Commit | Line | Data |
---|---|---|
4ed46869 | 1 | /* CCL (Code Conversion Language) interpreter. |
75c8c592 | 2 | Copyright (C) 1995, 1997 Electrotechnical Laboratory, JAPAN. |
8f924df7 | 3 | Licensed to the Free Software Foundation. |
d80dc57e | 4 | Copyright (C) 2001, 2002 Free Software Foundation, Inc. |
8f924df7 | 5 | Copyright (C) 2003 |
c10842ea KH |
6 | National Institute of Advanced Industrial Science and Technology (AIST) |
7 | Registration Number H13PRO009 | |
4ed46869 | 8 | |
369314dc KH |
9 | This file is part of GNU Emacs. |
10 | ||
11 | GNU Emacs is free software; you can redistribute it and/or modify | |
12 | it under the terms of the GNU General Public License as published by | |
13 | the Free Software Foundation; either version 2, or (at your option) | |
14 | any later version. | |
4ed46869 | 15 | |
369314dc KH |
16 | GNU Emacs is distributed in the hope that it will be useful, |
17 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
18 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
19 | GNU General Public License for more details. | |
4ed46869 | 20 | |
369314dc KH |
21 | You should have received a copy of the GNU General Public License |
22 | along with GNU Emacs; see the file COPYING. If not, write to | |
23 | the Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
24 | Boston, MA 02111-1307, USA. */ | |
4ed46869 | 25 | |
4ed46869 | 26 | #include <config.h> |
dfcf069d | 27 | |
68c45bf0 PE |
28 | #include <stdio.h> |
29 | ||
4ed46869 | 30 | #include "lisp.h" |
c10842ea | 31 | #include "character.h" |
4ed46869 KH |
32 | #include "charset.h" |
33 | #include "ccl.h" | |
34 | #include "coding.h" | |
35 | ||
c10842ea KH |
36 | Lisp_Object Qccl, Qcclp; |
37 | ||
20398ea4 | 38 | /* This contains all code conversion map available to CCL. */ |
8146262a | 39 | Lisp_Object Vcode_conversion_map_vector; |
e34b1164 | 40 | |
4ed46869 KH |
41 | /* Alist of fontname patterns vs corresponding CCL program. */ |
42 | Lisp_Object Vfont_ccl_encoder_alist; | |
43 | ||
6ae21908 KH |
44 | /* This symbol is a property which assocates with ccl program vector. |
45 | Ex: (get 'ccl-big5-encoder 'ccl-program) returns ccl program vector. */ | |
e34b1164 KH |
46 | Lisp_Object Qccl_program; |
47 | ||
8146262a KH |
48 | /* These symbols are properties which associate with code conversion |
49 | map and their ID respectively. */ | |
50 | Lisp_Object Qcode_conversion_map; | |
51 | Lisp_Object Qcode_conversion_map_id; | |
e34b1164 | 52 | |
6ae21908 KH |
53 | /* Symbols of ccl program have this property, a value of the property |
54 | is an index for Vccl_protram_table. */ | |
55 | Lisp_Object Qccl_program_idx; | |
56 | ||
5232fa7b KH |
57 | /* Table of registered CCL programs. Each element is a vector of |
58 | NAME, CCL_PROG, and RESOLVEDP where NAME (symbol) is the name of | |
59 | the program, CCL_PROG (vector) is the compiled code of the program, | |
60 | RESOLVEDP (t or nil) is the flag to tell if symbols in CCL_PROG is | |
61 | already resolved to index numbers or not. */ | |
4ed46869 KH |
62 | Lisp_Object Vccl_program_table; |
63 | ||
d80dc57e DL |
64 | /* Vector of registered hash tables for translation. */ |
65 | Lisp_Object Vtranslation_hash_table_vector; | |
66 | ||
67 | /* Return a hash table of id number ID. */ | |
68 | #define GET_HASH_TABLE(id) \ | |
69 | (XHASH_TABLE (XCDR(XVECTOR(Vtranslation_hash_table_vector)->contents[(id)]))) | |
d80dc57e | 70 | |
d325055a DL |
71 | extern int charset_unicode; |
72 | ||
4ed46869 KH |
73 | /* CCL (Code Conversion Language) is a simple language which has |
74 | operations on one input buffer, one output buffer, and 7 registers. | |
75 | The syntax of CCL is described in `ccl.el'. Emacs Lisp function | |
76 | `ccl-compile' compiles a CCL program and produces a CCL code which | |
77 | is a vector of integers. The structure of this vector is as | |
78 | follows: The 1st element: buffer-magnification, a factor for the | |
79 | size of output buffer compared with the size of input buffer. The | |
80 | 2nd element: address of CCL code to be executed when encountered | |
81 | with end of input stream. The 3rd and the remaining elements: CCL | |
82 | codes. */ | |
83 | ||
84 | /* Header of CCL compiled code */ | |
85 | #define CCL_HEADER_BUF_MAG 0 | |
86 | #define CCL_HEADER_EOF 1 | |
87 | #define CCL_HEADER_MAIN 2 | |
88 | ||
89 | /* CCL code is a sequence of 28-bit non-negative integers (i.e. the | |
90 | MSB is always 0), each contains CCL command and/or arguments in the | |
91 | following format: | |
92 | ||
93 | |----------------- integer (28-bit) ------------------| | |
94 | |------- 17-bit ------|- 3-bit --|- 3-bit --|- 5-bit -| | |
95 | |--constant argument--|-register-|-register-|-command-| | |
96 | ccccccccccccccccc RRR rrr XXXXX | |
97 | or | |
98 | |------- relative address -------|-register-|-command-| | |
99 | cccccccccccccccccccc rrr XXXXX | |
100 | or | |
101 | |------------- constant or other args ----------------| | |
102 | cccccccccccccccccccccccccccc | |
103 | ||
104 | where, `cc...c' is a non-negative integer indicating constant value | |
105 | (the left most `c' is always 0) or an absolute jump address, `RRR' | |
106 | and `rrr' are CCL register number, `XXXXX' is one of the following | |
107 | CCL commands. */ | |
108 | ||
109 | /* CCL commands | |
110 | ||
111 | Each comment fields shows one or more lines for command syntax and | |
112 | the following lines for semantics of the command. In semantics, IC | |
113 | stands for Instruction Counter. */ | |
114 | ||
115 | #define CCL_SetRegister 0x00 /* Set register a register value: | |
116 | 1:00000000000000000RRRrrrXXXXX | |
117 | ------------------------------ | |
118 | reg[rrr] = reg[RRR]; | |
119 | */ | |
120 | ||
121 | #define CCL_SetShortConst 0x01 /* Set register a short constant value: | |
122 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
123 | ------------------------------ | |
124 | reg[rrr] = CCCCCCCCCCCCCCCCCCC; | |
125 | */ | |
126 | ||
127 | #define CCL_SetConst 0x02 /* Set register a constant value: | |
128 | 1:00000000000000000000rrrXXXXX | |
129 | 2:CONSTANT | |
130 | ------------------------------ | |
131 | reg[rrr] = CONSTANT; | |
132 | IC++; | |
133 | */ | |
134 | ||
135 | #define CCL_SetArray 0x03 /* Set register an element of array: | |
136 | 1:CCCCCCCCCCCCCCCCCRRRrrrXXXXX | |
137 | 2:ELEMENT[0] | |
138 | 3:ELEMENT[1] | |
139 | ... | |
140 | ------------------------------ | |
141 | if (0 <= reg[RRR] < CC..C) | |
142 | reg[rrr] = ELEMENT[reg[RRR]]; | |
143 | IC += CC..C; | |
144 | */ | |
145 | ||
146 | #define CCL_Jump 0x04 /* Jump: | |
147 | 1:A--D--D--R--E--S--S-000XXXXX | |
148 | ------------------------------ | |
149 | IC += ADDRESS; | |
150 | */ | |
151 | ||
152 | /* Note: If CC..C is greater than 0, the second code is omitted. */ | |
153 | ||
154 | #define CCL_JumpCond 0x05 /* Jump conditional: | |
155 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
156 | ------------------------------ | |
157 | if (!reg[rrr]) | |
158 | IC += ADDRESS; | |
159 | */ | |
160 | ||
161 | ||
162 | #define CCL_WriteRegisterJump 0x06 /* Write register and jump: | |
163 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
164 | ------------------------------ | |
165 | write (reg[rrr]); | |
166 | IC += ADDRESS; | |
167 | */ | |
168 | ||
169 | #define CCL_WriteRegisterReadJump 0x07 /* Write register, read, and jump: | |
170 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
171 | 2:A--D--D--R--E--S--S-rrrYYYYY | |
172 | ----------------------------- | |
173 | write (reg[rrr]); | |
174 | IC++; | |
175 | read (reg[rrr]); | |
176 | IC += ADDRESS; | |
177 | */ | |
178 | /* Note: If read is suspended, the resumed execution starts from the | |
179 | second code (YYYYY == CCL_ReadJump). */ | |
180 | ||
181 | #define CCL_WriteConstJump 0x08 /* Write constant and jump: | |
182 | 1:A--D--D--R--E--S--S-000XXXXX | |
183 | 2:CONST | |
184 | ------------------------------ | |
185 | write (CONST); | |
186 | IC += ADDRESS; | |
187 | */ | |
188 | ||
189 | #define CCL_WriteConstReadJump 0x09 /* Write constant, read, and jump: | |
190 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
191 | 2:CONST | |
192 | 3:A--D--D--R--E--S--S-rrrYYYYY | |
193 | ----------------------------- | |
194 | write (CONST); | |
195 | IC += 2; | |
196 | read (reg[rrr]); | |
197 | IC += ADDRESS; | |
198 | */ | |
199 | /* Note: If read is suspended, the resumed execution starts from the | |
200 | second code (YYYYY == CCL_ReadJump). */ | |
201 | ||
202 | #define CCL_WriteStringJump 0x0A /* Write string and jump: | |
203 | 1:A--D--D--R--E--S--S-000XXXXX | |
204 | 2:LENGTH | |
205 | 3:0000STRIN[0]STRIN[1]STRIN[2] | |
206 | ... | |
207 | ------------------------------ | |
208 | write_string (STRING, LENGTH); | |
209 | IC += ADDRESS; | |
210 | */ | |
211 | ||
212 | #define CCL_WriteArrayReadJump 0x0B /* Write an array element, read, and jump: | |
213 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
214 | 2:LENGTH | |
215 | 3:ELEMENET[0] | |
216 | 4:ELEMENET[1] | |
217 | ... | |
218 | N:A--D--D--R--E--S--S-rrrYYYYY | |
219 | ------------------------------ | |
220 | if (0 <= reg[rrr] < LENGTH) | |
221 | write (ELEMENT[reg[rrr]]); | |
222 | IC += LENGTH + 2; (... pointing at N+1) | |
223 | read (reg[rrr]); | |
224 | IC += ADDRESS; | |
225 | */ | |
226 | /* Note: If read is suspended, the resumed execution starts from the | |
887bfbd7 | 227 | Nth code (YYYYY == CCL_ReadJump). */ |
4ed46869 KH |
228 | |
229 | #define CCL_ReadJump 0x0C /* Read and jump: | |
230 | 1:A--D--D--R--E--S--S-rrrYYYYY | |
231 | ----------------------------- | |
232 | read (reg[rrr]); | |
233 | IC += ADDRESS; | |
234 | */ | |
235 | ||
236 | #define CCL_Branch 0x0D /* Jump by branch table: | |
237 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
238 | 2:A--D--D--R--E-S-S[0]000XXXXX | |
239 | 3:A--D--D--R--E-S-S[1]000XXXXX | |
240 | ... | |
241 | ------------------------------ | |
242 | if (0 <= reg[rrr] < CC..C) | |
243 | IC += ADDRESS[reg[rrr]]; | |
244 | else | |
245 | IC += ADDRESS[CC..C]; | |
246 | */ | |
247 | ||
248 | #define CCL_ReadRegister 0x0E /* Read bytes into registers: | |
249 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
250 | 2:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
251 | ... | |
252 | ------------------------------ | |
253 | while (CCC--) | |
254 | read (reg[rrr]); | |
255 | */ | |
256 | ||
257 | #define CCL_WriteExprConst 0x0F /* write result of expression: | |
258 | 1:00000OPERATION000RRR000XXXXX | |
259 | 2:CONSTANT | |
260 | ------------------------------ | |
261 | write (reg[RRR] OPERATION CONSTANT); | |
262 | IC++; | |
263 | */ | |
264 | ||
265 | /* Note: If the Nth read is suspended, the resumed execution starts | |
266 | from the Nth code. */ | |
267 | ||
