Commit | Line | Data |
---|---|---|
4ed46869 | 1 | /* CCL (Code Conversion Language) interpreter. |
75c8c592 | 2 | Copyright (C) 1995, 1997 Electrotechnical Laboratory, JAPAN. |
d80dc57e | 3 | Copyright (C) 2001, 2002 Free Software Foundation, Inc. |
75c8c592 | 4 | Licensed to the Free Software Foundation. |
4ed46869 | 5 | |
369314dc KH |
6 | This file is part of GNU Emacs. |
7 | ||
8 | GNU Emacs is free software; you can redistribute it and/or modify | |
9 | it under the terms of the GNU General Public License as published by | |
10 | the Free Software Foundation; either version 2, or (at your option) | |
11 | any later version. | |
4ed46869 | 12 | |
369314dc KH |
13 | GNU Emacs is distributed in the hope that it will be useful, |
14 | but WITHOUT ANY WARRANTY; without even the implied warranty of | |
15 | MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the | |
16 | GNU General Public License for more details. | |
4ed46869 | 17 | |
369314dc KH |
18 | You should have received a copy of the GNU General Public License |
19 | along with GNU Emacs; see the file COPYING. If not, write to | |
20 | the Free Software Foundation, Inc., 59 Temple Place - Suite 330, | |
21 | Boston, MA 02111-1307, USA. */ | |
4ed46869 | 22 | |
4ed46869 | 23 | #ifdef emacs |
4ed46869 | 24 | #include <config.h> |
dfcf069d AS |
25 | #endif |
26 | ||
68c45bf0 PE |
27 | #include <stdio.h> |
28 | ||
29 | #ifdef emacs | |
30 | ||
4ed46869 KH |
31 | #include "lisp.h" |
32 | #include "charset.h" | |
33 | #include "ccl.h" | |
34 | #include "coding.h" | |
35 | ||
36 | #else /* not emacs */ | |
37 | ||
38 | #include "mulelib.h" | |
39 | ||
40 | #endif /* not emacs */ | |
41 | ||
20398ea4 | 42 | /* This contains all code conversion map available to CCL. */ |
8146262a | 43 | Lisp_Object Vcode_conversion_map_vector; |
e34b1164 | 44 | |
4ed46869 KH |
45 | /* Alist of fontname patterns vs corresponding CCL program. */ |
46 | Lisp_Object Vfont_ccl_encoder_alist; | |
47 | ||
6ae21908 KH |
48 | /* This symbol is a property which assocates with ccl program vector. |
49 | Ex: (get 'ccl-big5-encoder 'ccl-program) returns ccl program vector. */ | |
e34b1164 KH |
50 | Lisp_Object Qccl_program; |
51 | ||
8146262a KH |
52 | /* These symbols are properties which associate with code conversion |
53 | map and their ID respectively. */ | |
54 | Lisp_Object Qcode_conversion_map; | |
55 | Lisp_Object Qcode_conversion_map_id; | |
e34b1164 | 56 | |
6ae21908 KH |
57 | /* Symbols of ccl program have this property, a value of the property |
58 | is an index for Vccl_protram_table. */ | |
59 | Lisp_Object Qccl_program_idx; | |
60 | ||
5232fa7b KH |
61 | /* Table of registered CCL programs. Each element is a vector of |
62 | NAME, CCL_PROG, and RESOLVEDP where NAME (symbol) is the name of | |
63 | the program, CCL_PROG (vector) is the compiled code of the program, | |
64 | RESOLVEDP (t or nil) is the flag to tell if symbols in CCL_PROG is | |
65 | already resolved to index numbers or not. */ | |
4ed46869 KH |
66 | Lisp_Object Vccl_program_table; |
67 | ||
d80dc57e DL |
68 | /* Vector of registered hash tables for translation. */ |
69 | Lisp_Object Vtranslation_hash_table_vector; | |
70 | ||
71 | /* Return a hash table of id number ID. */ | |
72 | #define GET_HASH_TABLE(id) \ | |
73 | (XHASH_TABLE (XCDR(XVECTOR(Vtranslation_hash_table_vector)->contents[(id)]))) | |
d80dc57e | 74 | |
4ed46869 KH |
75 | /* CCL (Code Conversion Language) is a simple language which has |
76 | operations on one input buffer, one output buffer, and 7 registers. | |
77 | The syntax of CCL is described in `ccl.el'. Emacs Lisp function | |
78 | `ccl-compile' compiles a CCL program and produces a CCL code which | |
79 | is a vector of integers. The structure of this vector is as | |
80 | follows: The 1st element: buffer-magnification, a factor for the | |
81 | size of output buffer compared with the size of input buffer. The | |
82 | 2nd element: address of CCL code to be executed when encountered | |
83 | with end of input stream. The 3rd and the remaining elements: CCL | |
84 | codes. */ | |
85 | ||
86 | /* Header of CCL compiled code */ | |
87 | #define CCL_HEADER_BUF_MAG 0 | |
88 | #define CCL_HEADER_EOF 1 | |
89 | #define CCL_HEADER_MAIN 2 | |
90 | ||
91 | /* CCL code is a sequence of 28-bit non-negative integers (i.e. the | |
92 | MSB is always 0), each contains CCL command and/or arguments in the | |
93 | following format: | |
94 | ||
95 | |----------------- integer (28-bit) ------------------| | |
96 | |------- 17-bit ------|- 3-bit --|- 3-bit --|- 5-bit -| | |
97 | |--constant argument--|-register-|-register-|-command-| | |
98 | ccccccccccccccccc RRR rrr XXXXX | |
99 | or | |
100 | |------- relative address -------|-register-|-command-| | |
101 | cccccccccccccccccccc rrr XXXXX | |
102 | or | |
103 | |------------- constant or other args ----------------| | |
104 | cccccccccccccccccccccccccccc | |
105 | ||
106 | where, `cc...c' is a non-negative integer indicating constant value | |
107 | (the left most `c' is always 0) or an absolute jump address, `RRR' | |
108 | and `rrr' are CCL register number, `XXXXX' is one of the following | |
109 | CCL commands. */ | |
110 | ||
111 | /* CCL commands | |
