 to store the data here and lop off chunks as we need them. If we are | |
| 888 writing, we use this because the converter takes a Dynarr but we are | |
| 889 supposed to write into a fixed buffer. (NOTE: This introduces an extra | |
| 890 memory copy.) */ | |
| 891 unsigned_char_dynarr *convert_to; | |
| 892 | |
| 893 /* The conversion method might reject some of the data -- this typically | |
| 894 includes partial characters, partial escape sequences, etc. When | |
| 895 writing, we just pass the rejection up to the Lstream module, and it | |
| 896 will buffer the data. When reading, however, we need to do the | |
| 897 buffering ourselves, and we put it here, combined with newly read | |
| 898 data. */ | |
| 899 unsigned_char_dynarr *convert_from; | |
| 900 | |
| 901 /* If set, this is the last chunk of data being processed. When this is | |
| 902 finished, output any necessary terminating control characters, escape | |
| 903 sequences, etc. */ | |
| 904 unsigned int eof:1; | |
| 905 | |
| 906 /* CH holds a partially built-up character. This is really part of the | |
| 907 state-dependent data and should be moved there. */ | |
| 908 unsigned int ch; | |
| 909 | |
| 910 /* Coding-system-specific data holding extra state about the | |
| 911 conversion. Logically a struct TYPE_coding_stream; a pointer | |
| 800 | 912 to such a struct, with (when ERROR_CHECK_TYPES is defined) |
| 771 | 913 error-checking that this is really a structure of that type |
| 914 (checking the corresponding coding system type) can be retrieved using | |
| 915 CODING_STREAM_TYPE_DATA(). Allocated at the same time that | |
| 916 CODESYS is set (which may occur at any time, even multiple times, | |
| 917 during the lifetime of the stream). The size comes from | |
| 918 methods->coding_data_size. */ | |
| 919 void *data; | |
| 920 | |
| 921 enum encode_decode direction; | |
| 922 | |
| 800 | 923 /* If set, don't close the stream at the other end when being closed. */ |
| 924 unsigned int no_close_other:1; | |
| 802 | 925 /* If set, read only one byte at a time from other end to avoid any |
| 926 possible blocking. */ | |
| 927 unsigned int one_byte_at_a_time:1; | |
| 814 | 928 /* If set, and we're a read stream, we init char mode on ourselves as |
| 929 necessary to prevent the caller from getting partial characters. (the | |
| 930 default) */ | |
| 931 unsigned int set_char_mode_on_us_when_reading:1; | |
| 800 | 932 |
| 771 | 933 /* #### Temporary test */ |
| 934 unsigned int finalized:1; | |
| 935 }; | |
| 936 | |
| 937 #define CODING_STREAM_DATA(stream) LSTREAM_TYPE_DATA (stream, coding) | |
| 938 | |
| 800 | 939 #ifdef ERROR_CHECK_TYPES |
| 771 | 940 # define CODING_STREAM_TYPE_DATA(s, type) \ |
| 941 error_check_##type##_coding_stream_data (s) | |
| 942 #else | |
| 943 # define CODING_STREAM_TYPE_DATA(s, type) \ | |
| 944 ((struct type##_coding_stream *) (s)->data) | |
| 945 #endif | |
| 946 | |
| 947 /* C should be a binary character in the range 0 - 255; convert | |
| 948 to internal format and add to Dynarr DST. */ | |
| 949 | |
| 428 | 950 #ifdef MULE |
| 771 | 951 |
| 952 #define DECODE_ADD_BINARY_CHAR(c, dst) \ | |
| 953 do { \ | |
| 826 | 954 if (byte_ascii_p (c)) \ |
| 771 | 955 Dynarr_add (dst, c); \ |
| 826 | 956 else if (byte_c1_p (c)) \ |
| 771 | 957 { \ |
| 958 Dynarr_add (dst, LEADING_BYTE_CONTROL_1); \ | |
| 959 Dynarr_add (dst, c + 0x20); \ | |
| 960 } \ | |
| 961 else \ | |
| 962 { \ | |
| 963 Dynarr_add (dst, LEADING_BYTE_LATIN_ISO8859_1); \ | |
| 964 Dynarr_add (dst, c); \ | |
| 965 } \ | |
| 966 } while (0) | |
| 967 | |
| 968 #else /* not MULE */ | |
| 969 | |
| 970 #define DECODE_ADD_BINARY_CHAR(c, dst) \ | |
| 971 do { \ | |
| 972 Dynarr_add (dst, c); \ | |
| 973 } while (0) | |
| 974 | |
| 975 #endif /* MULE */ | |
| 976 | |
| 977 #define DECODE_OUTPUT_PARTIAL_CHAR(ch, dst) \ | |
| 978 do { \ | |
| 979 if (ch) \ | |
| 980 { \ | |