268 | #define CCL_ReadBranch 0x10 /* Read one byte into a register, | |
269 | and jump by branch table: | |
270 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
271 | 2:A--D--D--R--E-S-S[0]000XXXXX | |
272 | 3:A--D--D--R--E-S-S[1]000XXXXX | |
273 | ... | |
274 | ------------------------------ | |
275 | read (read[rrr]); | |
276 | if (0 <= reg[rrr] < CC..C) | |
277 | IC += ADDRESS[reg[rrr]]; | |
278 | else | |
279 | IC += ADDRESS[CC..C]; | |
280 | */ | |
281 | ||
282 | #define CCL_WriteRegister 0x11 /* Write registers: | |
283 | 1:CCCCCCCCCCCCCCCCCCCrrrXXXXX | |
284 | 2:CCCCCCCCCCCCCCCCCCCrrrXXXXX | |
285 | ... | |
286 | ------------------------------ | |
287 | while (CCC--) | |
288 | write (reg[rrr]); | |
289 | ... | |
290 | */ | |
291 | ||
292 | /* Note: If the Nth write is suspended, the resumed execution | |
293 | starts from the Nth code. */ | |
294 | ||
295 | #define CCL_WriteExprRegister 0x12 /* Write result of expression | |
296 | 1:00000OPERATIONRrrRRR000XXXXX | |
297 | ------------------------------ | |
298 | write (reg[RRR] OPERATION reg[Rrr]); | |
299 | */ | |
300 | ||
e34b1164 | 301 | #define CCL_Call 0x13 /* Call the CCL program whose ID is |
5232fa7b KH |
302 | CC..C or cc..c. |
303 | 1:CCCCCCCCCCCCCCCCCCCCFFFXXXXX | |
304 | [2:00000000cccccccccccccccccccc] | |
4ed46869 | 305 | ------------------------------ |
5232fa7b KH |
306 | if (FFF) |
307 | call (cc..c) | |
308 | IC++; | |
309 | else | |
310 | call (CC..C) | |
4ed46869 KH |
311 | */ |
312 | ||
313 | #define CCL_WriteConstString 0x14 /* Write a constant or a string: | |
314 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
315 | [2:0000STRIN[0]STRIN[1]STRIN[2]] | |
316 | [...] | |
317 | ----------------------------- | |
318 | if (!rrr) | |
319 | write (CC..C) | |
320 | else | |
321 | write_string (STRING, CC..C); | |
322 | IC += (CC..C + 2) / 3; | |
323 | */ | |
324 | ||
325 | #define CCL_WriteArray 0x15 /* Write an element of array: | |
326 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
327 | 2:ELEMENT[0] | |
328 | 3:ELEMENT[1] | |
329 | ... | |
330 | ------------------------------ | |
331 | if (0 <= reg[rrr] < CC..C) | |
332 | write (ELEMENT[reg[rrr]]); | |
333 | IC += CC..C; | |
334 | */ | |
335 | ||
336 | #define CCL_End 0x16 /* Terminate: | |
337 | 1:00000000000000000000000XXXXX | |
338 | ------------------------------ | |
339 | terminate (); | |
340 | */ | |
341 | ||
342 | /* The following two codes execute an assignment arithmetic/logical | |
343 | operation. The form of the operation is like REG OP= OPERAND. */ | |
344 | ||
345 | #define CCL_ExprSelfConst 0x17 /* REG OP= constant: | |
346 | 1:00000OPERATION000000rrrXXXXX | |
347 | 2:CONSTANT | |
348 | ------------------------------ | |
349 | reg[rrr] OPERATION= CONSTANT; | |
350 | */ | |
351 | ||
352 | #define CCL_ExprSelfReg 0x18 /* REG1 OP= REG2: | |
353 | 1:00000OPERATION000RRRrrrXXXXX | |
354 | ------------------------------ | |
355 | reg[rrr] OPERATION= reg[RRR]; | |
356 | */ | |
357 | ||
358 | /* The following codes execute an arithmetic/logical operation. The | |
359 | form of the operation is like REG_X = REG_Y OP OPERAND2. */ | |
360 | ||
361 | #define CCL_SetExprConst 0x19 /* REG_X = REG_Y OP constant: | |
362 | 1:00000OPERATION000RRRrrrXXXXX | |
363 | 2:CONSTANT | |
364 | ------------------------------ | |
365 | reg[rrr] = reg[RRR] OPERATION CONSTANT; | |
366 | IC++; | |
367 | */ | |
368 | ||
369 | #define CCL_SetExprReg 0x1A /* REG1 = REG2 OP REG3: | |
370 | 1:00000OPERATIONRrrRRRrrrXXXXX | |
371 | ------------------------------ | |
372 | reg[rrr] = reg[RRR] OPERATION reg[Rrr]; | |
373 | */ | |
374 | ||
375 | #define CCL_JumpCondExprConst 0x1B /* Jump conditional according to | |
376 | an operation on constant: | |
377 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
378 | 2:OPERATION | |
379 | 3:CONSTANT | |
380 | ----------------------------- | |
381 | reg[7] = reg[rrr] OPERATION CONSTANT; | |
382 | if (!(reg[7])) | |
383 | IC += ADDRESS; | |
384 | else | |
385 | IC += 2 | |
386 | */ | |
387 | ||
388 | #define CCL_JumpCondExprReg 0x1C /* Jump conditional according to | |
389 | an operation on register: | |
390 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
391 | 2:OPERATION | |
392 | 3:RRR | |
393 | ----------------------------- | |
394 | reg[7] = reg[rrr] OPERATION reg[RRR]; | |
395 | if (!reg[7]) | |
396 | IC += ADDRESS; | |
397 | else | |
398 | IC += 2; | |
399 | */ | |
400 | ||
401 | #define CCL_ReadJumpCondExprConst 0x1D /* Read and jump conditional according | |
402 | to an operation on constant: | |
403 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
404 | 2:OPERATION | |
405 | 3:CONSTANT | |
406 | ----------------------------- | |
407 | read (reg[rrr]); | |
408 | reg[7] = reg[rrr] OPERATION CONSTANT; | |
409 | if (!reg[7]) | |
410 | IC += ADDRESS; | |
411 | else | |
412 | IC += 2; | |
413 | */ | |
414 | ||
415 | #define CCL_ReadJumpCondExprReg 0x1E /* Read and jump conditional according | |
416 | to an operation on register: | |
417 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
418 | 2:OPERATION | |
419 | 3:RRR | |
420 | ----------------------------- | |
421 | read (reg[rrr]); | |
422 | reg[7] = reg[rrr] OPERATION reg[RRR]; | |
423 | if (!reg[7]) | |
424 | IC += ADDRESS; | |
425 | else | |
426 | IC += 2; | |
427 | */ | |
428 | ||
450ed226 | 429 | #define CCL_Extension 0x1F /* Extended CCL code |
4ed46869 KH |
430 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX |
431 | 2:ARGUEMENT | |
432 | 3:... | |
433 | ------------------------------ | |
434 | extended_command (rrr,RRR,Rrr,ARGS) | |
435 | */ | |
436 | ||
177c0ea7 | 437 | /* |
6ae21908 | 438 | Here after, Extended CCL Instructions. |
e34b1164 | 439 | Bit length of extended command is 14. |
6ae21908 | 440 | Therefore, the instruction code range is 0..16384(0x3fff). |
e34b1164 KH |
441 | */ |
442 | ||
6ae21908 KH |
443 | /* Read a multibyte characeter. |
444 | A code point is stored into reg[rrr]. A charset ID is stored into | |
445 | reg[RRR]. */ | |
446 | ||
447 | #define CCL_ReadMultibyteChar2 0x00 /* Read Multibyte Character | |
448 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX */ | |
449 | ||
450 | /* Write a multibyte character. | |
451 | Write a character whose code point is reg[rrr] and the charset ID | |
452 | is reg[RRR]. */ | |
453 | ||
454 | #define CCL_WriteMultibyteChar2 0x01 /* Write Multibyte Character | |
455 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX */ | |
456 | ||
8146262a | 457 | /* Translate a character whose code point is reg[rrr] and the charset |
f967223b | 458 | ID is reg[RRR] by a translation table whose ID is reg[Rrr]. |
6ae21908 | 459 | |
8146262a | 460 | A translated character is set in reg[rrr] (code point) and reg[RRR] |
6ae21908 KH |
461 | (charset ID). */ |
462 | ||
8146262a | 463 | #define CCL_TranslateCharacter 0x02 /* Translate a multibyte character |
6ae21908 KH |
464 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX */ |
465 | ||
8146262a | 466 | /* Translate a character whose code point is reg[rrr] and the charset |
f967223b | 467 | ID is reg[RRR] by a translation table whose ID is ARGUMENT. |
6ae21908 | 468 | |
8146262a | 469 | A translated character is set in reg[rrr] (code point) and reg[RRR] |
6ae21908 KH |
470 | (charset ID). */ |
471 | ||
8146262a KH |
472 | #define CCL_TranslateCharacterConstTbl 0x03 /* Translate a multibyte character |
473 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX | |
474 | 2:ARGUMENT(Translation Table ID) | |
475 | */ | |
6ae21908 | 476 | |
8146262a KH |
477 | /* Iterate looking up MAPs for reg[rrr] starting from the Nth (N = |
478 | reg[RRR]) MAP until some value is found. | |
6ae21908 | 479 | |
8146262a | 480 | Each MAP is a Lisp vector whose element is number, nil, t, or |
6ae21908 | 481 | lambda. |
8146262a | 482 | If the element is nil, ignore the map and proceed to the next map. |
6ae21908 KH |
483 | If the element is t or lambda, finish without changing reg[rrr]. |
484 | If the element is a number, set reg[rrr] to the number and finish. | |
485 | ||
8146262a KH |
486 | Detail of the map structure is descibed in the comment for |
487 | CCL_MapMultiple below. */ | |
6ae21908 | 488 | |
8146262a | 489 | #define CCL_IterateMultipleMap 0x10 /* Iterate multiple maps |
6ae21908 | 490 | 1:ExtendedCOMMNDXXXRRRrrrXXXXX |
8146262a KH |
491 | 2:NUMBER of MAPs |
492 | 3:MAP-ID1 | |
493 | 4:MAP-ID2 | |
6ae21908 | 494 | ... |
177c0ea7 | 495 | */ |
6ae21908 | 496 | |
8146262a KH |
497 | /* Map the code in reg[rrr] by MAPs starting from the Nth (N = |
498 | reg[RRR]) map. | |
6ae21908 | 499 | |
9b27b20d | 500 | MAPs are supplied in the succeeding CCL codes as follows: |
6ae21908 | 501 | |
8146262a KH |
502 | When CCL program gives this nested structure of map to this command: |
503 | ((MAP-ID11 | |
504 | MAP-ID12 | |
505 | (MAP-ID121 MAP-ID122 MAP-ID123) | |
506 | MAP-ID13) | |
507 | (MAP-ID21 | |
508 | (MAP-ID211 (MAP-ID2111) MAP-ID212) | |
509 | MAP-ID22)), | |
6ae21908 | 510 | the compiled CCL codes has this sequence: |
8146262a | 511 | CCL_MapMultiple (CCL code of this command) |
9b27b20d KH |
512 | 16 (total number of MAPs and SEPARATORs) |
513 | -7 (1st SEPARATOR) | |
8146262a KH |
514 | MAP-ID11 |
515 | MAP-ID12 | |
9b27b20d | 516 | -3 (2nd SEPARATOR) |
8146262a KH |
517 | MAP-ID121 |
518 | MAP-ID122 | |
519 | MAP-ID123 | |
520 | MAP-ID13 | |
9b27b20d | 521 | -7 (3rd SEPARATOR) |
8146262a | 522 | MAP-ID21 |
9b27b20d | 523 | -4 (4th SEPARATOR) |
8146262a | 524 | MAP-ID211 |
9b27b20d | 525 | -1 (5th SEPARATOR) |
8146262a KH |
526 | MAP_ID2111 |
527 | MAP-ID212 | |
528 | MAP-ID22 | |
6ae21908 | 529 | |
9b27b20d | 530 | A value of each SEPARATOR follows this rule: |
8146262a KH |
531 | MAP-SET := SEPARATOR [(MAP-ID | MAP-SET)]+ |
532 | SEPARATOR := -(number of MAP-IDs and SEPARATORs in the MAP-SET) | |
6ae21908 | 533 | |
8146262a | 534 | (*)....Nest level of MAP-SET must not be over than MAX_MAP_SET_LEVEL. |
6ae21908 | 535 | |
8146262a KH |
536 | When some map fails to map (i.e. it doesn't have a value for |
537 | reg[rrr]), the mapping is treated as identity. | |
6ae21908 | 538 | |
8146262a | 539 | The mapping is iterated for all maps in each map set (set of maps |
9b27b20d KH |
540 | separated by SEPARATOR) except in the case that lambda is |
541 | encountered. More precisely, the mapping proceeds as below: | |
542 | ||
543 | At first, VAL0 is set to reg[rrr], and it is translated by the | |
544 | first map to VAL1. Then, VAL1 is translated by the next map to | |
545 | VAL2. This mapping is iterated until the last map is used. The | |
54fa5bc1 KH |
546 | result of the mapping is the last value of VAL?. When the mapping |
547 | process reached to the end of the map set, it moves to the next | |