112 | ||
113 | Each comment fields shows one or more lines for command syntax and | |
114 | the following lines for semantics of the command. In semantics, IC | |
115 | stands for Instruction Counter. */ | |
116 | ||
117 | #define CCL_SetRegister 0x00 /* Set register a register value: | |
118 | 1:00000000000000000RRRrrrXXXXX | |
119 | ------------------------------ | |
120 | reg[rrr] = reg[RRR]; | |
121 | */ | |
122 | ||
123 | #define CCL_SetShortConst 0x01 /* Set register a short constant value: | |
124 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
125 | ------------------------------ | |
126 | reg[rrr] = CCCCCCCCCCCCCCCCCCC; | |
127 | */ | |
128 | ||
129 | #define CCL_SetConst 0x02 /* Set register a constant value: | |
130 | 1:00000000000000000000rrrXXXXX | |
131 | 2:CONSTANT | |
132 | ------------------------------ | |
133 | reg[rrr] = CONSTANT; | |
134 | IC++; | |
135 | */ | |
136 | ||
137 | #define CCL_SetArray 0x03 /* Set register an element of array: | |
138 | 1:CCCCCCCCCCCCCCCCCRRRrrrXXXXX | |
139 | 2:ELEMENT[0] | |
140 | 3:ELEMENT[1] | |
141 | ... | |
142 | ------------------------------ | |
143 | if (0 <= reg[RRR] < CC..C) | |
144 | reg[rrr] = ELEMENT[reg[RRR]]; | |
145 | IC += CC..C; | |
146 | */ | |
147 | ||
148 | #define CCL_Jump 0x04 /* Jump: | |
149 | 1:A--D--D--R--E--S--S-000XXXXX | |
150 | ------------------------------ | |
151 | IC += ADDRESS; | |
152 | */ | |
153 | ||
154 | /* Note: If CC..C is greater than 0, the second code is omitted. */ | |
155 | ||
156 | #define CCL_JumpCond 0x05 /* Jump conditional: | |
157 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
158 | ------------------------------ | |
159 | if (!reg[rrr]) | |
160 | IC += ADDRESS; | |
161 | */ | |
162 | ||
163 | ||
164 | #define CCL_WriteRegisterJump 0x06 /* Write register and jump: | |
165 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
166 | ------------------------------ | |
167 | write (reg[rrr]); | |
168 | IC += ADDRESS; | |
169 | */ | |
170 | ||
171 | #define CCL_WriteRegisterReadJump 0x07 /* Write register, read, and jump: | |
172 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
173 | 2:A--D--D--R--E--S--S-rrrYYYYY | |
174 | ----------------------------- | |
175 | write (reg[rrr]); | |
176 | IC++; | |
177 | read (reg[rrr]); | |
178 | IC += ADDRESS; | |
179 | */ | |
180 | /* Note: If read is suspended, the resumed execution starts from the | |
181 | second code (YYYYY == CCL_ReadJump). */ | |
182 | ||
183 | #define CCL_WriteConstJump 0x08 /* Write constant and jump: | |
184 | 1:A--D--D--R--E--S--S-000XXXXX | |
185 | 2:CONST | |
186 | ------------------------------ | |
187 | write (CONST); | |
188 | IC += ADDRESS; | |
189 | */ | |
190 | ||
191 | #define CCL_WriteConstReadJump 0x09 /* Write constant, read, and jump: | |
192 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
193 | 2:CONST | |
194 | 3:A--D--D--R--E--S--S-rrrYYYYY | |
195 | ----------------------------- | |
196 | write (CONST); | |
197 | IC += 2; | |
198 | read (reg[rrr]); | |
199 | IC += ADDRESS; | |
200 | */ | |
201 | /* Note: If read is suspended, the resumed execution starts from the | |
202 | second code (YYYYY == CCL_ReadJump). */ | |
203 | ||
204 | #define CCL_WriteStringJump 0x0A /* Write string and jump: | |
205 | 1:A--D--D--R--E--S--S-000XXXXX | |
206 | 2:LENGTH | |
207 | 3:0000STRIN[0]STRIN[1]STRIN[2] | |
208 | ... | |
209 | ------------------------------ | |
210 | write_string (STRING, LENGTH); | |
211 | IC += ADDRESS; | |
212 | */ | |
213 | ||
214 | #define CCL_WriteArrayReadJump 0x0B /* Write an array element, read, and jump: | |
215 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
216 | 2:LENGTH | |
217 | 3:ELEMENET[0] | |
218 | 4:ELEMENET[1] | |
219 | ... | |
220 | N:A--D--D--R--E--S--S-rrrYYYYY | |
221 | ------------------------------ | |
222 | if (0 <= reg[rrr] < LENGTH) | |
223 | write (ELEMENT[reg[rrr]]); | |
224 | IC += LENGTH + 2; (... pointing at N+1) | |
225 | read (reg[rrr]); | |
226 | IC += ADDRESS; | |
227 | */ | |
228 | /* Note: If read is suspended, the resumed execution starts from the | |
887bfbd7 | 229 | Nth code (YYYYY == CCL_ReadJump). */ |
4ed46869 KH |
230 | |
231 | #define CCL_ReadJump 0x0C /* Read and jump: | |
232 | 1:A--D--D--R--E--S--S-rrrYYYYY | |
233 | ----------------------------- | |
234 | read (reg[rrr]); | |
235 | IC += ADDRESS; | |
236 | */ | |
237 | ||
238 | #define CCL_Branch 0x0D /* Jump by branch table: | |
239 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
240 | 2:A--D--D--R--E-S-S[0]000XXXXX | |
241 | 3:A--D--D--R--E-S-S[1]000XXXXX | |
242 | ... | |
243 | ------------------------------ | |
244 | if (0 <= reg[rrr] < CC..C) | |
245 | IC += ADDRESS[reg[rrr]]; | |
246 | else | |
247 | IC += ADDRESS[CC..C]; | |
248 | */ | |
249 | ||
250 | #define CCL_ReadRegister 0x0E /* Read bytes into registers: | |
251 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
252 | 2:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
253 | ... | |
254 | ------------------------------ | |
255 | while (CCC--) | |
256 | read (reg[rrr]); | |
257 | */ | |
258 | ||
259 | #define CCL_WriteExprConst 0x0F /* write result of expression: | |
260 | 1:00000OPERATION000RRR000XXXXX | |
261 | 2:CONSTANT | |