| 981 DECODE_ADD_BINARY_CHAR (ch, dst); \ | |
| 982 ch = 0; \ | |
| 983 } \ | |
| 984 } while (0) | |
| 428 | 985 |
| 986 #ifdef MULE | |
| 987 /* Convert shift-JIS code (sj1, sj2) into internal string | |
| 988 representation (c1, c2). (The leading byte is assumed.) */ | |
| 989 | |
| 771 | 990 #define DECODE_SHIFT_JIS(sj1, sj2, c1, c2) \ |
| 428 | 991 do { \ |
| 992 int I1 = sj1, I2 = sj2; \ | |
| 993 if (I2 >= 0x9f) \ | |
| 994 c1 = (I1 << 1) - ((I1 >= 0xe0) ? 0xe0 : 0x60), \ | |
| 995 c2 = I2 + 2; \ | |
| 996 else \ | |
| 997 c1 = (I1 << 1) - ((I1 >= 0xe0) ? 0xe1 : 0x61), \ | |
| 998 c2 = I2 + ((I2 >= 0x7f) ? 0x60 : 0x61); \ | |
| 999 } while (0) | |
| 1000 | |
| 1001 /* Convert the internal string representation of a Shift-JIS character | |
| 1002 (c1, c2) into Shift-JIS code (sj1, sj2). The leading byte is | |
| 1003 assumed. */ | |
| 1004 | |
| 771 | 1005 #define ENCODE_SHIFT_JIS(c1, c2, sj1, sj2) \ |
| 428 | 1006 do { \ |
| 1007 int I1 = c1, I2 = c2; \ | |
| 1008 if (I1 & 1) \ | |
| 1009 sj1 = (I1 >> 1) + ((I1 < 0xdf) ? 0x31 : 0x71), \ | |
| 1010 sj2 = I2 - ((I2 >= 0xe0) ? 0x60 : 0x61); \ | |
| 1011 else \ | |
| 1012 sj1 = (I1 >> 1) + ((I1 < 0xdf) ? 0x30 : 0x70), \ | |
| 1013 sj2 = I2 - 2; \ | |
| 1014 } while (0) | |
| 1015 #endif /* MULE */ | |
| 1016 | |
| 771 | 1017 DECLARE_CODING_SYSTEM_TYPE (no_conversion); |
| 1018 DECLARE_CODING_SYSTEM_TYPE (convert_eol); | |
| 1019 #if 0 | |
| 1020 DECLARE_CODING_SYSTEM_TYPE (text_file_wrapper); | |
| 1021 #endif /* 0 */ | |
| 1022 DECLARE_CODING_SYSTEM_TYPE (undecided); | |
| 1023 DECLARE_CODING_SYSTEM_TYPE (chain); | |
| 1024 | |
| 1025 #ifdef DEBUG_XEMACS | |
| 1026 DECLARE_CODING_SYSTEM_TYPE (internal); | |
| 1027 #endif | |
| 1028 | |
| 1029 #ifdef MULE | |
| 1030 DECLARE_CODING_SYSTEM_TYPE (iso2022); | |
| 1031 DECLARE_CODING_SYSTEM_TYPE (ccl); | |
| 1032 DECLARE_CODING_SYSTEM_TYPE (shift_jis); | |
| 1033 DECLARE_CODING_SYSTEM_TYPE (big5); | |
| 1034 #endif | |
| 1035 | |
| 1036 #ifdef HAVE_ZLIB | |
| 1037 DECLARE_CODING_SYSTEM_TYPE (gzip); | |
| 1038 #endif | |
| 428 | 1039 |
| 771 | 1040 DECLARE_CODING_SYSTEM_TYPE (unicode); |
| 428 | 1041 |
| 1315 | 1042 #ifdef WIN32_ANY |
| 771 | 1043 DECLARE_CODING_SYSTEM_TYPE (mswindows_multibyte_to_unicode); |
| 1044 DECLARE_CODING_SYSTEM_TYPE (mswindows_multibyte); | |
| 428 | 1045 #endif |
| 771 | 1046 |
| 1047 Lisp_Object coding_stream_detected_coding_system (Lstream *stream); | |
| 1048 Lisp_Object coding_stream_coding_system (Lstream *stream); | |
| 1049 void set_coding_stream_coding_system (Lstream *stream, | |
| 1050 Lisp_Object codesys); | |
| 1051 Lisp_Object detect_coding_stream (Lisp_Object stream); | |
| 867 | 1052 Ichar decode_big5_char (int o1, int o2); |
| 771 | 1053 void add_entry_to_coding_system_type_list (struct coding_system_methods *m); |
| 1054 Lisp_Object make_internal_coding_system (Lisp_Object existing, | |
|
4528
726060ee587c
First draft of g++ 4.3 warning removal patch. Builds. *Needs ChangeLogs.*
Stephen J. Turnbull <stephen@xemacs.org>
parents:
4522
diff
changeset
|
1055 const Ascbyte *prefix, |
| 771 | 1056 Lisp_Object type, |
| 1057 Lisp_Object description, | |
| 1058 Lisp_Object props); | |
| 802 | 1059 |
| 814 | 1060 #define LSTREAM_FL_NO_CLOSE_OTHER (1 << 16) |
| 1061 #define LSTREAM_FL_READ_ONE_BYTE_AT_A_TIME (1 << 17) | |
| 1062 #define LSTREAM_FL_NO_INIT_CHAR_MODE_WHEN_READING (1 << 18) | |
| 1063 | |
| 771 | 1064 Lisp_Object make_coding_input_stream (Lstream *stream, Lisp_Object codesys, |
| 800 | 1065 enum encode_decode direction, |
| 802 | 1066 int flags); |
| 771 | 1067 Lisp_Object make_coding_output_stream (Lstream *stream, Lisp_Object codesys, |
| 800 | 1068 enum encode_decode direction, |
| 802 | 1069 int flags); |
| 771 | 1070 void set_detection_results (struct detection_state *st, int detector, |
| 1071 int given); | |
| 428 | 1072 |
| 440 | 1073 #endif /* INCLUDED_file_coding_h_ */ |
| 1074 |