548 | map set. If the next does not exit, the mapping process terminates, | |
549 | and regard the last value as a result. | |
9b27b20d KH |
550 | |
551 | But, when VALm is mapped to VALn and VALn is not a number, the | |
552 | mapping proceed as below: | |
553 | ||
554 | If VALn is nil, the lastest map is ignored and the mapping of VALm | |
555 | proceed to the next map. | |
556 | ||
557 | In VALn is t, VALm is reverted to reg[rrr] and the mapping of VALm | |
558 | proceed to the next map. | |
559 | ||
54fa5bc1 KH |
560 | If VALn is lambda, move to the next map set like reaching to the |
561 | end of the current map set. | |
562 | ||
563 | If VALn is a symbol, call the CCL program refered by it. | |
564 | Then, use reg[rrr] as a mapped value except for -1, -2 and -3. | |
565 | Such special values are regarded as nil, t, and lambda respectively. | |
6ae21908 | 566 | |
8146262a | 567 | Each map is a Lisp vector of the following format (a) or (b): |
6ae21908 KH |
568 | (a)......[STARTPOINT VAL1 VAL2 ...] |
569 | (b)......[t VAL STARTPOINT ENDPOINT], | |
570 | where | |
8146262a | 571 | STARTPOINT is an offset to be used for indexing a map, |
9b27b20d | 572 | ENDPOINT is a maximum index number of a map, |
177c0ea7 | 573 | VAL and VALn is a number, nil, t, or lambda. |
6ae21908 | 574 | |
8146262a KH |
575 | Valid index range of a map of type (a) is: |
576 | STARTPOINT <= index < STARTPOINT + map_size - 1 | |
577 | Valid index range of a map of type (b) is: | |
9b27b20d | 578 | STARTPOINT <= index < ENDPOINT */ |
6ae21908 | 579 | |
8146262a | 580 | #define CCL_MapMultiple 0x11 /* Mapping by multiple code conversion maps |
6ae21908 KH |
581 | 1:ExtendedCOMMNDXXXRRRrrrXXXXX |
582 | 2:N-2 | |
583 | 3:SEPARATOR_1 (< 0) | |
8146262a KH |
584 | 4:MAP-ID_1 |
585 | 5:MAP-ID_2 | |
6ae21908 KH |
586 | ... |
587 | M:SEPARATOR_x (< 0) | |
8146262a | 588 | M+1:MAP-ID_y |
6ae21908 KH |
589 | ... |
590 | N:SEPARATOR_z (< 0) | |
591 | */ | |
592 | ||
54fa5bc1 | 593 | #define MAX_MAP_SET_LEVEL 30 |
6ae21908 KH |
594 | |
595 | typedef struct | |
596 | { | |
597 | int rest_length; | |
598 | int orig_val; | |
599 | } tr_stack; | |
600 | ||
8146262a KH |
601 | static tr_stack mapping_stack[MAX_MAP_SET_LEVEL]; |
602 | static tr_stack *mapping_stack_pointer; | |
6ae21908 | 603 | |
54fa5bc1 KH |
604 | /* If this variable is non-zero, it indicates the stack_idx |
605 | of immediately called by CCL_MapMultiple. */ | |
be57900b | 606 | static int stack_idx_of_map_multiple; |
54fa5bc1 KH |
607 | |
608 | #define PUSH_MAPPING_STACK(restlen, orig) \ | |
a89f435d PJ |
609 | do \ |
610 | { \ | |
54fa5bc1 KH |
611 | mapping_stack_pointer->rest_length = (restlen); \ |
612 | mapping_stack_pointer->orig_val = (orig); \ | |
613 | mapping_stack_pointer++; \ | |
a89f435d PJ |
614 | } \ |
615 | while (0) | |
54fa5bc1 KH |
616 | |
617 | #define POP_MAPPING_STACK(restlen, orig) \ | |
a89f435d PJ |
618 | do \ |
619 | { \ | |
54fa5bc1 KH |
620 | mapping_stack_pointer--; \ |
621 | (restlen) = mapping_stack_pointer->rest_length; \ | |
622 | (orig) = mapping_stack_pointer->orig_val; \ | |
a89f435d PJ |
623 | } \ |
624 | while (0) | |
6ae21908 | 625 | |
54fa5bc1 | 626 | #define CCL_CALL_FOR_MAP_INSTRUCTION(symbol, ret_ic) \ |
a89f435d | 627 | do \ |
0ee1088b | 628 | { \ |
54fa5bc1 KH |
629 | struct ccl_program called_ccl; \ |
630 | if (stack_idx >= 256 \ | |
631 | || (setup_ccl_program (&called_ccl, (symbol)) != 0)) \ | |
632 | { \ | |
633 | if (stack_idx > 0) \ | |
634 | { \ | |
635 | ccl_prog = ccl_prog_stack_struct[0].ccl_prog; \ | |
636 | ic = ccl_prog_stack_struct[0].ic; \ | |
637 | } \ | |
638 | CCL_INVALID_CMD; \ | |
639 | } \ | |
640 | ccl_prog_stack_struct[stack_idx].ccl_prog = ccl_prog; \ | |
641 | ccl_prog_stack_struct[stack_idx].ic = (ret_ic); \ | |
642 | stack_idx++; \ | |
643 | ccl_prog = called_ccl.prog; \ | |
644 | ic = CCL_HEADER_MAIN; \ | |
645 | goto ccl_repeat; \ | |
0ee1088b | 646 | } \ |
a89f435d | 647 | while (0) |
6ae21908 | 648 | |
8146262a | 649 | #define CCL_MapSingle 0x12 /* Map by single code conversion map |
6ae21908 | 650 | 1:ExtendedCOMMNDXXXRRRrrrXXXXX |
8146262a | 651 | 2:MAP-ID |
6ae21908 | 652 | ------------------------------ |
8146262a KH |
653 | Map reg[rrr] by MAP-ID. |
654 | If some valid mapping is found, | |
6ae21908 KH |
655 | set reg[rrr] to the result, |
656 | else | |
657 | set reg[RRR] to -1. | |
658 | */ | |
4ed46869 | 659 | |
d80dc57e DL |
660 | #define CCL_LookupIntConstTbl 0x13 /* Lookup multibyte character by |
661 | integer key. Afterwards R7 set | |
662 | to 1 iff lookup succeeded. | |
663 | 1:ExtendedCOMMNDRrrRRRXXXXXXXX | |
664 | 2:ARGUMENT(Hash table ID) */ | |
665 | ||
666 | #define CCL_LookupCharConstTbl 0x14 /* Lookup integer by multibyte | |
667 | character key. Afterwards R7 set | |
668 | to 1 iff lookup succeeded. | |
669 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX | |
670 | 2:ARGUMENT(Hash table ID) */ | |
671 | ||
4ed46869 KH |
672 | /* CCL arithmetic/logical operators. */ |
673 | #define CCL_PLUS 0x00 /* X = Y + Z */ | |
674 | #define CCL_MINUS 0x01 /* X = Y - Z */ | |
675 | #define CCL_MUL 0x02 /* X = Y * Z */ | |
676 | #define CCL_DIV 0x03 /* X = Y / Z */ | |
677 | #define CCL_MOD 0x04 /* X = Y % Z */ | |
678 | #define CCL_AND 0x05 /* X = Y & Z */ | |
679 | #define CCL_OR 0x06 /* X = Y | Z */ | |
680 | #define CCL_XOR 0x07 /* X = Y ^ Z */ | |
681 | #define CCL_LSH 0x08 /* X = Y << Z */ | |
682 | #define CCL_RSH 0x09 /* X = Y >> Z */ | |
683 | #define CCL_LSH8 0x0A /* X = (Y << 8) | Z */ | |
684 | #define CCL_RSH8 0x0B /* X = Y >> 8, r[7] = Y & 0xFF */ | |
685 | #define CCL_DIVMOD 0x0C /* X = Y / Z, r[7] = Y % Z */ | |
686 | #define CCL_LS 0x10 /* X = (X < Y) */ | |
687 | #define CCL_GT 0x11 /* X = (X > Y) */ | |
688 | #define CCL_EQ 0x12 /* X = (X == Y) */ | |
689 | #define CCL_LE 0x13 /* X = (X <= Y) */ | |
690 | #define CCL_GE 0x14 /* X = (X >= Y) */ | |
691 | #define CCL_NE 0x15 /* X = (X != Y) */ | |
692 | ||
51520e8a | 693 | #define CCL_DECODE_SJIS 0x16 /* X = HIGHER_BYTE (DE-SJIS (Y, Z)) |
4ed46869 | 694 | r[7] = LOWER_BYTE (DE-SJIS (Y, Z)) */ |
51520e8a KH |
695 | #define CCL_ENCODE_SJIS 0x17 /* X = HIGHER_BYTE (SJIS (Y, Z)) |
696 | r[7] = LOWER_BYTE (SJIS (Y, Z) */ | |
4ed46869 | 697 | |
4ed46869 | 698 | /* Terminate CCL program successfully. */ |
0ee1088b | 699 | #define CCL_SUCCESS \ |
a89f435d | 700 | do \ |
0ee1088b | 701 | { \ |
4ed46869 | 702 | ccl->status = CCL_STAT_SUCCESS; \ |
0ee1088b KH |
703 | goto ccl_finish; \ |
704 | } \ | |
a89f435d | 705 | while(0) |
4ed46869 KH |
706 | |
707 | /* Suspend CCL program because of reading from empty input buffer or | |
708 | writing to full output buffer. When this program is resumed, the | |
709 | same I/O command is executed. */ | |
e34b1164 | 710 | #define CCL_SUSPEND(stat) \ |
a89f435d | 711 | do \ |
0ee1088b | 712 | { \ |
e34b1164 KH |
713 | ic--; \ |
714 | ccl->status = stat; \ | |
715 | goto ccl_finish; \ | |
0ee1088b | 716 | } \ |
a89f435d | 717 | while (0) |
4ed46869 KH |
718 | |
719 | /* Terminate CCL program because of invalid command. Should not occur | |
720 | in the normal case. */ | |
721 | #define CCL_INVALID_CMD \ | |
a89f435d | 722 | do \ |
0ee1088b | 723 | { \ |
4ed46869 KH |
724 | ccl->status = CCL_STAT_INVALID_CMD; \ |
725 | goto ccl_error_handler; \ | |
0ee1088b | 726 | } \ |
a89f435d | 727 | while(0) |
4ed46869 KH |
728 | |
729 | /* Encode one character CH to multibyte form and write to the current | |
887bfbd7 | 730 | output buffer. If CH is less than 256, CH is written as is. */ |
c10842ea KH |
731 | #define CCL_WRITE_CHAR(ch) \ |
732 | do { \ | |
733 | if (! dst) \ | |
734 | CCL_INVALID_CMD; \ | |
735 | else if (dst < dst_end) \ | |
736 | *dst++ = (ch); \ | |
737 | else \ | |
738 | CCL_SUSPEND (CCL_STAT_SUSPEND_BY_DST); \ | |
a8302ba3 KH |
739 | } while (0) |
740 | ||
4ed46869 KH |
741 | /* Write a string at ccl_prog[IC] of length LEN to the current output |
742 | buffer. */ | |
743 | #define CCL_WRITE_STRING(len) \ | |
744 | do { \ | |
c10842ea | 745 | int i; \ |
4ed46869 KH |
746 | if (!dst) \ |
747 | CCL_INVALID_CMD; \ | |
c10842ea | 748 | else if (dst + len <= dst_end) \ |
4ed46869 KH |
749 | for (i = 0; i < len; i++) \ |
750 | *dst++ = ((XFASTINT (ccl_prog[ic + (i / 3)])) \ | |
751 | >> ((2 - (i % 3)) * 8)) & 0xFF; \ | |
752 | else \ | |
e34b1164 | 753 | CCL_SUSPEND (CCL_STAT_SUSPEND_BY_DST); \ |
4ed46869 KH |
754 | } while (0) |
755 | ||
c10842ea KH |
756 | /* Read one byte from the current input buffer into Rth register. */ |
757 | #define CCL_READ_CHAR(r) \ | |
758 | do { \ | |
759 | if (! src) \ | |
760 | CCL_INVALID_CMD; \ | |
761 | else if (src < src_end) \ | |
762 | r = *src++; \ | |
763 | else if (ccl->last_block) \ | |
764 | { \ | |
765 | ic = ccl->eof_ic; \ | |
766 | goto ccl_repeat; \ | |
767 | } \ | |
768 | else \ | |
769 | CCL_SUSPEND (CCL_STAT_SUSPEND_BY_SRC); \ | |
770 | } while (0) | |
771 | ||
bda731af KH |
772 | /* Decode CODE by a charset whose id is ID. If ID is 0, return CODE |
773 | as is for backward compatibility. Assume that we can use the | |
774 | variable `charset'. */ | |
775 | ||
776 | #define CCL_DECODE_CHAR(id, code) \ | |
777 | ((id) == 0 ? (code) \ | |
778 | : (charset = CHARSET_FROM_ID ((id)), DECODE_CHAR (charset, (code)))) | |
779 | ||
bda731af KH |
780 | /* Encode character C by some of charsets in CHARSET_LIST. Set ID to |
781 | the id of the used charset, ENCODED to the resulf of encoding. | |
782 | Assume that we can use the variable `charset'. */ | |
783 | ||
8f924df7 | 784 | #define CCL_ENCODE_CHAR(c, charset_list, id, encoded) \ |
4ffd4870 | 785 | do { \ |
8f924df7 | 786 | unsigned code; \ |
4ffd4870 | 787 | \ |
8f924df7 KH |
788 | charset = char_charset ((c), (charset_list), &code); \ |
789 | if (! charset && ! NILP (charset_list)) \ | |
790 | charset = char_charset ((c), Qnil, &code); \ | |
791 | if (charset) \ | |
792 | { \ | |
793 | (id) = CHARSET_ID (charset); \ | |
794 | (encoded) = code; \ | |
4ffd4870 | 795 | } \ |
8f924df7 | 796 | } while (0) |
4ffd4870 | 797 | |
c10842ea KH |
798 | /* Execute CCL code on characters at SOURCE (length SRC_SIZE). The |
799 | resulting text goes to a place pointed by DESTINATION, the length | |
800 | of which should not exceed DST_SIZE. As a side effect, how many | |
801 | characters are consumed and produced are recorded in CCL->consumed | |
802 | and CCL->produced, and the contents of CCL registers are updated. | |
803 | If SOURCE or DESTINATION is NULL, only operations on registers are | |
804 | permitted. */ | |
4ed46869 KH |
805 | |
806 | #ifdef CCL_DEBUG | |
807 | #define CCL_DEBUG_BACKTRACE_LEN 256 | |
f9bd23fd | 808 | int ccl_backtrace_table[CCL_DEBUG_BACKTRACE_LEN]; |
4ed46869 KH |
809 | int ccl_backtrace_idx; |
810 | #endif | |
811 | ||
812 | struct ccl_prog_stack | |
813 | { | |