262 | ------------------------------ | |
263 | write (reg[RRR] OPERATION CONSTANT); | |
264 | IC++; | |
265 | */ | |
266 | ||
267 | /* Note: If the Nth read is suspended, the resumed execution starts | |
268 | from the Nth code. */ | |
269 | ||
270 | #define CCL_ReadBranch 0x10 /* Read one byte into a register, | |
271 | and jump by branch table: | |
272 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
273 | 2:A--D--D--R--E-S-S[0]000XXXXX | |
274 | 3:A--D--D--R--E-S-S[1]000XXXXX | |
275 | ... | |
276 | ------------------------------ | |
277 | read (read[rrr]); | |
278 | if (0 <= reg[rrr] < CC..C) | |
279 | IC += ADDRESS[reg[rrr]]; | |
280 | else | |
281 | IC += ADDRESS[CC..C]; | |
282 | */ | |
283 | ||
284 | #define CCL_WriteRegister 0x11 /* Write registers: | |
285 | 1:CCCCCCCCCCCCCCCCCCCrrrXXXXX | |
286 | 2:CCCCCCCCCCCCCCCCCCCrrrXXXXX | |
287 | ... | |
288 | ------------------------------ | |
289 | while (CCC--) | |
290 | write (reg[rrr]); | |
291 | ... | |
292 | */ | |
293 | ||
294 | /* Note: If the Nth write is suspended, the resumed execution | |
295 | starts from the Nth code. */ | |
296 | ||
297 | #define CCL_WriteExprRegister 0x12 /* Write result of expression | |
298 | 1:00000OPERATIONRrrRRR000XXXXX | |
299 | ------------------------------ | |
300 | write (reg[RRR] OPERATION reg[Rrr]); | |
301 | */ | |
302 | ||
e34b1164 | 303 | #define CCL_Call 0x13 /* Call the CCL program whose ID is |
5232fa7b KH |
304 | CC..C or cc..c. |
305 | 1:CCCCCCCCCCCCCCCCCCCCFFFXXXXX | |
306 | [2:00000000cccccccccccccccccccc] | |
4ed46869 | 307 | ------------------------------ |
5232fa7b KH |
308 | if (FFF) |
309 | call (cc..c) | |
310 | IC++; | |
311 | else | |
312 | call (CC..C) | |
4ed46869 KH |
313 | */ |
314 | ||
315 | #define CCL_WriteConstString 0x14 /* Write a constant or a string: | |
316 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
317 | [2:0000STRIN[0]STRIN[1]STRIN[2]] | |
318 | [...] | |
319 | ----------------------------- | |
320 | if (!rrr) | |
321 | write (CC..C) | |
322 | else | |
323 | write_string (STRING, CC..C); | |
324 | IC += (CC..C + 2) / 3; | |
325 | */ | |
326 | ||
327 | #define CCL_WriteArray 0x15 /* Write an element of array: | |
328 | 1:CCCCCCCCCCCCCCCCCCCCrrrXXXXX | |
329 | 2:ELEMENT[0] | |
330 | 3:ELEMENT[1] | |
331 | ... | |
332 | ------------------------------ | |
333 | if (0 <= reg[rrr] < CC..C) | |
334 | write (ELEMENT[reg[rrr]]); | |
335 | IC += CC..C; | |
336 | */ | |
337 | ||
338 | #define CCL_End 0x16 /* Terminate: | |
339 | 1:00000000000000000000000XXXXX | |
340 | ------------------------------ | |
341 | terminate (); | |
342 | */ | |
343 | ||
344 | /* The following two codes execute an assignment arithmetic/logical | |
345 | operation. The form of the operation is like REG OP= OPERAND. */ | |
346 | ||
347 | #define CCL_ExprSelfConst 0x17 /* REG OP= constant: | |
348 | 1:00000OPERATION000000rrrXXXXX | |
349 | 2:CONSTANT | |
350 | ------------------------------ | |
351 | reg[rrr] OPERATION= CONSTANT; | |
352 | */ | |
353 | ||
354 | #define CCL_ExprSelfReg 0x18 /* REG1 OP= REG2: | |
355 | 1:00000OPERATION000RRRrrrXXXXX | |
356 | ------------------------------ | |
357 | reg[rrr] OPERATION= reg[RRR]; | |
358 | */ | |
359 | ||
360 | /* The following codes execute an arithmetic/logical operation. The | |
361 | form of the operation is like REG_X = REG_Y OP OPERAND2. */ | |
362 | ||
363 | #define CCL_SetExprConst 0x19 /* REG_X = REG_Y OP constant: | |
364 | 1:00000OPERATION000RRRrrrXXXXX | |
365 | 2:CONSTANT | |
366 | ------------------------------ | |
367 | reg[rrr] = reg[RRR] OPERATION CONSTANT; | |
368 | IC++; | |
369 | */ | |
370 | ||
371 | #define CCL_SetExprReg 0x1A /* REG1 = REG2 OP REG3: | |
372 | 1:00000OPERATIONRrrRRRrrrXXXXX | |
373 | ------------------------------ | |
374 | reg[rrr] = reg[RRR] OPERATION reg[Rrr]; | |
375 | */ | |
376 | ||
377 | #define CCL_JumpCondExprConst 0x1B /* Jump conditional according to | |
378 | an operation on constant: | |
379 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
380 | 2:OPERATION | |
381 | 3:CONSTANT | |
382 | ----------------------------- | |
383 | reg[7] = reg[rrr] OPERATION CONSTANT; | |
384 | if (!(reg[7])) | |
385 | IC += ADDRESS; | |
386 | else | |
387 | IC += 2 | |
388 | */ | |
389 | ||
390 | #define CCL_JumpCondExprReg 0x1C /* Jump conditional according to | |
391 | an operation on register: | |
392 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
393 | 2:OPERATION | |
394 | 3:RRR | |
395 | ----------------------------- | |
396 | reg[7] = reg[rrr] OPERATION reg[RRR]; | |
397 | if (!reg[7]) | |
398 | IC += ADDRESS; | |
399 | else | |
400 | IC += 2; | |
401 | */ | |
402 | ||
403 | #define CCL_ReadJumpCondExprConst 0x1D /* Read and jump conditional according | |
404 | to an operation on constant: | |
405 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
406 | 2:OPERATION | |
407 | 3:CONSTANT | |
408 | ----------------------------- | |
409 | read (reg[rrr]); | |
410 | reg[7] = reg[rrr] OPERATION CONSTANT; | |
411 | if (!reg[7]) | |
412 | IC += ADDRESS; | |
413 | else | |
414 | IC += 2; | |
415 | */ | |
416 | ||
417 | #define CCL_ReadJumpCondExprReg 0x1E /* Read and jump conditional according | |
418 | to an operation on register: | |