a9f1cc19 | 814 | Lisp_Object *ccl_prog; /* Pointer to an array of CCL code. */ |
4ed46869 KH |
815 | int ic; /* Instruction Counter. */ |
816 | }; | |
817 | ||
177c0ea7 | 818 | /* For the moment, we only support depth 256 of stack. */ |
c13362d8 KH |
819 | static struct ccl_prog_stack ccl_prog_stack_struct[256]; |
820 | ||
c10842ea | 821 | void |
bda731af | 822 | ccl_driver (ccl, source, destination, src_size, dst_size, charset_list) |
4ed46869 | 823 | struct ccl_program *ccl; |
c10842ea KH |
824 | int *source, *destination; |
825 | int src_size, dst_size; | |
bda731af | 826 | Lisp_Object charset_list; |
4ed46869 KH |
827 | { |
828 | register int *reg = ccl->reg; | |
829 | register int ic = ccl->ic; | |
8a1ae4dd | 830 | register int code = 0, field1, field2; |
e995085f | 831 | register Lisp_Object *ccl_prog = ccl->prog; |
c10842ea KH |
832 | int *src = source, *src_end = src + src_size; |
833 | int *dst = destination, *dst_end = dst + dst_size; | |
4ed46869 | 834 | int jump_address; |
8a1ae4dd | 835 | int i = 0, j, op; |
c13362d8 | 836 | int stack_idx = ccl->stack_idx; |
519bf146 | 837 | /* Instruction counter of the current CCL code. */ |
8a1ae4dd | 838 | int this_ic = 0; |
c10842ea | 839 | struct charset *charset; |
4ed46869 KH |
840 | |
841 | if (ic >= ccl->eof_ic) | |
842 | ic = CCL_HEADER_MAIN; | |
843 | ||
c10842ea | 844 | if (ccl->buf_magnification == 0) /* We can't read/produce any bytes. */ |
12abd7d1 KH |
845 | dst = NULL; |
846 | ||
54fa5bc1 KH |
847 | /* Set mapping stack pointer. */ |
848 | mapping_stack_pointer = mapping_stack; | |
849 | ||
4ed46869 KH |
850 | #ifdef CCL_DEBUG |
851 | ccl_backtrace_idx = 0; | |
852 | #endif | |
853 | ||
854 | for (;;) | |
855 | { | |
4ccd0d4a | 856 | ccl_repeat: |
4ed46869 KH |
857 | #ifdef CCL_DEBUG |
858 | ccl_backtrace_table[ccl_backtrace_idx++] = ic; | |
859 | if (ccl_backtrace_idx >= CCL_DEBUG_BACKTRACE_LEN) | |
860 | ccl_backtrace_idx = 0; | |
861 | ccl_backtrace_table[ccl_backtrace_idx] = 0; | |
862 | #endif | |
863 | ||
864 | if (!NILP (Vquit_flag) && NILP (Vinhibit_quit)) | |
865 | { | |
866 | /* We can't just signal Qquit, instead break the loop as if | |
867 | the whole data is processed. Don't reset Vquit_flag, it | |
868 | must be handled later at a safer place. */ | |
c10842ea KH |
869 | if (src) |
870 | src = source + src_size; | |
4ed46869 KH |
871 | ccl->status = CCL_STAT_QUIT; |
872 | break; | |
873 | } | |
874 | ||
519bf146 | 875 | this_ic = ic; |
4ed46869 KH |
876 | code = XINT (ccl_prog[ic]); ic++; |
877 | field1 = code >> 8; | |
878 | field2 = (code & 0xFF) >> 5; | |
879 | ||
880 | #define rrr field2 | |
881 | #define RRR (field1 & 7) | |
882 | #define Rrr ((field1 >> 3) & 7) | |
883 | #define ADDR field1 | |
e34b1164 | 884 | #define EXCMD (field1 >> 6) |
4ed46869 KH |
885 | |
886 | switch (code & 0x1F) | |
887 | { | |
888 | case CCL_SetRegister: /* 00000000000000000RRRrrrXXXXX */ | |
889 | reg[rrr] = reg[RRR]; | |
890 | break; | |
891 | ||
892 | case CCL_SetShortConst: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
893 | reg[rrr] = field1; | |
894 | break; | |
895 | ||
896 | case CCL_SetConst: /* 00000000000000000000rrrXXXXX */ | |
897 | reg[rrr] = XINT (ccl_prog[ic]); | |
898 | ic++; | |
899 | break; | |
900 | ||
901 | case CCL_SetArray: /* CCCCCCCCCCCCCCCCCCCCRRRrrrXXXXX */ | |
902 | i = reg[RRR]; | |
903 | j = field1 >> 3; | |
904 | if ((unsigned int) i < j) | |
905 | reg[rrr] = XINT (ccl_prog[ic + i]); | |
906 | ic += j; | |
907 | break; | |
908 | ||
909 | case CCL_Jump: /* A--D--D--R--E--S--S-000XXXXX */ | |
910 | ic += ADDR; | |
911 | break; | |
912 | ||
913 | case CCL_JumpCond: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
914 | if (!reg[rrr]) | |
915 | ic += ADDR; | |
916 | break; | |
917 | ||
918 | case CCL_WriteRegisterJump: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
919 | i = reg[rrr]; | |
920 | CCL_WRITE_CHAR (i); | |
921 | ic += ADDR; | |
922 | break; | |
923 | ||
924 | case CCL_WriteRegisterReadJump: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
925 | i = reg[rrr]; | |
926 | CCL_WRITE_CHAR (i); | |
927 | ic++; | |
928 | CCL_READ_CHAR (reg[rrr]); | |
929 | ic += ADDR - 1; | |
930 | break; | |
931 | ||
932 | case CCL_WriteConstJump: /* A--D--D--R--E--S--S-000XXXXX */ | |
933 | i = XINT (ccl_prog[ic]); | |
934 | CCL_WRITE_CHAR (i); | |
935 | ic += ADDR; | |
936 | break; | |
937 | ||
938 | case CCL_WriteConstReadJump: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
939 | i = XINT (ccl_prog[ic]); | |
940 | CCL_WRITE_CHAR (i); | |
941 | ic++; | |
942 | CCL_READ_CHAR (reg[rrr]); | |
943 | ic += ADDR - 1; | |
944 | break; | |
945 | ||
946 | case CCL_WriteStringJump: /* A--D--D--R--E--S--S-000XXXXX */ | |
947 | j = XINT (ccl_prog[ic]); | |
948 | ic++; | |
949 | CCL_WRITE_STRING (j); | |
950 | ic += ADDR - 1; | |
951 | break; | |
952 | ||
953 | case CCL_WriteArrayReadJump: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
954 | i = reg[rrr]; | |
2e34157c | 955 | j = XINT (ccl_prog[ic]); |
4ed46869 KH |
956 | if ((unsigned int) i < j) |
957 | { | |
887bfbd7 | 958 | i = XINT (ccl_prog[ic + 1 + i]); |
4ed46869 KH |
959 | CCL_WRITE_CHAR (i); |
960 | } | |
887bfbd7 | 961 | ic += j + 2; |
4ed46869 KH |
962 | CCL_READ_CHAR (reg[rrr]); |
963 | ic += ADDR - (j + 2); | |
964 | break; | |
965 | ||
966 | case CCL_ReadJump: /* A--D--D--R--E--S--S-rrrYYYYY */ | |
967 | CCL_READ_CHAR (reg[rrr]); | |
968 | ic += ADDR; | |
969 | break; | |
970 | ||
971 | case CCL_ReadBranch: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
972 | CCL_READ_CHAR (reg[rrr]); | |
973 | /* fall through ... */ | |
974 | case CCL_Branch: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
975 | if ((unsigned int) reg[rrr] < field1) | |
976 | ic += XINT (ccl_prog[ic + reg[rrr]]); | |
977 | else | |
978 | ic += XINT (ccl_prog[ic + field1]); | |
979 | break; | |
980 | ||
981 | case CCL_ReadRegister: /* CCCCCCCCCCCCCCCCCCCCrrXXXXX */ | |
982 | while (1) | |
983 | { | |
984 | CCL_READ_CHAR (reg[rrr]); | |
985 | if (!field1) break; | |
986 | code = XINT (ccl_prog[ic]); ic++; | |
987 | field1 = code >> 8; | |
988 | field2 = (code & 0xFF) >> 5; | |
989 | } | |
990 | break; | |
991 | ||
992 | case CCL_WriteExprConst: /* 1:00000OPERATION000RRR000XXXXX */ | |
993 | rrr = 7; | |
994 | i = reg[RRR]; | |
995 | j = XINT (ccl_prog[ic]); | |
996 | op = field1 >> 6; | |
25660570 | 997 | jump_address = ic + 1; |
4ed46869 KH |
998 | goto ccl_set_expr; |
999 | ||
1000 | case CCL_WriteRegister: /* CCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
1001 | while (1) | |
1002 | { | |
1003 | i = reg[rrr]; | |
1004 | CCL_WRITE_CHAR (i); | |
1005 | if (!field1) break; | |
1006 | code = XINT (ccl_prog[ic]); ic++; | |
1007 | field1 = code >> 8; | |
1008 | field2 = (code & 0xFF) >> 5; | |
1009 | } | |
1010 | break; | |
1011 | ||
1012 | case CCL_WriteExprRegister: /* 1:00000OPERATIONRrrRRR000XXXXX */ | |
1013 | rrr = 7; | |
1014 | i = reg[RRR]; | |
1015 | j = reg[Rrr]; | |
1016 | op = field1 >> 6; | |
25660570 | 1017 | jump_address = ic; |
4ed46869 KH |
1018 | goto ccl_set_expr; |
1019 | ||
5232fa7b | 1020 | case CCL_Call: /* 1:CCCCCCCCCCCCCCCCCCCCFFFXXXXX */ |
4ed46869 KH |
1021 | { |
1022 | Lisp_Object slot; | |
5232fa7b KH |
1023 | int prog_id; |
1024 | ||
1025 | /* If FFF is nonzero, the CCL program ID is in the | |
1026 | following code. */ | |
1027 | if (rrr) | |
1028 | { | |
1029 | prog_id = XINT (ccl_prog[ic]); | |
1030 | ic++; | |
1031 | } | |
1032 | else | |
1033 | prog_id = field1; | |
4ed46869 KH |
1034 | |
1035 | if (stack_idx >= 256 | |
5232fa7b | 1036 | || prog_id < 0 |
64ef2921 SM |
1037 | || prog_id >= ASIZE (Vccl_program_table) |
1038 | || (slot = AREF (Vccl_program_table, prog_id), !VECTORP (slot)) | |
1039 | || !VECTORP (AREF (slot, 1))) | |
4ed46869 KH |
1040 | { |
1041 | if (stack_idx > 0) | |
1042 | { | |
1043 | ccl_prog = ccl_prog_stack_struct[0].ccl_prog; | |
1044 | ic = ccl_prog_stack_struct[0].ic; | |
1045 | } | |
1046 | CCL_INVALID_CMD; | |
1047 | } | |
177c0ea7 | 1048 | |
4ed46869 KH |
1049 | ccl_prog_stack_struct[stack_idx].ccl_prog = ccl_prog; |
1050 | ccl_prog_stack_struct[stack_idx].ic = ic; | |
1051 | stack_idx++; | |
64ef2921 | 1052 | ccl_prog = XVECTOR (AREF (slot, 1))->contents; |
4ed46869 KH |
1053 | ic = CCL_HEADER_MAIN; |
1054 | } | |
1055 | break; | |
1056 | ||
1057 | case CCL_WriteConstString: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
1058 | if (!rrr) | |
1059 | CCL_WRITE_CHAR (field1); | |
1060 | else | |
1061 | { | |
1062 | CCL_WRITE_STRING (field1); | |
1063 | ic += (field1 + 2) / 3; | |
1064 | } | |
1065 | break; | |
1066 | ||
1067 | case CCL_WriteArray: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
1068 | i = reg[rrr]; | |
1069 | if ((unsigned int) i < field1) | |
1070 | { | |
1071 | j = XINT (ccl_prog[ic + i]); | |
1072 | CCL_WRITE_CHAR (j); | |
1073 | } | |
1074 | ic += field1; | |
1075 | break; | |
1076 | ||
1077 | case CCL_End: /* 0000000000000000000000XXXXX */ | |
d3a478e2 | 1078 | if (stack_idx > 0) |
4ed46869 | 1079 | { |
d3a478e2 | 1080 | stack_idx--; |
4ed46869 KH |
1081 | ccl_prog = ccl_prog_stack_struct[stack_idx].ccl_prog; |
1082 | ic = ccl_prog_stack_struct[stack_idx].ic; | |
1083 | break; | |
1084 | } | |
ad3d1b1d KH |
1085 | if (src) |
1086 | src = src_end; | |
1087 | /* ccl->ic should points to this command code again to | |
1088 | suppress further processing. */ | |
1089 | ic--; | |
4ed46869 KH |
1090 | CCL_SUCCESS; |
1091 | ||
1092 | case CCL_ExprSelfConst: /* 00000OPERATION000000rrrXXXXX */ | |
1093 | i = XINT (ccl_prog[ic]); | |
1094 | ic++; | |
1095 | op = field1 >> 6; | |
1096 | goto ccl_expr_self; | |
1097 | ||
1098 | case CCL_ExprSelfReg: /* 00000OPERATION000RRRrrrXXXXX */ | |
1099 | i = reg[RRR]; | |
1100 | op = field1 >> 6; | |
1101 | ||
1102 | ccl_expr_self: | |
1103 | switch (op) | |
1104 | { | |
1105 | case CCL_PLUS: reg[rrr] += i; break; | |
1106 | case CCL_MINUS: reg[rrr] -= i; break; | |
1107 | case CCL_MUL: reg[rrr] *= i; break; | |
1108 | case CCL_DIV: reg[rrr] /= i; break; | |
1109 | case CCL_MOD: reg[rrr] %= i; break; | |
1110 | case CCL_AND: reg[rrr] &= i; break; | |
1111 | case CCL_OR: reg[rrr] |= i; break; | |
1112 | case CCL_XOR: reg[rrr] ^= i; break; | |
1113 | case CCL_LSH: reg[rrr] <<= i; break; | |
1114 | case CCL_RSH: reg[rrr] >>= i; break; | |
1115 | case CCL_LSH8: reg[rrr] <<= 8; reg[rrr] |= i; break; | |
1116 | case CCL_RSH8: reg[7] = reg[rrr] & 0xFF; reg[rrr] >>= 8; break; | |
1117 | case CCL_DIVMOD: reg[7] = reg[rrr] % i; reg[rrr] /= i; break; | |
1118 | case CCL_LS: reg[rrr] = reg[rrr] < i; break; | |
1119 | case CCL_GT: reg[rrr] = reg[rrr] > i; break; | |
1120 | case CCL_EQ: reg[rrr] = reg[rrr] == i; break; | |
1121 | case CCL_LE: reg[rrr] = reg[rrr] <= i; break; | |
1122 | case CCL_GE: reg[rrr] = reg[rrr] >= i; break; | |