419 | 1:A--D--D--R--E--S--S-rrrXXXXX | |
420 | 2:OPERATION | |
421 | 3:RRR | |
422 | ----------------------------- | |
423 | read (reg[rrr]); | |
424 | reg[7] = reg[rrr] OPERATION reg[RRR]; | |
425 | if (!reg[7]) | |
426 | IC += ADDRESS; | |
427 | else | |
428 | IC += 2; | |
429 | */ | |
430 | ||
450ed226 | 431 | #define CCL_Extension 0x1F /* Extended CCL code |
4ed46869 KH |
432 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX |
433 | 2:ARGUEMENT | |
434 | 3:... | |
435 | ------------------------------ | |
436 | extended_command (rrr,RRR,Rrr,ARGS) | |
437 | */ | |
438 | ||
e34b1164 | 439 | /* |
6ae21908 | 440 | Here after, Extended CCL Instructions. |
e34b1164 | 441 | Bit length of extended command is 14. |
6ae21908 | 442 | Therefore, the instruction code range is 0..16384(0x3fff). |
e34b1164 KH |
443 | */ |
444 | ||
6ae21908 KH |
445 | /* Read a multibyte characeter. |
446 | A code point is stored into reg[rrr]. A charset ID is stored into | |
447 | reg[RRR]. */ | |
448 | ||
449 | #define CCL_ReadMultibyteChar2 0x00 /* Read Multibyte Character | |
450 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX */ | |
451 | ||
452 | /* Write a multibyte character. | |
453 | Write a character whose code point is reg[rrr] and the charset ID | |
454 | is reg[RRR]. */ | |
455 | ||
456 | #define CCL_WriteMultibyteChar2 0x01 /* Write Multibyte Character | |
457 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX */ | |
458 | ||
8146262a | 459 | /* Translate a character whose code point is reg[rrr] and the charset |
f967223b | 460 | ID is reg[RRR] by a translation table whose ID is reg[Rrr]. |
6ae21908 | 461 | |
8146262a | 462 | A translated character is set in reg[rrr] (code point) and reg[RRR] |
6ae21908 KH |
463 | (charset ID). */ |
464 | ||
8146262a | 465 | #define CCL_TranslateCharacter 0x02 /* Translate a multibyte character |
6ae21908 KH |
466 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX */ |
467 | ||
8146262a | 468 | /* Translate a character whose code point is reg[rrr] and the charset |
f967223b | 469 | ID is reg[RRR] by a translation table whose ID is ARGUMENT. |
6ae21908 | 470 | |
8146262a | 471 | A translated character is set in reg[rrr] (code point) and reg[RRR] |
6ae21908 KH |
472 | (charset ID). */ |
473 | ||
8146262a KH |
474 | #define CCL_TranslateCharacterConstTbl 0x03 /* Translate a multibyte character |
475 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX | |
476 | 2:ARGUMENT(Translation Table ID) | |
477 | */ | |
6ae21908 | 478 | |
8146262a KH |
479 | /* Iterate looking up MAPs for reg[rrr] starting from the Nth (N = |
480 | reg[RRR]) MAP until some value is found. | |
6ae21908 | 481 | |
8146262a | 482 | Each MAP is a Lisp vector whose element is number, nil, t, or |
6ae21908 | 483 | lambda. |
8146262a | 484 | If the element is nil, ignore the map and proceed to the next map. |
6ae21908 KH |
485 | If the element is t or lambda, finish without changing reg[rrr]. |
486 | If the element is a number, set reg[rrr] to the number and finish. | |
487 | ||
8146262a KH |
488 | Detail of the map structure is descibed in the comment for |
489 | CCL_MapMultiple below. */ | |
6ae21908 | 490 | |
8146262a | 491 | #define CCL_IterateMultipleMap 0x10 /* Iterate multiple maps |
6ae21908 | 492 | 1:ExtendedCOMMNDXXXRRRrrrXXXXX |
8146262a KH |
493 | 2:NUMBER of MAPs |
494 | 3:MAP-ID1 | |
495 | 4:MAP-ID2 | |
6ae21908 KH |
496 | ... |
497 | */ | |
498 | ||
8146262a KH |
499 | /* Map the code in reg[rrr] by MAPs starting from the Nth (N = |
500 | reg[RRR]) map. | |
6ae21908 | 501 | |
9b27b20d | 502 | MAPs are supplied in the succeeding CCL codes as follows: |
6ae21908 | 503 | |
8146262a KH |
504 | When CCL program gives this nested structure of map to this command: |
505 | ((MAP-ID11 | |
506 | MAP-ID12 | |
507 | (MAP-ID121 MAP-ID122 MAP-ID123) | |
508 | MAP-ID13) | |
509 | (MAP-ID21 | |
510 | (MAP-ID211 (MAP-ID2111) MAP-ID212) | |
511 | MAP-ID22)), | |
6ae21908 | 512 | the compiled CCL codes has this sequence: |
8146262a | 513 | CCL_MapMultiple (CCL code of this command) |
9b27b20d KH |
514 | 16 (total number of MAPs and SEPARATORs) |
515 | -7 (1st SEPARATOR) | |
8146262a KH |
516 | MAP-ID11 |
517 | MAP-ID12 | |
9b27b20d | 518 | -3 (2nd SEPARATOR) |
8146262a KH |
519 | MAP-ID121 |
520 | MAP-ID122 | |
521 | MAP-ID123 | |
522 | MAP-ID13 | |
9b27b20d | 523 | -7 (3rd SEPARATOR) |
8146262a | 524 | MAP-ID21 |
9b27b20d | 525 | -4 (4th SEPARATOR) |
8146262a | 526 | MAP-ID211 |
9b27b20d | 527 | -1 (5th SEPARATOR) |
8146262a KH |
528 | MAP_ID2111 |
529 | MAP-ID212 | |
530 | MAP-ID22 | |
6ae21908 | 531 | |
9b27b20d | 532 | A value of each SEPARATOR follows this rule: |
8146262a KH |
533 | MAP-SET := SEPARATOR [(MAP-ID | MAP-SET)]+ |
534 | SEPARATOR := -(number of MAP-IDs and SEPARATORs in the MAP-SET) | |
6ae21908 | 535 | |
8146262a | 536 | (*)....Nest level of MAP-SET must not be over than MAX_MAP_SET_LEVEL. |
6ae21908 | 537 | |
8146262a KH |
538 | When some map fails to map (i.e. it doesn't have a value for |
539 | reg[rrr]), the mapping is treated as identity. | |
6ae21908 | 540 | |
8146262a | 541 | The mapping is iterated for all maps in each map set (set of maps |
9b27b20d KH |
542 | separated by SEPARATOR) except in the case that lambda is |
543 | encountered. More precisely, the mapping proceeds as below: | |
544 | ||
545 | At first, VAL0 is set to reg[rrr], and it is translated by the | |
546 | first map to VAL1. Then, VAL1 is translated by the next map to | |
547 | VAL2. This mapping is iterated until the last map is used. The | |
54fa5bc1 KH |
548 | result of the mapping is the last value of VAL?. When the mapping |
549 | process reached to the end of the map set, it moves to the next | |
550 | map set. If the next does not exit, the mapping process terminates, | |
551 | and regard the last value as a result. | |
9b27b20d KH |
552 | |
553 | But, when VALm is mapped to VALn and VALn is not a number, the | |
554 | mapping proceed as below: | |
555 | ||
556 | If VALn is nil, the lastest map is ignored and the mapping of VALm | |
557 | proceed to the next map. | |
558 | ||
559 | In VALn is t, VALm is reverted to reg[rrr] and the mapping of VALm | |
560 | proceed to the next map. | |
561 | ||
54fa5bc1 KH |
562 | If VALn is lambda, move to the next map set like reaching to the |
563 | end of the current map set. | |
564 | ||
565 | If VALn is a symbol, call the CCL program refered by it. | |
566 | Then, use reg[rrr] as a mapped value except for -1, -2 and -3. | |
567 | Such special values are regarded as nil, t, and lambda respectively. | |
6ae21908 | 568 | |
8146262a | 569 | Each map is a Lisp vector of the following format (a) or (b): |
6ae21908 KH |
570 | (a)......[STARTPOINT VAL1 VAL2 ...] |
571 | (b)......[t VAL STARTPOINT ENDPOINT], | |
572 | where | |
8146262a | 573 | STARTPOINT is an offset to be used for indexing a map, |
9b27b20d | 574 | ENDPOINT is a maximum index number of a map, |
6ae21908 KH |
575 | VAL and VALn is a number, nil, t, or lambda. |
576 | ||
8146262a KH |
577 | Valid index range of a map of type (a) is: |
578 | STARTPOINT <= index < STARTPOINT + map_size - 1 | |
579 | Valid index range of a map of type (b) is: | |
9b27b20d | 580 | STARTPOINT <= index < ENDPOINT */ |
6ae21908 | 581 | |
8146262a | 582 | #define CCL_MapMultiple 0x11 /* Mapping by multiple code conversion maps |
6ae21908 KH |
583 | 1:ExtendedCOMMNDXXXRRRrrrXXXXX |
584 | 2:N-2 | |
585 | 3:SEPARATOR_1 (< 0) | |
8146262a KH |
586 | 4:MAP-ID_1 |
587 | 5:MAP-ID_2 | |
6ae21908 KH |
588 | ... |
589 | M:SEPARATOR_x (< 0) | |
8146262a | 590 | M+1:MAP-ID_y |
6ae21908 KH |
591 | ... |
592 | N:SEPARATOR_z (< 0) | |
593 | */ | |
594 | ||
54fa5bc1 | 595 | #define MAX_MAP_SET_LEVEL 30 |
6ae21908 KH |
596 | |
597 | typedef struct | |
598 | { | |
599 | int rest_length; | |
600 | int orig_val; | |
601 | } tr_stack; | |
602 | ||
8146262a KH |
603 | static tr_stack mapping_stack[MAX_MAP_SET_LEVEL]; |
604 | static tr_stack *mapping_stack_pointer; | |
6ae21908 | 605 | |
54fa5bc1 KH |
606 | /* If this variable is non-zero, it indicates the stack_idx |
607 | of immediately called by CCL_MapMultiple. */ | |
be57900b | 608 | static int stack_idx_of_map_multiple; |
54fa5bc1 KH |
609 | |
610 | #define PUSH_MAPPING_STACK(restlen, orig) \ | |
a89f435d PJ |
611 | do \ |
612 | { \ | |
54fa5bc1 KH |
613 | mapping_stack_pointer->rest_length = (restlen); \ |
614 | mapping_stack_pointer->orig_val = (orig); \ | |
615 | mapping_stack_pointer++; \ | |
a89f435d PJ |
616 | } \ |
617 | while (0) | |
54fa5bc1 KH |
618 | |
619 | #define POP_MAPPING_STACK(restlen, orig) \ | |
a89f435d PJ |
620 | do \ |
621 | { \ | |
54fa5bc1 KH |
622 | mapping_stack_pointer--; \ |
623 | (restlen) = mapping_stack_pointer->rest_length; \ | |
624 | (orig) = mapping_stack_pointer->orig_val; \ | |
a89f435d PJ |
625 | } \ |
626 | while (0) | |
6ae21908 | 627 | |
54fa5bc1 | 628 | #define CCL_CALL_FOR_MAP_INSTRUCTION(symbol, ret_ic) \ |
a89f435d | 629 | do \ |
0ee1088b | 630 | { \ |
54fa5bc1 KH |
631 | struct ccl_program called_ccl; \ |
632 | if (stack_idx >= 256 \ | |
633 | || (setup_ccl_program (&called_ccl, (symbol)) != 0)) \ | |
634 | { \ | |
635 | if (stack_idx > 0) \ | |
636 | { \ | |
637 | ccl_prog = ccl_prog_stack_struct[0].ccl_prog; \ | |
638 | ic = ccl_prog_stack_struct[0].ic; \ | |
639 | } \ | |
640 | CCL_INVALID_CMD; \ | |
641 | } \ | |
642 | ccl_prog_stack_struct[stack_idx].ccl_prog = ccl_prog; \ | |
643 | ccl_prog_stack_struct[stack_idx].ic = (ret_ic); \ | |
644 | stack_idx++; \ | |
645 | ccl_prog = called_ccl.prog; \ | |
646 | ic = CCL_HEADER_MAIN; \ | |
647 | goto ccl_repeat; \ | |
0ee1088b | 648 | } \ |
a89f435d | 649 | while (0) |
6ae21908 | 650 | |
8146262a | 651 | #define CCL_MapSingle 0x12 /* Map by single code conversion map |
6ae21908 | 652 | 1:ExtendedCOMMNDXXXRRRrrrXXXXX |
8146262a | 653 | 2:MAP-ID |
6ae21908 | 654 | ------------------------------ |
8146262a KH |
655 | Map reg[rrr] by MAP-ID. |
656 | If some valid mapping is found, | |
6ae21908 KH |
657 | set reg[rrr] to the result, |
658 | else | |
659 | set reg[RRR] to -1. | |
660 | */ | |
4ed46869 | 661 | |
d80dc57e DL |
662 | #define CCL_LookupIntConstTbl 0x13 /* Lookup multibyte character by |
663 | integer key. Afterwards R7 set | |
664 | to 1 iff lookup succeeded. | |
665 | 1:ExtendedCOMMNDRrrRRRXXXXXXXX | |
666 | 2:ARGUMENT(Hash table ID) */ | |
667 | ||
668 | #define CCL_LookupCharConstTbl 0x14 /* Lookup integer by multibyte | |