1123 | case CCL_NE: reg[rrr] = reg[rrr] != i; break; | |
1124 | default: CCL_INVALID_CMD; | |
1125 | } | |
1126 | break; | |
1127 | ||
1128 | case CCL_SetExprConst: /* 00000OPERATION000RRRrrrXXXXX */ | |
1129 | i = reg[RRR]; | |
1130 | j = XINT (ccl_prog[ic]); | |
1131 | op = field1 >> 6; | |
1132 | jump_address = ++ic; | |
1133 | goto ccl_set_expr; | |
1134 | ||
1135 | case CCL_SetExprReg: /* 00000OPERATIONRrrRRRrrrXXXXX */ | |
1136 | i = reg[RRR]; | |
1137 | j = reg[Rrr]; | |
1138 | op = field1 >> 6; | |
1139 | jump_address = ic; | |
1140 | goto ccl_set_expr; | |
1141 | ||
1142 | case CCL_ReadJumpCondExprConst: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
1143 | CCL_READ_CHAR (reg[rrr]); | |
1144 | case CCL_JumpCondExprConst: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
1145 | i = reg[rrr]; | |
1146 | op = XINT (ccl_prog[ic]); | |
1147 | jump_address = ic++ + ADDR; | |
1148 | j = XINT (ccl_prog[ic]); | |
1149 | ic++; | |
1150 | rrr = 7; | |
1151 | goto ccl_set_expr; | |
1152 | ||
1153 | case CCL_ReadJumpCondExprReg: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
1154 | CCL_READ_CHAR (reg[rrr]); | |
1155 | case CCL_JumpCondExprReg: | |
1156 | i = reg[rrr]; | |
1157 | op = XINT (ccl_prog[ic]); | |
1158 | jump_address = ic++ + ADDR; | |
1159 | j = reg[XINT (ccl_prog[ic])]; | |
1160 | ic++; | |
1161 | rrr = 7; | |
1162 | ||
1163 | ccl_set_expr: | |
1164 | switch (op) | |
1165 | { | |
1166 | case CCL_PLUS: reg[rrr] = i + j; break; | |
1167 | case CCL_MINUS: reg[rrr] = i - j; break; | |
1168 | case CCL_MUL: reg[rrr] = i * j; break; | |
1169 | case CCL_DIV: reg[rrr] = i / j; break; | |
1170 | case CCL_MOD: reg[rrr] = i % j; break; | |
1171 | case CCL_AND: reg[rrr] = i & j; break; | |
1172 | case CCL_OR: reg[rrr] = i | j; break; | |
1173 | case CCL_XOR: reg[rrr] = i ^ j;; break; | |
1174 | case CCL_LSH: reg[rrr] = i << j; break; | |
1175 | case CCL_RSH: reg[rrr] = i >> j; break; | |
1176 | case CCL_LSH8: reg[rrr] = (i << 8) | j; break; | |
1177 | case CCL_RSH8: reg[rrr] = i >> 8; reg[7] = i & 0xFF; break; | |
1178 | case CCL_DIVMOD: reg[rrr] = i / j; reg[7] = i % j; break; | |
1179 | case CCL_LS: reg[rrr] = i < j; break; | |
1180 | case CCL_GT: reg[rrr] = i > j; break; | |
1181 | case CCL_EQ: reg[rrr] = i == j; break; | |
1182 | case CCL_LE: reg[rrr] = i <= j; break; | |
1183 | case CCL_GE: reg[rrr] = i >= j; break; | |
1184 | case CCL_NE: reg[rrr] = i != j; break; | |
c10842ea KH |
1185 | case CCL_DECODE_SJIS: |
1186 | { | |
1187 | i = (i << 8) | j; | |
1188 | SJIS_TO_JIS (i); | |
1189 | reg[rrr] = i >> 8; | |
1190 | reg[7] = i & 0xFF; | |
1191 | break; | |
1192 | } | |
1193 | case CCL_ENCODE_SJIS: | |
1194 | { | |
1195 | i = (i << 8) | j; | |
1196 | JIS_TO_SJIS (i); | |
1197 | reg[rrr] = i >> 8; | |
1198 | reg[7] = i & 0xFF; | |
1199 | break; | |
1200 | } | |
4ed46869 KH |
1201 | default: CCL_INVALID_CMD; |
1202 | } | |
1203 | code &= 0x1F; | |
1204 | if (code == CCL_WriteExprConst || code == CCL_WriteExprRegister) | |
1205 | { | |
1206 | i = reg[rrr]; | |
1207 | CCL_WRITE_CHAR (i); | |
25660570 | 1208 | ic = jump_address; |
4ed46869 KH |
1209 | } |
1210 | else if (!reg[rrr]) | |
1211 | ic = jump_address; | |
1212 | break; | |
1213 | ||
450ed226 | 1214 | case CCL_Extension: |
e34b1164 KH |
1215 | switch (EXCMD) |
1216 | { | |
6ae21908 | 1217 | case CCL_ReadMultibyteChar2: |
e34b1164 KH |
1218 | if (!src) |
1219 | CCL_INVALID_CMD; | |
c10842ea | 1220 | CCL_READ_CHAR (i); |
bda731af | 1221 | CCL_ENCODE_CHAR (i, charset_list, reg[RRR], reg[rrr]); |
e34b1164 KH |
1222 | break; |
1223 | ||
6ae21908 | 1224 | case CCL_WriteMultibyteChar2: |
c10842ea KH |
1225 | if (! dst) |
1226 | CCL_INVALID_CMD; | |
bda731af | 1227 | i = CCL_DECODE_CHAR (reg[RRR], reg[rrr]); |
c10842ea | 1228 | CCL_WRITE_CHAR (i); |
e34b1164 KH |
1229 | break; |
1230 | ||
8146262a | 1231 | case CCL_TranslateCharacter: |
bda731af | 1232 | i = CCL_DECODE_CHAR (reg[RRR], reg[rrr]); |
c10842ea | 1233 | op = translate_char (GET_TRANSLATION_TABLE (reg[Rrr]), i); |
bda731af | 1234 | CCL_ENCODE_CHAR (op, charset_list, reg[RRR], reg[rrr]); |
e34b1164 KH |
1235 | break; |
1236 | ||
8146262a | 1237 | case CCL_TranslateCharacterConstTbl: |
e34b1164 KH |
1238 | op = XINT (ccl_prog[ic]); /* table */ |
1239 | ic++; | |
bda731af | 1240 | i = CCL_DECODE_CHAR (reg[RRR], reg[rrr]); |
c10842ea | 1241 | op = translate_char (GET_TRANSLATION_TABLE (op), i); |
bda731af | 1242 | CCL_ENCODE_CHAR (op, charset_list, reg[RRR], reg[rrr]); |
e34b1164 KH |
1243 | break; |
1244 | ||
d80dc57e DL |
1245 | case CCL_LookupIntConstTbl: |
1246 | op = XINT (ccl_prog[ic]); /* table */ | |
1247 | ic++; | |
177c0ea7 | 1248 | { |
d80dc57e DL |
1249 | struct Lisp_Hash_Table *h = GET_HASH_TABLE (op); |
1250 | ||
1251 | op = hash_lookup (h, make_number (reg[RRR]), NULL); | |
1252 | if (op >= 0) | |
1253 | { | |
f9bd23fd DL |
1254 | Lisp_Object opl; |
1255 | opl = HASH_VALUE (h, op); | |
0bc6bafd | 1256 | if (! CHARACTERP (opl)) |
d80dc57e | 1257 | CCL_INVALID_CMD; |
bda731af KH |
1258 | reg[RRR] = charset_unicode; |
1259 | reg[rrr] = op; | |
d80dc57e DL |
1260 | reg[7] = 1; /* r7 true for success */ |
1261 | } | |
1262 | else | |
1263 | reg[7] = 0; | |
1264 | } | |
1265 | break; | |
1266 | ||
1267 | case CCL_LookupCharConstTbl: | |
1268 | op = XINT (ccl_prog[ic]); /* table */ | |
1269 | ic++; | |
bda731af | 1270 | i = CCL_DECODE_CHAR (reg[RRR], reg[rrr]); |
177c0ea7 | 1271 | { |
d80dc57e DL |
1272 | struct Lisp_Hash_Table *h = GET_HASH_TABLE (op); |
1273 | ||
1274 | op = hash_lookup (h, make_number (i), NULL); | |
1275 | if (op >= 0) | |
1276 | { | |
f9bd23fd DL |
1277 | Lisp_Object opl; |
1278 | opl = HASH_VALUE (h, op); | |
1279 | if (!INTEGERP (opl)) | |
d80dc57e | 1280 | CCL_INVALID_CMD; |
f9bd23fd | 1281 | reg[RRR] = XINT (opl); |
d80dc57e DL |
1282 | reg[7] = 1; /* r7 true for success */ |
1283 | } | |
1284 | else | |
1285 | reg[7] = 0; | |
1286 | } | |
1287 | break; | |
1288 | ||
e34b1164 KH |
1289 | case CCL_IterateMultipleMap: |
1290 | { | |
8146262a | 1291 | Lisp_Object map, content, attrib, value; |
e34b1164 KH |
1292 | int point, size, fin_ic; |
1293 | ||
8146262a | 1294 | j = XINT (ccl_prog[ic++]); /* number of maps. */ |
e34b1164 KH |
1295 | fin_ic = ic + j; |
1296 | op = reg[rrr]; | |
1297 | if ((j > reg[RRR]) && (j >= 0)) | |
1298 | { | |
1299 | ic += reg[RRR]; | |
1300 | i = reg[RRR]; | |
1301 | } | |
1302 | else | |
1303 | { | |
1304 | reg[RRR] = -1; | |
1305 | ic = fin_ic; | |
1306 | break; | |
1307 | } | |
1308 | ||
1309 | for (;i < j;i++) | |
1310 | { | |
1311 | ||
64ef2921 | 1312 | size = ASIZE (Vcode_conversion_map_vector); |
d387866a | 1313 | point = XINT (ccl_prog[ic++]); |
e34b1164 | 1314 | if (point >= size) continue; |
64ef2921 | 1315 | map = AREF (Vcode_conversion_map_vector, point); |
8146262a KH |
1316 | |
1317 | /* Check map varidity. */ | |
1318 | if (!CONSP (map)) continue; | |
03699b14 | 1319 | map = XCDR (map); |
8146262a | 1320 | if (!VECTORP (map)) continue; |
64ef2921 | 1321 | size = ASIZE (map); |
e34b1164 | 1322 | if (size <= 1) continue; |
6ae21908 | 1323 | |
64ef2921 | 1324 | content = AREF (map, 0); |
6ae21908 | 1325 | |
8146262a | 1326 | /* check map type, |
6ae21908 KH |
1327 | [STARTPOINT VAL1 VAL2 ...] or |
1328 | [t ELELMENT STARTPOINT ENDPOINT] */ | |
1329 | if (NUMBERP (content)) | |
1330 | { | |
1331 | point = XUINT (content); | |
1332 | point = op - point + 1; | |
1333 | if (!((point >= 1) && (point < size))) continue; | |
64ef2921 | 1334 | content = AREF (map, point); |
6ae21908 KH |
1335 | } |
1336 | else if (EQ (content, Qt)) | |
1337 | { | |
1338 | if (size != 4) continue; | |
64ef2921 SM |
1339 | if ((op >= XUINT (AREF (map, 2))) |
1340 | && (op < XUINT (AREF (map, 3)))) | |
1341 | content = AREF (map, 1); | |
6ae21908 KH |
1342 | else |
1343 | continue; | |
1344 | } | |
177c0ea7 | 1345 | else |
6ae21908 | 1346 | continue; |
e34b1164 KH |
1347 | |
1348 | if (NILP (content)) | |
1349 | continue; | |
1350 | else if (NUMBERP (content)) | |
1351 | { | |
1352 | reg[RRR] = i; | |
6ae21908 | 1353 | reg[rrr] = XINT(content); |
e34b1164 KH |
1354 | break; |
1355 | } | |
1356 | else if (EQ (content, Qt) || EQ (content, Qlambda)) | |
1357 | { | |
1358 | reg[RRR] = i; | |
1359 | break; | |
1360 | } | |
1361 | else if (CONSP (content)) | |
1362 | { | |
03699b14 KR |
1363 | attrib = XCAR (content); |
1364 | value = XCDR (content); | |
e34b1164 KH |
1365 | if (!NUMBERP (attrib) || !NUMBERP (value)) |
1366 | continue; | |
1367 | reg[RRR] = i; | |
6ae21908 | 1368 | reg[rrr] = XUINT (value); |
e34b1164 KH |
1369 | break; |
1370 | } | |
54fa5bc1 KH |
1371 | else if (SYMBOLP (content)) |
1372 | CCL_CALL_FOR_MAP_INSTRUCTION (content, fin_ic); | |
1373 | else | |
1374 | CCL_INVALID_CMD; | |
e34b1164 KH |
1375 | } |
1376 | if (i == j) | |
1377 | reg[RRR] = -1; | |
1378 | ic = fin_ic; | |
1379 | } | |
1380 | break; | |
177c0ea7 | 1381 | |
8146262a | 1382 | case CCL_MapMultiple: |
e34b1164 | 1383 | { |
8146262a KH |
1384 | Lisp_Object map, content, attrib, value; |
1385 | int point, size, map_vector_size; | |
1386 | int map_set_rest_length, fin_ic; | |
54fa5bc1 KH |
1387 | int current_ic = this_ic; |
1388 | ||
1389 | /* inhibit recursive call on MapMultiple. */ | |
1390 | if (stack_idx_of_map_multiple > 0) | |
1391 | { | |
1392 | if (stack_idx_of_map_multiple <= stack_idx) | |
1393 | { | |
1394 | stack_idx_of_map_multiple = 0; | |
1395 | mapping_stack_pointer = mapping_stack; | |
1396 | CCL_INVALID_CMD; | |
1397 | } | |
1398 | } | |
1399 | else | |
1400 | mapping_stack_pointer = mapping_stack; | |
1401 | stack_idx_of_map_multiple = 0; | |
8146262a KH |
1402 | |
1403 | map_set_rest_length = | |
1404 | XINT (ccl_prog[ic++]); /* number of maps and separators. */ | |
1405 | fin_ic = ic + map_set_rest_length; | |
54fa5bc1 KH |
1406 | op = reg[rrr]; |
1407 | ||
8146262a | 1408 | if ((map_set_rest_length > reg[RRR]) && (reg[RRR] >= 0)) |
e34b1164 KH |
1409 | { |
1410 | ic += reg[RRR]; | |
1411 | i = reg[RRR]; | |
8146262a | 1412 | map_set_rest_length -= i; |
e34b1164 KH |
1413 | } |
1414 | else | |
1415 | { | |
1416 | ic = fin_ic; | |
1417 | reg[RRR] = -1; | |
54fa5bc1 | 1418 | mapping_stack_pointer = mapping_stack; |
e34b1164 KH |
1419 | break; |
1420 | } | |
6ae21908 | 1421 | |
54fa5bc1 KH |
1422 | if (mapping_stack_pointer <= (mapping_stack + 1)) |
1423 | { | |
1424 | /* Set up initial state. */ | |
1425 | mapping_stack_pointer = mapping_stack; | |
1426 | PUSH_MAPPING_STACK (0, op); | |
1427 | reg[RRR] = -1; | |
1428 | } | |
1429 | else | |
1430 | { | |
1431 | /* Recover after calling other ccl program. */ | |
1432 | int orig_op; | |
e34b1164 | 1433 | |
54fa5bc1 KH |
1434 | POP_MAPPING_STACK (map_set_rest_length, orig_op); |
1435 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1436 | switch (op) | |