669 | character key. Afterwards R7 set | |
670 | to 1 iff lookup succeeded. | |
671 | 1:ExtendedCOMMNDRrrRRRrrrXXXXX | |
672 | 2:ARGUMENT(Hash table ID) */ | |
673 | ||
4ed46869 KH |
674 | /* CCL arithmetic/logical operators. */ |
675 | #define CCL_PLUS 0x00 /* X = Y + Z */ | |
676 | #define CCL_MINUS 0x01 /* X = Y - Z */ | |
677 | #define CCL_MUL 0x02 /* X = Y * Z */ | |
678 | #define CCL_DIV 0x03 /* X = Y / Z */ | |
679 | #define CCL_MOD 0x04 /* X = Y % Z */ | |
680 | #define CCL_AND 0x05 /* X = Y & Z */ | |
681 | #define CCL_OR 0x06 /* X = Y | Z */ | |
682 | #define CCL_XOR 0x07 /* X = Y ^ Z */ | |
683 | #define CCL_LSH 0x08 /* X = Y << Z */ | |
684 | #define CCL_RSH 0x09 /* X = Y >> Z */ | |
685 | #define CCL_LSH8 0x0A /* X = (Y << 8) | Z */ | |
686 | #define CCL_RSH8 0x0B /* X = Y >> 8, r[7] = Y & 0xFF */ | |
687 | #define CCL_DIVMOD 0x0C /* X = Y / Z, r[7] = Y % Z */ | |
688 | #define CCL_LS 0x10 /* X = (X < Y) */ | |
689 | #define CCL_GT 0x11 /* X = (X > Y) */ | |
690 | #define CCL_EQ 0x12 /* X = (X == Y) */ | |
691 | #define CCL_LE 0x13 /* X = (X <= Y) */ | |
692 | #define CCL_GE 0x14 /* X = (X >= Y) */ | |
693 | #define CCL_NE 0x15 /* X = (X != Y) */ | |
694 | ||
51520e8a | 695 | #define CCL_DECODE_SJIS 0x16 /* X = HIGHER_BYTE (DE-SJIS (Y, Z)) |
4ed46869 | 696 | r[7] = LOWER_BYTE (DE-SJIS (Y, Z)) */ |
51520e8a KH |
697 | #define CCL_ENCODE_SJIS 0x17 /* X = HIGHER_BYTE (SJIS (Y, Z)) |
698 | r[7] = LOWER_BYTE (SJIS (Y, Z) */ | |
4ed46869 | 699 | |
4ed46869 | 700 | /* Terminate CCL program successfully. */ |
0ee1088b | 701 | #define CCL_SUCCESS \ |
a89f435d | 702 | do \ |
0ee1088b | 703 | { \ |
4ed46869 | 704 | ccl->status = CCL_STAT_SUCCESS; \ |
0ee1088b KH |
705 | goto ccl_finish; \ |
706 | } \ | |
a89f435d | 707 | while(0) |
4ed46869 KH |
708 | |
709 | /* Suspend CCL program because of reading from empty input buffer or | |
710 | writing to full output buffer. When this program is resumed, the | |
711 | same I/O command is executed. */ | |
e34b1164 | 712 | #define CCL_SUSPEND(stat) \ |
a89f435d | 713 | do \ |
0ee1088b | 714 | { \ |
e34b1164 KH |
715 | ic--; \ |
716 | ccl->status = stat; \ | |
717 | goto ccl_finish; \ | |
0ee1088b | 718 | } \ |
a89f435d | 719 | while (0) |
4ed46869 KH |
720 | |
721 | /* Terminate CCL program because of invalid command. Should not occur | |
722 | in the normal case. */ | |
723 | #define CCL_INVALID_CMD \ | |
a89f435d | 724 | do \ |
0ee1088b | 725 | { \ |
4ed46869 KH |
726 | ccl->status = CCL_STAT_INVALID_CMD; \ |
727 | goto ccl_error_handler; \ | |
0ee1088b | 728 | } \ |
a89f435d | 729 | while(0) |
4ed46869 KH |
730 | |
731 | /* Encode one character CH to multibyte form and write to the current | |
887bfbd7 | 732 | output buffer. If CH is less than 256, CH is written as is. */ |
a37520c6 KH |
733 | #define CCL_WRITE_CHAR(ch) \ |
734 | do { \ | |
735 | int bytes = SINGLE_BYTE_CHAR_P (ch) ? 1: CHAR_BYTES (ch); \ | |
736 | if (!dst) \ | |
737 | CCL_INVALID_CMD; \ | |
738 | else if (dst + bytes + extra_bytes < (dst_bytes ? dst_end : src)) \ | |
739 | { \ | |
740 | if (bytes == 1) \ | |
741 | { \ | |
742 | *dst++ = (ch); \ | |
743 | if ((ch) >= 0x80 && (ch) < 0xA0) \ | |
744 | /* We may have to convert this eight-bit char to \ | |
745 | multibyte form later. */ \ | |
746 | extra_bytes++; \ | |
747 | } \ | |
31165028 | 748 | else if (CHAR_VALID_P (ch, 0)) \ |
a37520c6 | 749 | dst += CHAR_STRING (ch, dst); \ |
31165028 KH |
750 | else \ |
751 | CCL_INVALID_CMD; \ | |
a37520c6 KH |
752 | } \ |
753 | else \ | |
754 | CCL_SUSPEND (CCL_STAT_SUSPEND_BY_DST); \ | |
4ed46869 KH |
755 | } while (0) |
756 | ||
a8302ba3 KH |
757 | /* Encode one character CH to multibyte form and write to the current |
758 | output buffer. The output bytes always forms a valid multibyte | |
759 | sequence. */ | |
760 | #define CCL_WRITE_MULTIBYTE_CHAR(ch) \ | |
761 | do { \ | |
762 | int bytes = CHAR_BYTES (ch); \ | |
763 | if (!dst) \ | |
764 | CCL_INVALID_CMD; \ | |
765 | else if (dst + bytes + extra_bytes < (dst_bytes ? dst_end : src)) \ | |
766 | { \ | |
767 | if (CHAR_VALID_P ((ch), 0)) \ | |
768 | dst += CHAR_STRING ((ch), dst); \ | |
769 | else \ | |
770 | CCL_INVALID_CMD; \ | |
771 | } \ | |
772 | else \ | |
773 | CCL_SUSPEND (CCL_STAT_SUSPEND_BY_DST); \ | |
774 | } while (0) | |
775 | ||
4ed46869 KH |
776 | /* Write a string at ccl_prog[IC] of length LEN to the current output |
777 | buffer. */ | |
778 | #define CCL_WRITE_STRING(len) \ | |
779 | do { \ | |
780 | if (!dst) \ | |
781 | CCL_INVALID_CMD; \ | |
e34b1164 | 782 | else if (dst + len <= (dst_bytes ? dst_end : src)) \ |
4ed46869 KH |
783 | for (i = 0; i < len; i++) \ |
784 | *dst++ = ((XFASTINT (ccl_prog[ic + (i / 3)])) \ | |
785 | >> ((2 - (i % 3)) * 8)) & 0xFF; \ | |
786 | else \ | |
e34b1164 | 787 | CCL_SUSPEND (CCL_STAT_SUSPEND_BY_DST); \ |
4ed46869 KH |
788 | } while (0) |
789 | ||
9977c491 KH |
790 | /* Read one byte from the current input buffer into REGth register. */ |
791 | #define CCL_READ_CHAR(REG) \ | |
17312e44 KH |
792 | do { \ |
793 | if (!src) \ | |
794 | CCL_INVALID_CMD; \ | |
795 | else if (src < src_end) \ | |
796 | { \ | |
9977c491 KH |
797 | REG = *src++; \ |
798 | if (REG == '\n' \ | |
17312e44 KH |