e34b1164 | 1437 | { |
54fa5bc1 KH |
1438 | case -1: |
1439 | /* Regard it as Qnil. */ | |
1440 | op = orig_op; | |
1441 | i++; | |
1442 | ic++; | |
1443 | map_set_rest_length--; | |
1444 | break; | |
1445 | case -2: | |
1446 | /* Regard it as Qt. */ | |
e34b1164 | 1447 | op = reg[rrr]; |
54fa5bc1 KH |
1448 | i++; |
1449 | ic++; | |
1450 | map_set_rest_length--; | |
1451 | break; | |
1452 | case -3: | |
1453 | /* Regard it as Qlambda. */ | |
1454 | op = orig_op; | |
1455 | i += map_set_rest_length; | |
1456 | ic += map_set_rest_length; | |
1457 | map_set_rest_length = 0; | |
1458 | break; | |
1459 | default: | |
1460 | /* Regard it as normal mapping. */ | |
8146262a | 1461 | i += map_set_rest_length; |
54fa5bc1 | 1462 | ic += map_set_rest_length; |
8146262a | 1463 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); |
6ae21908 KH |
1464 | break; |
1465 | } | |
e34b1164 | 1466 | } |
64ef2921 | 1467 | map_vector_size = ASIZE (Vcode_conversion_map_vector); |
177c0ea7 | 1468 | |
54fa5bc1 KH |
1469 | do { |
1470 | for (;map_set_rest_length > 0;i++, ic++, map_set_rest_length--) | |
1471 | { | |
1472 | point = XINT(ccl_prog[ic]); | |
1473 | if (point < 0) | |
1474 | { | |
1475 | /* +1 is for including separator. */ | |
1476 | point = -point + 1; | |
1477 | if (mapping_stack_pointer | |
1478 | >= &mapping_stack[MAX_MAP_SET_LEVEL]) | |
1479 | CCL_INVALID_CMD; | |
1480 | PUSH_MAPPING_STACK (map_set_rest_length - point, | |
1481 | reg[rrr]); | |
1482 | map_set_rest_length = point; | |
1483 | reg[rrr] = op; | |
1484 | continue; | |
1485 | } | |
1486 | ||
1487 | if (point >= map_vector_size) continue; | |
64ef2921 | 1488 | map = AREF (Vcode_conversion_map_vector, point); |
54fa5bc1 KH |
1489 | |
1490 | /* Check map varidity. */ | |
1491 | if (!CONSP (map)) continue; | |
1492 | map = XCDR (map); | |
1493 | if (!VECTORP (map)) continue; | |
64ef2921 | 1494 | size = ASIZE (map); |
54fa5bc1 KH |
1495 | if (size <= 1) continue; |
1496 | ||
64ef2921 | 1497 | content = AREF (map, 0); |
54fa5bc1 KH |
1498 | |
1499 | /* check map type, | |
1500 | [STARTPOINT VAL1 VAL2 ...] or | |
1501 | [t ELEMENT STARTPOINT ENDPOINT] */ | |
1502 | if (NUMBERP (content)) | |
1503 | { | |
1504 | point = XUINT (content); | |
1505 | point = op - point + 1; | |
1506 | if (!((point >= 1) && (point < size))) continue; | |
64ef2921 | 1507 | content = AREF (map, point); |
54fa5bc1 KH |
1508 | } |
1509 | else if (EQ (content, Qt)) | |
1510 | { | |
1511 | if (size != 4) continue; | |
64ef2921 SM |
1512 | if ((op >= XUINT (AREF (map, 2))) && |
1513 | (op < XUINT (AREF (map, 3)))) | |
1514 | content = AREF (map, 1); | |
54fa5bc1 KH |
1515 | else |
1516 | continue; | |
1517 | } | |
177c0ea7 | 1518 | else |
54fa5bc1 KH |
1519 | continue; |
1520 | ||
1521 | if (NILP (content)) | |
1522 | continue; | |
1523 | ||
1524 | reg[RRR] = i; | |
1525 | if (NUMBERP (content)) | |
1526 | { | |
1527 | op = XINT (content); | |
1528 | i += map_set_rest_length - 1; | |
1529 | ic += map_set_rest_length - 1; | |
1530 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1531 | map_set_rest_length++; | |
1532 | } | |
1533 | else if (CONSP (content)) | |
1534 | { | |
1535 | attrib = XCAR (content); | |
1536 | value = XCDR (content); | |
1537 | if (!NUMBERP (attrib) || !NUMBERP (value)) | |
1538 | continue; | |
1539 | op = XUINT (value); | |
1540 | i += map_set_rest_length - 1; | |
1541 | ic += map_set_rest_length - 1; | |
1542 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1543 | map_set_rest_length++; | |
1544 | } | |
1545 | else if (EQ (content, Qt)) | |
1546 | { | |
1547 | op = reg[rrr]; | |
1548 | } | |
1549 | else if (EQ (content, Qlambda)) | |
1550 | { | |
1551 | i += map_set_rest_length; | |
1552 | ic += map_set_rest_length; | |
1553 | break; | |
1554 | } | |
1555 | else if (SYMBOLP (content)) | |
1556 | { | |
1557 | if (mapping_stack_pointer | |
1558 | >= &mapping_stack[MAX_MAP_SET_LEVEL]) | |
1559 | CCL_INVALID_CMD; | |
1560 | PUSH_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1561 | PUSH_MAPPING_STACK (map_set_rest_length, op); | |
1562 | stack_idx_of_map_multiple = stack_idx + 1; | |
1563 | CCL_CALL_FOR_MAP_INSTRUCTION (content, current_ic); | |
1564 | } | |
1565 | else | |
1566 | CCL_INVALID_CMD; | |
1567 | } | |
1568 | if (mapping_stack_pointer <= (mapping_stack + 1)) | |
1569 | break; | |
1570 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1571 | i += map_set_rest_length; | |
1572 | ic += map_set_rest_length; | |
1573 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1574 | } while (1); | |
1575 | ||
e34b1164 KH |
1576 | ic = fin_ic; |
1577 | } | |
1578 | reg[rrr] = op; | |
1579 | break; | |
1580 | ||
8146262a | 1581 | case CCL_MapSingle: |
e34b1164 | 1582 | { |
8146262a | 1583 | Lisp_Object map, attrib, value, content; |
e34b1164 | 1584 | int size, point; |
8146262a | 1585 | j = XINT (ccl_prog[ic++]); /* map_id */ |
e34b1164 | 1586 | op = reg[rrr]; |
64ef2921 | 1587 | if (j >= ASIZE (Vcode_conversion_map_vector)) |
e34b1164 KH |
1588 | { |
1589 | reg[RRR] = -1; | |
1590 | break; | |
1591 | } | |
64ef2921 | 1592 | map = AREF (Vcode_conversion_map_vector, j); |
8146262a | 1593 | if (!CONSP (map)) |
e34b1164 KH |
1594 | { |
1595 | reg[RRR] = -1; | |
1596 | break; | |
1597 | } | |
03699b14 | 1598 | map = XCDR (map); |
8146262a | 1599 | if (!VECTORP (map)) |
e34b1164 KH |
1600 | { |
1601 | reg[RRR] = -1; | |
1602 | break; | |
1603 | } | |
64ef2921 SM |
1604 | size = ASIZE (map); |
1605 | point = XUINT (AREF (map, 0)); | |
e34b1164 KH |
1606 | point = op - point + 1; |
1607 | reg[RRR] = 0; | |
1608 | if ((size <= 1) || | |
1609 | (!((point >= 1) && (point < size)))) | |
1610 | reg[RRR] = -1; | |
1611 | else | |
1612 | { | |
b1cab202 | 1613 | reg[RRR] = 0; |
64ef2921 | 1614 | content = AREF (map, point); |
e34b1164 KH |
1615 | if (NILP (content)) |
1616 | reg[RRR] = -1; | |
1617 | else if (NUMBERP (content)) | |
6ae21908 | 1618 | reg[rrr] = XINT (content); |
b1cab202 | 1619 | else if (EQ (content, Qt)); |
e34b1164 KH |
1620 | else if (CONSP (content)) |
1621 | { | |
03699b14 KR |
1622 | attrib = XCAR (content); |
1623 | value = XCDR (content); | |
e34b1164 KH |
1624 | if (!NUMBERP (attrib) || !NUMBERP (value)) |
1625 | continue; | |
1626 | reg[rrr] = XUINT(value); | |
1627 | break; | |
1628 | } | |
54fa5bc1 KH |
1629 | else if (SYMBOLP (content)) |
1630 | CCL_CALL_FOR_MAP_INSTRUCTION (content, ic); | |
e34b1164 KH |
1631 | else |
1632 | reg[RRR] = -1; | |
1633 | } | |
1634 | } | |
1635 | break; | |
177c0ea7 | 1636 | |
e34b1164 KH |
1637 | default: |
1638 | CCL_INVALID_CMD; | |
1639 | } | |
1640 | break; | |
1641 | ||
4ed46869 KH |
1642 | default: |
1643 | CCL_INVALID_CMD; | |
1644 | } | |
1645 | } | |
1646 | ||
1647 | ccl_error_handler: | |
0fb94c7f EZ |
1648 | /* The suppress_error member is set when e.g. a CCL-based coding |
1649 | system is used for terminal output. */ | |
1650 | if (!ccl->suppress_error && destination) | |
4ed46869 KH |
1651 | { |
1652 | /* We can insert an error message only if DESTINATION is | |
1653 | specified and we still have a room to store the message | |
1654 | there. */ | |
1655 | char msg[256]; | |
1656 | int msglen; | |
1657 | ||
12abd7d1 KH |
1658 | if (!dst) |
1659 | dst = destination; | |
1660 | ||
4ed46869 KH |
1661 | switch (ccl->status) |
1662 | { | |
1663 | case CCL_STAT_INVALID_CMD: | |
1664 | sprintf(msg, "\nCCL: Invalid command %x (ccl_code = %x) at %d.", | |
519bf146 | 1665 | code & 0x1F, code, this_ic); |
4ed46869 KH |
1666 | #ifdef CCL_DEBUG |
1667 | { | |
1668 | int i = ccl_backtrace_idx - 1; | |
1669 | int j; | |
1670 | ||
1671 | msglen = strlen (msg); | |
12abd7d1 | 1672 | if (dst + msglen <= (dst_bytes ? dst_end : src)) |
4ed46869 KH |
1673 | { |
1674 | bcopy (msg, dst, msglen); | |
1675 | dst += msglen; | |
1676 | } | |
1677 | ||
1678 | for (j = 0; j < CCL_DEBUG_BACKTRACE_LEN; j++, i--) | |
1679 | { | |
1680 | if (i < 0) i = CCL_DEBUG_BACKTRACE_LEN - 1; | |
1681 | if (ccl_backtrace_table[i] == 0) | |
1682 | break; | |
1683 | sprintf(msg, " %d", ccl_backtrace_table[i]); | |
1684 | msglen = strlen (msg); | |
12abd7d1 | 1685 | if (dst + msglen > (dst_bytes ? dst_end : src)) |
4ed46869 KH |
1686 | break; |
1687 | bcopy (msg, dst, msglen); | |
1688 | dst += msglen; | |
1689 | } | |
12abd7d1 | 1690 | goto ccl_finish; |
4ed46869 | 1691 | } |
4ed46869 | 1692 | #endif |
12abd7d1 | 1693 | break; |
4ed46869 KH |
1694 | |
1695 | case CCL_STAT_QUIT: | |
1696 | sprintf(msg, "\nCCL: Quited."); | |
1697 | break; | |
1698 | ||
1699 | default: | |
1700 | sprintf(msg, "\nCCL: Unknown error type (%d).", ccl->status); | |
1701 | } | |
1702 | ||
1703 | msglen = strlen (msg); | |
c10842ea | 1704 | if (dst + msglen <= dst_end) |
4ed46869 | 1705 | { |
c10842ea KH |
1706 | for (i = 0; i < msglen; i++) |
1707 | *dst++ = msg[i]; | |
4ed46869 | 1708 | } |
177c0ea7 | 1709 | |
31165028 KH |
1710 | if (ccl->status == CCL_STAT_INVALID_CMD) |
1711 | { | |
8a1ae4dd GM |
1712 | #if 0 /* If the remaining bytes contain 0x80..0x9F, copying them |
1713 | results in an invalid multibyte sequence. */ | |
1714 | ||
31165028 KH |
1715 | /* Copy the remaining source data. */ |
1716 | int i = src_end - src; | |
1717 | if (dst_bytes && (dst_end - dst) < i) | |
1718 | i = dst_end - dst; | |
1719 | bcopy (src, dst, i); | |
1720 | src += i; | |
1721 | dst += i; | |
8a1ae4dd GM |
1722 | #else |
1723 | /* Signal that we've consumed everything. */ | |
1724 | src = src_end; | |
1725 | #endif | |
31165028 | 1726 | } |
4ed46869 KH |
1727 | } |
1728 | ||
1729 | ccl_finish: | |
1730 | ccl->ic = ic; | |
c13362d8 KH |
1731 | ccl->stack_idx = stack_idx; |
1732 | ccl->prog = ccl_prog; | |
c10842ea KH |
1733 | ccl->consumed = src - source; |
1734 | ccl->produced = dst - destination; | |
4ed46869 KH |
1735 | } |
1736 | ||
5232fa7b KH |
1737 | /* Resolve symbols in the specified CCL code (Lisp vector). This |
1738 | function converts symbols of code conversion maps and character | |
1739 | translation tables embeded in the CCL code into their ID numbers. | |
1740 | ||
1741 | The return value is a vector (CCL itself or a new vector in which | |
1742 | all symbols are resolved), Qt if resolving of some symbol failed, | |
1743 | or nil if CCL contains invalid data. */ | |
1744 | ||
1745 | static Lisp_Object | |
1746 | resolve_symbol_ccl_program (ccl) | |
1747 | Lisp_Object ccl; | |
1748 | { | |
1749 | int i, veclen, unresolved = 0; | |
1750 | Lisp_Object result, contents, val; | |
1751 | ||
1752 | result = ccl; | |
64ef2921 | 1753 | veclen = ASIZE (result); |
5232fa7b KH |
1754 | |
1755 | for (i = 0; i < veclen; i++) | |
1756 | { | |
64ef2921 | 1757 | contents = AREF (result, i); |
5232fa7b KH |
1758 | if (INTEGERP (contents)) |
1759 | continue; | |
1760 | else if (CONSP (contents) | |
03699b14 KR |
1761 | && SYMBOLP (XCAR (contents)) |
1762 | && SYMBOLP (XCDR (contents))) | |
5232fa7b KH |
1763 | { |
1764 | /* This is the new style for embedding symbols. The form is | |
1765 | (SYMBOL . PROPERTY). (get SYMBOL PROPERTY) should give | |
1766 | an index number. */ | |
1767 | ||
1768 | if (EQ (result, ccl)) | |
1769 | result = Fcopy_sequence (ccl); | |
1770 | ||
03699b14 | 1771 | val = Fget (XCAR (contents), XCDR (contents)); |
5232fa7b | 1772 | if (NATNUMP (val)) |
64ef2921 | 1773 | AREF (result, i) = val; |
5232fa7b KH |
1774 | else |
1775 | unresolved = 1; | |
1776 | continue; | |
1777 | } | |
1778 | else if (SYMBOLP (contents)) | |
1779 | { | |
1780 | /* This is the old style for embedding symbols. This style | |
1781 | may lead to a bug if, for instance, a translation table | |
1782 | and a code conversion map have the same name. */ | |
1783 | if (EQ (result, ccl)) | |
1784 | result = Fcopy_sequence (ccl); | |
1785 | ||
1786 | val = Fget (contents, Qtranslation_table_id); | |
1787 | if (NATNUMP (val)) | |
64ef2921 | 1788 | AREF (result, i) = val; |
5232fa7b KH |
1789 | else |
1790 | { | |
1791 | val = Fget (contents, Qcode_conversion_map_id); | |
1792 | if (NATNUMP (val)) | |
64ef2921 | 1793 | AREF (result, i) = val; |
5232fa7b KH |
1794 | else |
1795 | { | |
1796 | val = Fget (contents, Qccl_program_idx); | |
1797 | if (NATNUMP (val)) | |
64ef2921 | 1798 | AREF (result, i) = val; |
5232fa7b KH |
1799 | else |
1800 | unresolved = 1; | |
1801 | } | |
1802 | } | |
1803 | continue; | |
1804 | } | |
1805 | return Qnil; | |
1806 | } | |
1807 | ||
1808 | return (unresolved ? Qt : result); | |
1809 | } | |
1810 | ||
1811 | /* Return the compiled code (vector) of CCL program CCL_PROG. | |
1812 | CCL_PROG is a name (symbol) of the program or already compiled | |
1813 | code. If necessary, resolve symbols in the compiled code to index | |
1814 | numbers. If we failed to get the compiled code or to resolve | |
1815 | symbols, return Qnil. */ | |
1816 | ||
1817 | static Lisp_Object | |
1818 | ccl_get_compiled_code (ccl_prog) | |
1819 | Lisp_Object ccl_prog; | |
1820 | { | |
1821 | Lisp_Object val, slot; | |
1822 | ||
1823 | if (VECTORP (ccl_prog)) | |
1824 | { | |
1825 | val = resolve_symbol_ccl_program (ccl_prog); | |
1826 | return (VECTORP (val) ? val : Qnil); | |
1827 | } | |
1828 | if (!SYMBOLP (ccl_prog)) | |
1829 | return Qnil; | |
1830 | ||
1831 | val = Fget (ccl_prog, Qccl_program_idx); | |
1832 | if (! NATNUMP (val) | |
64ef2921 | 1833 | || XINT (val) >= ASIZE (Vccl_program_table)) |
5232fa7b | 1834 | return Qnil; |
64ef2921 | 1835 | slot = AREF (Vccl_program_table, XINT (val)); |
5232fa7b | 1836 | if (! VECTORP (slot) |
64ef2921 SM |
1837 | || ASIZE (slot) != 3 |
1838 | || ! VECTORP (AREF (slot, 1))) | |
5232fa7b | 1839 | return Qnil; |
64ef2921 | 1840 | if (NILP (AREF (slot, 2))) |
5232fa7b | 1841 | { |
64ef2921 | 1842 | val = resolve_symbol_ccl_program (AREF (slot, 1)); |
5232fa7b KH |
1843 | if (! VECTORP (val)) |
1844 | return Qnil; | |
64ef2921 SM |
1845 | AREF (slot, 1) = val; |
1846 | AREF (slot, 2) = Qt; | |
5232fa7b | 1847 | } |
64ef2921 | 1848 | return AREF (slot, 1); |
5232fa7b KH |
1849 | } |
1850 | ||
4ed46869 | 1851 | /* Setup fields of the structure pointed by CCL appropriately for the |
5232fa7b KH |
1852 | execution of CCL program CCL_PROG. CCL_PROG is the name (symbol) |
1853 | of the CCL program or the already compiled code (vector). | |
1854 | Return 0 if we succeed this setup, else return -1. | |
1855 | ||
1856 | If CCL_PROG is nil, we just reset the structure pointed by CCL. */ | |
1857 | int | |
1858 | setup_ccl_program (ccl, ccl_prog) | |
4ed46869 | 1859 | struct ccl_program *ccl; |
5232fa7b | 1860 | Lisp_Object ccl_prog; |
4ed46869 KH |
1861 | { |
1862 | int i; | |
1863 | ||
5232fa7b | 1864 | if (! NILP (ccl_prog)) |
ad3d1b1d | 1865 | { |
5232fa7b | 1866 | struct Lisp_Vector *vp; |
ad3d1b1d | 1867 | |
5232fa7b KH |
1868 | ccl_prog = ccl_get_compiled_code (ccl_prog); |
1869 | if (! VECTORP (ccl_prog)) | |
1870 | return -1; | |
1871 | vp = XVECTOR (ccl_prog); | |
ad3d1b1d KH |
1872 | ccl->size = vp->size; |
1873 | ccl->prog = vp->contents; | |
1874 | ccl->eof_ic = XINT (vp->contents[CCL_HEADER_EOF]); | |
1875 | ccl->buf_magnification = XINT (vp->contents[CCL_HEADER_BUF_MAG]); | |
1876 | } | |
4ed46869 | 1877 | ccl->ic = CCL_HEADER_MAIN; |
4ed46869 KH |
1878 | for (i = 0; i < 8; i++) |
1879 | ccl->reg[i] = 0; | |
1880 | ccl->last_block = 0; | |
e34b1164 | 1881 | ccl->private_state = 0; |
4ed46869 | 1882 | ccl->status = 0; |
c13362d8 | 1883 | ccl->stack_idx = 0; |
ae08ba36 | 1884 | ccl->suppress_error = 0; |
fd40a25f | 1885 | ccl->eight_bit_control = 0; |
5232fa7b | 1886 | return 0; |
4ed46869 KH |
1887 | } |
1888 | ||
5232fa7b | 1889 | DEFUN ("ccl-program-p", Fccl_program_p, Sccl_program_p, 1, 1, 0, |
fdb82f93 PJ |
1890 | doc: /* Return t if OBJECT is a CCL program name or a compiled CCL program code. |
1891 | See the documentation of `define-ccl-program' for the detail of CCL program. */) | |
1892 | (object) | |
5232fa7b | 1893 | Lisp_Object object; |
6ae21908 | 1894 | { |
5232fa7b | 1895 | Lisp_Object val; |
6ae21908 | 1896 | |
5232fa7b | 1897 | if (VECTORP (object)) |
6ae21908 | 1898 | { |
5232fa7b KH |
1899 | val = resolve_symbol_ccl_program (object); |
1900 | return (VECTORP (val) ? Qt : Qnil); | |
6ae21908 | 1901 | } |
5232fa7b KH |
1902 | if (!SYMBOLP (object)) |
1903 | return Qnil; | |
6ae21908 | 1904 | |
5232fa7b KH |
1905 | val = Fget (object, Qccl_program_idx); |
1906 | return ((! NATNUMP (val) | |
64ef2921 | 1907 | || XINT (val) >= ASIZE (Vccl_program_table)) |
5232fa7b | 1908 | ? Qnil : Qt); |
6ae21908 KH |
1909 | } |
1910 | ||
4ed46869 | 1911 | DEFUN ("ccl-execute", Fccl_execute, Sccl_execute, 2, 2, 0, |
fdb82f93 PJ |
1912 | doc: /* Execute CCL-PROGRAM with registers initialized by REGISTERS. |
1913 | ||
1914 | CCL-PROGRAM is a CCL program name (symbol) | |
1915 | or compiled code generated by `ccl-compile' (for backward compatibility. | |
1916 | In the latter case, the execution overhead is bigger than in the former). | |
1917 | No I/O commands should appear in CCL-PROGRAM. | |
1918 | ||
1919 | REGISTERS is a vector of [R0 R1 ... R7] where RN is an initial value | |
1920 | for the Nth register. | |
1921 | ||
1922 | As side effect, each element of REGISTERS holds the value of | |
1923 | the corresponding register after the execution. | |
1924 | ||
1925 | See the documentation of `define-ccl-program' for a definition of CCL | |
1926 | programs. */) | |
1927 | (ccl_prog, reg) | |
4ed46869 KH |
1928 | Lisp_Object ccl_prog, reg; |
1929 | { | |
1930 | struct ccl_program ccl; | |
1931 | int i; | |
1932 | ||
5232fa7b KH |
1933 | if (setup_ccl_program (&ccl, ccl_prog) < 0) |
1934 | error ("Invalid CCL program"); | |
6ae21908 | 1935 | |
b7826503 | 1936 | CHECK_VECTOR (reg); |
64ef2921 | 1937 | if (ASIZE (reg) != 8) |
d7e1fe1f | 1938 | error ("Length of vector REGISTERS is not 8"); |
4ed46869 | 1939 | |
4ed46869 | 1940 | for (i = 0; i < 8; i++) |
64ef2921 SM |
1941 | ccl.reg[i] = (INTEGERP (AREF (reg, i)) |
1942 | ? XINT (AREF (reg, i)) | |
4ed46869 KH |
1943 | : 0); |
1944 | ||
bda731af | 1945 | ccl_driver (&ccl, NULL, NULL, 0, 0, Qnil); |
4ed46869 KH |
1946 | QUIT; |
1947 | if (ccl.status != CCL_STAT_SUCCESS) | |
1948 | error ("Error in CCL program at %dth code", ccl.ic); | |
1949 | ||
1950 | for (i = 0; i < 8; i++) | |
64ef2921 | 1951 | XSETINT (AREF (reg, i), ccl.reg[i]); |
4ed46869 KH |
1952 | return Qnil; |
1953 | } | |
1954 | ||
1955 | DEFUN ("ccl-execute-on-string", Fccl_execute_on_string, Sccl_execute_on_string, | |
39a68837 | 1956 | 3, 5, 0, |
fdb82f93 PJ |
1957 | doc: /* Execute CCL-PROGRAM with initial STATUS on STRING. |
1958 | ||
1959 | CCL-PROGRAM is a symbol registered by register-ccl-program, | |
1960 | or a compiled code generated by `ccl-compile' (for backward compatibility, | |
1961 | in this case, the execution is slower). | |
1962 | ||
1963 | Read buffer is set to STRING, and write buffer is allocated automatically. | |
1964 | ||
1965 | STATUS is a vector of [R0 R1 ... R7 IC], where | |
1966 | R0..R7 are initial values of corresponding registers, | |
1967 | IC is the instruction counter specifying from where to start the program. | |
1968 | If R0..R7 are nil, they are initialized to 0. | |
1969 | If IC is nil, it is initialized to head of the CCL program. | |
1970 | ||
1971 | If optional 4th arg CONTINUE is non-nil, keep IC on read operation | |
1972 | when read buffer is exausted, else, IC is always set to the end of | |
1973 | CCL-PROGRAM on exit. | |
1974 | ||
1975 | It returns the contents of write buffer as a string, | |
1976 | and as side effect, STATUS is updated. | |
1977 | If the optional 5th arg UNIBYTE-P is non-nil, the returned string | |
1978 | is a unibyte string. By default it is a multibyte string. | |
1979 | ||
1980 | See the documentation of `define-ccl-program' for the detail of CCL program. */) | |
1981 | (ccl_prog, status, str, contin, unibyte_p) | |
39a68837 | 1982 | Lisp_Object ccl_prog, status, str, contin, unibyte_p; |
4ed46869 KH |
1983 | { |
1984 | Lisp_Object val; | |
1985 | struct ccl_program ccl; | |
c10842ea | 1986 | int i; |
4ed46869 | 1987 | int outbufsize; |
c10842ea KH |
1988 | unsigned char *outbuf, *outp; |
1989 | int str_chars, str_bytes; | |
1990 | #define CCL_EXECUTE_BUF_SIZE 1024 | |
1991 | int source[CCL_EXECUTE_BUF_SIZE], destination[CCL_EXECUTE_BUF_SIZE]; | |
1992 | int consumed_chars, consumed_bytes, produced_chars; | |
6ae21908 | 1993 | |
5232fa7b KH |
1994 | if (setup_ccl_program (&ccl, ccl_prog) < 0) |
1995 | error ("Invalid CCL program"); | |
4ed46869 | 1996 | |
b7826503 | 1997 | CHECK_VECTOR (status); |
64ef2921 | 1998 | if (ASIZE (status) != 9) |
5232fa7b | 1999 | error ("Length of vector STATUS is not 9"); |
b7826503 | 2000 | CHECK_STRING (str); |
4ed46869 | 2001 | |
8f924df7 KH |
2002 | str_chars = SCHARS (str); |
2003 | str_bytes = SBYTES (str); | |
5232fa7b | 2004 | |
4ed46869 KH |
2005 | for (i = 0; i < 8; i++) |
2006 | { | |
64ef2921 SM |
2007 | if (NILP (AREF (status, i))) |
2008 | XSETINT (AREF (status, i), 0); | |
2009 | if (INTEGERP (AREF (status, i))) | |
2010 | ccl.reg[i] = XINT (AREF (status, i)); | |
4ed46869 | 2011 | } |
64ef2921 | 2012 | if (INTEGERP (AREF (status, i))) |
4ed46869 | 2013 | { |
64ef2921 | 2014 | i = XFASTINT (AREF (status, 8)); |
4ed46869 KH |
2015 | if (ccl.ic < i && i < ccl.size) |
2016 | ccl.ic = i; | |
2017 | } | |
4ed46869 | 2018 | |
c10842ea KH |
2019 | outbufsize = (ccl.buf_magnification |
2020 | ? str_bytes * ccl.buf_magnification + 256 | |
2021 | : str_bytes + 256); | |
2022 | outp = outbuf = (unsigned char *) xmalloc (outbufsize); | |
2023 | ||
2024 | consumed_chars = consumed_bytes = 0; | |
2025 | produced_chars = 0; | |