799 | && ccl->eol_type != CODING_EOL_LF) \ |
800 | { \ | |
801 | /* We are encoding. */ \ | |
802 | if (ccl->eol_type == CODING_EOL_CRLF) \ | |
803 | { \ | |
804 | if (ccl->cr_consumed) \ | |
805 | ccl->cr_consumed = 0; \ | |
806 | else \ | |
807 | { \ | |
808 | ccl->cr_consumed = 1; \ | |
9977c491 | 809 | REG = '\r'; \ |
17312e44 KH |
810 | src--; \ |
811 | } \ | |
812 | } \ | |
813 | else \ | |
9977c491 | 814 | REG = '\r'; \ |
17312e44 | 815 | } \ |
9977c491 | 816 | if (REG == LEADING_CODE_8_BIT_CONTROL \ |
17312e44 | 817 | && ccl->multibyte) \ |
9977c491 | 818 | REG = *src++ - 0x20; \ |
17312e44 KH |
819 | } \ |
820 | else if (ccl->last_block) \ | |
821 | { \ | |
822 | ic = ccl->eof_ic; \ | |
823 | goto ccl_repeat; \ | |
824 | } \ | |
825 | else \ | |
826 | CCL_SUSPEND (CCL_STAT_SUSPEND_BY_SRC); \ | |
4ed46869 KH |
827 | } while (0) |
828 | ||
829 | ||
4ffd4870 KH |
830 | /* Set C to the character code made from CHARSET and CODE. This is |
831 | like MAKE_CHAR but check the validity of CHARSET and CODE. If they | |
832 | are not valid, set C to (CODE & 0xFF) because that is usually the | |
833 | case that CCL_ReadMultibyteChar2 read an invalid code and it set | |
834 | CODE to that invalid byte. */ | |
835 | ||
836 | #define CCL_MAKE_CHAR(charset, code, c) \ | |
837 | do { \ | |
838 | if (charset == CHARSET_ASCII) \ | |
839 | c = code & 0xFF; \ | |
840 | else if (CHARSET_DEFINED_P (charset) \ | |
841 | && (code & 0x7F) >= 32 \ | |
842 | && (code < 256 || ((code >> 7) & 0x7F) >= 32)) \ | |
843 | { \ | |
844 | int c1 = code & 0x7F, c2 = 0; \ | |
845 | \ | |
846 | if (code >= 256) \ | |
847 | c2 = c1, c1 = (code >> 7) & 0x7F; \ | |
bd045987 | 848 | c = MAKE_CHAR (charset, c1, c2); \ |
4ffd4870 KH |
849 | } \ |
850 | else \ | |
bd045987 | 851 | c = code & 0xFF; \ |
4ffd4870 KH |
852 | } while (0) |
853 | ||
854 | ||
4ed46869 KH |
855 | /* Execute CCL code on SRC_BYTES length text at SOURCE. The resulting |
856 | text goes to a place pointed by DESTINATION, the length of which | |
857 | should not exceed DST_BYTES. The bytes actually processed is | |
858 | returned as *CONSUMED. The return value is the length of the | |
859 | resulting text. As a side effect, the contents of CCL registers | |
860 | are updated. If SOURCE or DESTINATION is NULL, only operations on | |
861 | registers are permitted. */ | |
862 | ||
863 | #ifdef CCL_DEBUG | |
864 | #define CCL_DEBUG_BACKTRACE_LEN 256 | |
865 | int ccl_backtrace_table[CCL_BACKTRACE_TABLE]; | |
866 | int ccl_backtrace_idx; | |
867 | #endif | |
868 | ||
869 | struct ccl_prog_stack | |
870 | { | |
a9f1cc19 | 871 | Lisp_Object *ccl_prog; /* Pointer to an array of CCL code. */ |
4ed46869 KH |
872 | int ic; /* Instruction Counter. */ |
873 | }; | |
874 | ||
c13362d8 KH |
875 | /* For the moment, we only support depth 256 of stack. */ |
876 | static struct ccl_prog_stack ccl_prog_stack_struct[256]; | |
877 | ||
dfcf069d | 878 | int |
4ed46869 KH |
879 | ccl_driver (ccl, source, destination, src_bytes, dst_bytes, consumed) |
880 | struct ccl_program *ccl; | |
881 | unsigned char *source, *destination; | |
882 | int src_bytes, dst_bytes; | |
883 | int *consumed; | |
884 | { | |
885 | register int *reg = ccl->reg; | |
886 | register int ic = ccl->ic; | |
8a1ae4dd | 887 | register int code = 0, field1, field2; |
e995085f | 888 | register Lisp_Object *ccl_prog = ccl->prog; |
4ed46869 KH |
889 | unsigned char *src = source, *src_end = src + src_bytes; |
890 | unsigned char *dst = destination, *dst_end = dst + dst_bytes; | |
891 | int jump_address; | |
8a1ae4dd | 892 | int i = 0, j, op; |
c13362d8 | 893 | int stack_idx = ccl->stack_idx; |
519bf146 | 894 | /* Instruction counter of the current CCL code. */ |
8a1ae4dd | 895 | int this_ic = 0; |
a37520c6 KH |
896 | /* CCL_WRITE_CHAR will produce 8-bit code of range 0x80..0x9F. But, |
897 | each of them will be converted to multibyte form of 2-byte | |
898 | sequence. For that conversion, we remember how many more bytes | |
899 | we must keep in DESTINATION in this variable. */ | |
900 | int extra_bytes = 0; | |
4ed46869 KH |
901 | |
902 | if (ic >= ccl->eof_ic) | |
903 | ic = CCL_HEADER_MAIN; | |
904 | ||
8a1ae4dd | 905 | if (ccl->buf_magnification == 0) /* We can't produce any bytes. */ |
12abd7d1 KH |
906 | dst = NULL; |
907 | ||
54fa5bc1 KH |
908 | /* Set mapping stack pointer. */ |
909 | mapping_stack_pointer = mapping_stack; | |
910 | ||
4ed46869 KH |
911 | #ifdef CCL_DEBUG |
912 | ccl_backtrace_idx = 0; | |
913 | #endif | |
914 | ||
915 | for (;;) | |
916 | { | |
4ccd0d4a | 917 | ccl_repeat: |
4ed46869 KH |
918 | #ifdef CCL_DEBUG |
919 | ccl_backtrace_table[ccl_backtrace_idx++] = ic; | |
920 | if (ccl_backtrace_idx >= CCL_DEBUG_BACKTRACE_LEN) | |
921 | ccl_backtrace_idx = 0; | |
922 | ccl_backtrace_table[ccl_backtrace_idx] = 0; | |
923 | #endif | |
924 | ||
925 | if (!NILP (Vquit_flag) && NILP (Vinhibit_quit)) | |
926 | { | |
927 | /* We can't just signal Qquit, instead break the loop as if | |
928 | the whole data is processed. Don't reset Vquit_flag, it | |
929 | must be handled later at a safer place. */ | |
930 | if (consumed) | |
931 | src = source + src_bytes; | |
932 | ccl->status = CCL_STAT_QUIT; | |
933 | break; | |