99e293b5 | 2026 | while (1) |
a3d8fcf2 | 2027 | { |
8f924df7 KH |
2028 | const unsigned char *p = SDATA (str) + consumed_bytes; |
2029 | const unsigned char *endp = SDATA (str) + str_bytes; | |
c10842ea KH |
2030 | int i = 0; |
2031 | int *src, src_size; | |
2032 | ||
2033 | if (endp - p == str_chars - consumed_chars) | |
2034 | while (i < CCL_EXECUTE_BUF_SIZE && p < endp) | |
2035 | source[i++] = *p++; | |
2036 | else | |
2037 | while (i < CCL_EXECUTE_BUF_SIZE && p < endp) | |
2038 | source[i++] = STRING_CHAR_ADVANCE (p); | |
2039 | consumed_chars += i; | |
8f924df7 | 2040 | consumed_bytes = p - SDATA (str); |
c10842ea KH |
2041 | |
2042 | if (consumed_bytes == str_bytes) | |
2043 | ccl.last_block = NILP (contin); | |
2044 | src = source; | |
2045 | src_size = i; | |
2046 | while (1) | |
2047 | { | |
bda731af KH |
2048 | ccl_driver (&ccl, src, destination, src_size, CCL_EXECUTE_BUF_SIZE, |
2049 | Qnil); | |
c10842ea KH |
2050 | produced_chars += ccl.produced; |
2051 | if (NILP (unibyte_p)) | |
2052 | { | |
2053 | if (outp - outbuf + MAX_MULTIBYTE_LENGTH * ccl.produced | |
2054 | > outbufsize) | |
2055 | { | |
2056 | int offset = outp - outbuf; | |
2057 | outbufsize += MAX_MULTIBYTE_LENGTH * ccl.produced; | |
2058 | outbuf = (unsigned char *) xrealloc (outbuf, outbufsize); | |
2059 | outp = outbuf + offset; | |
2060 | } | |
2061 | for (i = 0; i < ccl.produced; i++) | |
2062 | CHAR_STRING_ADVANCE (destination[i], outp); | |
2063 | } | |
2064 | else | |
2065 | { | |
2066 | if (outp - outbuf + ccl.produced > outbufsize) | |
2067 | { | |
2068 | int offset = outp - outbuf; | |
2069 | outbufsize += ccl.produced; | |
2070 | outbuf = (unsigned char *) xrealloc (outbuf, outbufsize); | |
2071 | outp = outbuf + offset; | |
2072 | } | |
2073 | for (i = 0; i < ccl.produced; i++) | |
2074 | *outp++ = destination[i]; | |
2075 | } | |
2076 | src += ccl.consumed; | |
2077 | src_size -= ccl.consumed; | |
99e293b5 KH |
2078 | if (ccl.status != CCL_STAT_SUSPEND_BY_DST) |
2079 | break; | |
c10842ea | 2080 | } |
a3d8fcf2 | 2081 | |
c10842ea KH |
2082 | if (ccl.status != CCL_STAT_SUSPEND_BY_SRC) |
2083 | break; | |
a3d8fcf2 | 2084 | } |
a3d8fcf2 | 2085 | |
4ed46869 | 2086 | if (ccl.status != CCL_STAT_SUCCESS |
a3d8fcf2 | 2087 | && ccl.status != CCL_STAT_SUSPEND_BY_SRC) |
4ed46869 KH |
2088 | error ("Error in CCL program at %dth code", ccl.ic); |
2089 | ||
c10842ea KH |
2090 | for (i = 0; i < 8; i++) |
2091 | XSET (XVECTOR (status)->contents[i], Lisp_Int, ccl.reg[i]); | |
2092 | XSETINT (XVECTOR (status)->contents[8], ccl.ic); | |
2093 | ||
2094 | if (NILP (unibyte_p)) | |
2095 | val = make_multibyte_string ((char *) outbuf, produced_chars, | |
2096 | outp - outbuf); | |
2097 | else | |
2098 | val = make_unibyte_string ((char *) outbuf, produced_chars); | |
2099 | xfree (outbuf); | |
4ed46869 KH |
2100 | |
2101 | return val; | |
2102 | } | |
2103 | ||
2104 | DEFUN ("register-ccl-program", Fregister_ccl_program, Sregister_ccl_program, | |
2105 | 2, 2, 0, | |
fdb82f93 PJ |
2106 | doc: /* Register CCL program CCL_PROG as NAME in `ccl-program-table'. |
2107 | CCL_PROG should be a compiled CCL program (vector), or nil. | |
2108 | If it is nil, just reserve NAME as a CCL program name. | |
2109 | Return index number of the registered CCL program. */) | |
2110 | (name, ccl_prog) | |
4ed46869 KH |
2111 | Lisp_Object name, ccl_prog; |
2112 | { | |
64ef2921 | 2113 | int len = ASIZE (Vccl_program_table); |
5232fa7b KH |
2114 | int idx; |
2115 | Lisp_Object resolved; | |
4ed46869 | 2116 | |
b7826503 | 2117 | CHECK_SYMBOL (name); |
5232fa7b | 2118 | resolved = Qnil; |
4ed46869 | 2119 | if (!NILP (ccl_prog)) |
6ae21908 | 2120 | { |
b7826503 | 2121 | CHECK_VECTOR (ccl_prog); |
5232fa7b | 2122 | resolved = resolve_symbol_ccl_program (ccl_prog); |
4d247a1f KH |
2123 | if (NILP (resolved)) |
2124 | error ("Error in CCL program"); | |
2125 | if (VECTORP (resolved)) | |
5232fa7b KH |
2126 | { |
2127 | ccl_prog = resolved; | |
2128 | resolved = Qt; | |
2129 | } | |
4d247a1f KH |
2130 | else |
2131 | resolved = Qnil; | |
6ae21908 | 2132 | } |
5232fa7b KH |
2133 | |
2134 | for (idx = 0; idx < len; idx++) | |
4ed46869 | 2135 | { |
5232fa7b | 2136 | Lisp_Object slot; |
4ed46869 | 2137 | |
64ef2921 | 2138 | slot = AREF (Vccl_program_table, idx); |
5232fa7b KH |
2139 | if (!VECTORP (slot)) |
2140 | /* This is the first unsed slot. Register NAME here. */ | |
4ed46869 KH |
2141 | break; |
2142 | ||
64ef2921 | 2143 | if (EQ (name, AREF (slot, 0))) |
4ed46869 | 2144 | { |
5232fa7b | 2145 | /* Update this slot. */ |
64ef2921 SM |
2146 | AREF (slot, 1) = ccl_prog; |
2147 | AREF (slot, 2) = resolved; | |
5232fa7b | 2148 | return make_number (idx); |
4ed46869 KH |
2149 | } |
2150 | } | |
2151 | ||
5232fa7b | 2152 | if (idx == len) |
4ed46869 | 2153 | { |
5232fa7b KH |
2154 | /* Extend the table. */ |
2155 | Lisp_Object new_table; | |
4ed46869 KH |
2156 | int j; |
2157 | ||
5232fa7b | 2158 | new_table = Fmake_vector (make_number (len * 2), Qnil); |
4ed46869 | 2159 | for (j = 0; j < len; j++) |
64ef2921 SM |
2160 | AREF (new_table, j) |
2161 | = AREF (Vccl_program_table, j); | |
4ed46869 KH |
2162 | Vccl_program_table = new_table; |
2163 | } | |
2164 | ||
5232fa7b KH |
2165 | { |
2166 | Lisp_Object elt; | |
2167 | ||
2168 | elt = Fmake_vector (make_number (3), Qnil); | |
64ef2921 SM |
2169 | AREF (elt, 0) = name; |
2170 | AREF (elt, 1) = ccl_prog; | |
2171 | AREF (elt, 2) = resolved; | |
2172 | AREF (Vccl_program_table, idx) = elt; | |
5232fa7b KH |
2173 | } |
2174 | ||
2175 | Fput (name, Qccl_program_idx, make_number (idx)); | |
2176 | return make_number (idx); | |
4ed46869 KH |
2177 | } |
2178 | ||
8146262a KH |
2179 | /* Register code conversion map. |
2180 | A code conversion map consists of numbers, Qt, Qnil, and Qlambda. | |
d617f6df DL |
2181 | The first element is the start code point. |
2182 | The other elements are mapped numbers. | |
8146262a KH |
2183 | Symbol t means to map to an original number before mapping. |
2184 | Symbol nil means that the corresponding element is empty. | |
d617f6df | 2185 | Symbol lambda means to terminate mapping here. |
e34b1164 KH |
2186 | */ |
2187 | ||
8146262a KH |
2188 | DEFUN ("register-code-conversion-map", Fregister_code_conversion_map, |
2189 | Sregister_code_conversion_map, | |
e34b1164 | 2190 | 2, 2, 0, |
fdb82f93 PJ |
2191 | doc: /* Register SYMBOL as code conversion map MAP. |
2192 | Return index number of the registered map. */) | |
2193 | (symbol, map) | |
8146262a | 2194 | Lisp_Object symbol, map; |
e34b1164 | 2195 | { |
64ef2921 | 2196 | int len = ASIZE (Vcode_conversion_map_vector); |
e34b1164 KH |
2197 | int i; |
2198 | Lisp_Object index; | |
2199 | ||
b7826503 PJ |
2200 | CHECK_SYMBOL (symbol); |
2201 | CHECK_VECTOR (map); | |
177c0ea7 | 2202 | |
e34b1164 KH |
2203 | for (i = 0; i < len; i++) |
2204 | { | |
64ef2921 | 2205 | Lisp_Object slot = AREF (Vcode_conversion_map_vector, i); |
e34b1164 KH |
2206 | |
2207 | if (!CONSP (slot)) | |
2208 | break; | |
2209 | ||
03699b14 | 2210 | if (EQ (symbol, XCAR (slot))) |
e34b1164 KH |
2211 | { |
2212 | index = make_number (i); | |
f3fbd155 | 2213 | XSETCDR (slot, map); |
8146262a KH |
2214 | Fput (symbol, Qcode_conversion_map, map); |
2215 | Fput (symbol, Qcode_conversion_map_id, index); | |
e34b1164 KH |
2216 | return index; |
2217 | } | |
2218 | } | |
2219 | ||
2220 | if (i == len) | |
2221 | { | |
2222 | Lisp_Object new_vector = Fmake_vector (make_number (len * 2), Qnil); | |
2223 | int j; | |
2224 | ||
2225 | for (j = 0; j < len; j++) | |
64ef2921 SM |
2226 | AREF (new_vector, j) |
2227 | = AREF (Vcode_conversion_map_vector, j); | |
8146262a | 2228 | Vcode_conversion_map_vector = new_vector; |
e34b1164 KH |
2229 | } |
2230 | ||
2231 | index = make_number (i); | |
8146262a KH |
2232 | Fput (symbol, Qcode_conversion_map, map); |
2233 | Fput (symbol, Qcode_conversion_map_id, index); | |
64ef2921 | 2234 | AREF (Vcode_conversion_map_vector, i) = Fcons (symbol, map); |
e34b1164 KH |
2235 | return index; |
2236 | } | |
2237 | ||
2238 | ||
dfcf069d | 2239 | void |
4ed46869 KH |
2240 | syms_of_ccl () |
2241 | { | |
2242 | staticpro (&Vccl_program_table); | |
6703ac4f | 2243 | Vccl_program_table = Fmake_vector (make_number (32), Qnil); |
4ed46869 | 2244 | |
c10842ea KH |
2245 | Qccl = intern ("ccl"); |
2246 | staticpro (&Qccl); | |
2247 | ||
2248 | Qcclp = intern ("cclp"); | |
2249 | staticpro (&Qcclp); | |
2250 | ||
6ae21908 KH |
2251 | Qccl_program = intern ("ccl-program"); |
2252 | staticpro (&Qccl_program); | |
2253 | ||
2254 | Qccl_program_idx = intern ("ccl-program-idx"); | |
2255 | staticpro (&Qccl_program_idx); | |
e34b1164 | 2256 | |
8146262a KH |
2257 | Qcode_conversion_map = intern ("code-conversion-map"); |
2258 | staticpro (&Qcode_conversion_map); | |
6ae21908 | 2259 | |
8146262a KH |
2260 | Qcode_conversion_map_id = intern ("code-conversion-map-id"); |
2261 | staticpro (&Qcode_conversion_map_id); | |
6ae21908 | 2262 | |
8146262a | 2263 | DEFVAR_LISP ("code-conversion-map-vector", &Vcode_conversion_map_vector, |
fdb82f93 | 2264 | doc: /* Vector of code conversion maps. */); |
8146262a | 2265 | Vcode_conversion_map_vector = Fmake_vector (make_number (16), Qnil); |
e34b1164 | 2266 | |
4ed46869 | 2267 | DEFVAR_LISP ("font-ccl-encoder-alist", &Vfont_ccl_encoder_alist, |
fdb82f93 PJ |
2268 | doc: /* Alist of fontname patterns vs corresponding CCL program. |
2269 | Each element looks like (REGEXP . CCL-CODE), | |
2270 | where CCL-CODE is a compiled CCL program. | |
2271 | When a font whose name matches REGEXP is used for displaying a character, | |
2272 | CCL-CODE is executed to calculate the code point in the font | |
2273 | from the charset number and position code(s) of the character which are set | |
2274 | in CCL registers R0, R1, and R2 before the execution. | |
2275 | The code point in the font is set in CCL registers R1 and R2 | |
2276 | when the execution terminated. | |
2277 | If the font is single-byte font, the register R2 is not used. */); | |
4ed46869 KH |
2278 | Vfont_ccl_encoder_alist = Qnil; |
2279 | ||
d80dc57e DL |
2280 | DEFVAR_LISP ("translation-hash-table-vector", &Vtranslation_hash_table_vector, |
2281 | doc: /* Vector containing all translation hash tables ever defined. | |
2282 | Comprises pairs (SYMBOL . TABLE) where SYMBOL and TABLE were set up by calls | |
2283 | to `define-translation-hash-table'. The vector is indexed by the table id | |
2284 | used by CCL. */); | |
2285 | Vtranslation_hash_table_vector = Qnil; | |
2286 | ||
5232fa7b | 2287 | defsubr (&Sccl_program_p); |
4ed46869 KH |
2288 | defsubr (&Sccl_execute); |
2289 | defsubr (&Sccl_execute_on_string); | |
2290 | defsubr (&Sregister_ccl_program); | |
8146262a | 2291 | defsubr (&Sregister_code_conversion_map); |
4ed46869 | 2292 | } |