934 | } | |
935 | ||
519bf146 | 936 | this_ic = ic; |
4ed46869 KH |
937 | code = XINT (ccl_prog[ic]); ic++; |
938 | field1 = code >> 8; | |
939 | field2 = (code & 0xFF) >> 5; | |
940 | ||
941 | #define rrr field2 | |
942 | #define RRR (field1 & 7) | |
943 | #define Rrr ((field1 >> 3) & 7) | |
944 | #define ADDR field1 | |
e34b1164 | 945 | #define EXCMD (field1 >> 6) |
4ed46869 KH |
946 | |
947 | switch (code & 0x1F) | |
948 | { | |
949 | case CCL_SetRegister: /* 00000000000000000RRRrrrXXXXX */ | |
950 | reg[rrr] = reg[RRR]; | |
951 | break; | |
952 | ||
953 | case CCL_SetShortConst: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
954 | reg[rrr] = field1; | |
955 | break; | |
956 | ||
957 | case CCL_SetConst: /* 00000000000000000000rrrXXXXX */ | |
958 | reg[rrr] = XINT (ccl_prog[ic]); | |
959 | ic++; | |
960 | break; | |
961 | ||
962 | case CCL_SetArray: /* CCCCCCCCCCCCCCCCCCCCRRRrrrXXXXX */ | |
963 | i = reg[RRR]; | |
964 | j = field1 >> 3; | |
965 | if ((unsigned int) i < j) | |
966 | reg[rrr] = XINT (ccl_prog[ic + i]); | |
967 | ic += j; | |
968 | break; | |
969 | ||
970 | case CCL_Jump: /* A--D--D--R--E--S--S-000XXXXX */ | |
971 | ic += ADDR; | |
972 | break; | |
973 | ||
974 | case CCL_JumpCond: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
975 | if (!reg[rrr]) | |
976 | ic += ADDR; | |
977 | break; | |
978 | ||
979 | case CCL_WriteRegisterJump: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
980 | i = reg[rrr]; | |
981 | CCL_WRITE_CHAR (i); | |
982 | ic += ADDR; | |
983 | break; | |
984 | ||
985 | case CCL_WriteRegisterReadJump: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
986 | i = reg[rrr]; | |
987 | CCL_WRITE_CHAR (i); | |
988 | ic++; | |
989 | CCL_READ_CHAR (reg[rrr]); | |
990 | ic += ADDR - 1; | |
991 | break; | |
992 | ||
993 | case CCL_WriteConstJump: /* A--D--D--R--E--S--S-000XXXXX */ | |
994 | i = XINT (ccl_prog[ic]); | |
995 | CCL_WRITE_CHAR (i); | |
996 | ic += ADDR; | |
997 | break; | |
998 | ||
999 | case CCL_WriteConstReadJump: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
1000 | i = XINT (ccl_prog[ic]); | |
1001 | CCL_WRITE_CHAR (i); | |
1002 | ic++; | |
1003 | CCL_READ_CHAR (reg[rrr]); | |
1004 | ic += ADDR - 1; | |
1005 | break; | |
1006 | ||
1007 | case CCL_WriteStringJump: /* A--D--D--R--E--S--S-000XXXXX */ | |
1008 | j = XINT (ccl_prog[ic]); | |
1009 | ic++; | |
1010 | CCL_WRITE_STRING (j); | |
1011 | ic += ADDR - 1; | |
1012 | break; | |
1013 | ||
1014 | case CCL_WriteArrayReadJump: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
1015 | i = reg[rrr]; | |
2e34157c | 1016 | j = XINT (ccl_prog[ic]); |
4ed46869 KH |
1017 | if ((unsigned int) i < j) |
1018 | { | |
887bfbd7 | 1019 | i = XINT (ccl_prog[ic + 1 + i]); |
4ed46869 KH |
1020 | CCL_WRITE_CHAR (i); |
1021 | } | |
887bfbd7 | 1022 | ic += j + 2; |
4ed46869 KH |
1023 | CCL_READ_CHAR (reg[rrr]); |
1024 | ic += ADDR - (j + 2); | |
1025 | break; | |
1026 | ||
1027 | case CCL_ReadJump: /* A--D--D--R--E--S--S-rrrYYYYY */ | |
1028 | CCL_READ_CHAR (reg[rrr]); | |
1029 | ic += ADDR; | |
1030 | break; | |
1031 | ||
1032 | case CCL_ReadBranch: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
1033 | CCL_READ_CHAR (reg[rrr]); | |
1034 | /* fall through ... */ | |
1035 | case CCL_Branch: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
1036 | if ((unsigned int) reg[rrr] < field1) | |
1037 | ic += XINT (ccl_prog[ic + reg[rrr]]); | |
1038 | else | |
1039 | ic += XINT (ccl_prog[ic + field1]); | |
1040 | break; | |
1041 | ||
1042 | case CCL_ReadRegister: /* CCCCCCCCCCCCCCCCCCCCrrXXXXX */ | |
1043 | while (1) | |
1044 | { | |
1045 | CCL_READ_CHAR (reg[rrr]); | |
1046 | if (!field1) break; | |
1047 | code = XINT (ccl_prog[ic]); ic++; | |
1048 | field1 = code >> 8; | |
1049 | field2 = (code & 0xFF) >> 5; | |
1050 | } | |
1051 | break; | |
1052 | ||
1053 | case CCL_WriteExprConst: /* 1:00000OPERATION000RRR000XXXXX */ | |
1054 | rrr = 7; | |
1055 | i = reg[RRR]; | |
1056 | j = XINT (ccl_prog[ic]); | |
1057 | op = field1 >> 6; | |
25660570 | 1058 | jump_address = ic + 1; |
4ed46869 KH |
1059 | goto ccl_set_expr; |
1060 | ||
1061 | case CCL_WriteRegister: /* CCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
1062 | while (1) | |
1063 | { | |
1064 | i = reg[rrr]; | |
1065 | CCL_WRITE_CHAR (i); | |
1066 | if (!field1) break; | |
1067 | code = XINT (ccl_prog[ic]); ic++; | |
1068 | field1 = code >> 8; | |
1069 | field2 = (code & 0xFF) >> 5; | |
1070 | } | |
1071 | break; | |
1072 | ||
1073 | case CCL_WriteExprRegister: /* 1:00000OPERATIONRrrRRR000XXXXX */ | |
1074 | rrr = 7; | |
1075 | i = reg[RRR]; | |
1076 | j = reg[Rrr]; | |
1077 | op = field1 >> 6; | |
25660570 | 1078 | jump_address = ic; |
4ed46869 KH |
1079 | goto ccl_set_expr; |
1080 | ||
5232fa7b | 1081 | case CCL_Call: /* 1:CCCCCCCCCCCCCCCCCCCCFFFXXXXX */ |
4ed46869 KH |
1082 | { |
1083 | Lisp_Object slot; | |
5232fa7b KH |
1084 | int prog_id; |
1085 | ||
1086 | /* If FFF is nonzero, the CCL program ID is in the | |
1087 | following code. */ | |
1088 | if (rrr) | |
1089 | { | |
1090 | prog_id = XINT (ccl_prog[ic]); | |
1091 | ic++; | |
1092 | } | |
1093 | else | |
1094 | prog_id = field1; | |
4ed46869 KH |
1095 | |
1096 | if (stack_idx >= 256 | |
5232fa7b | 1097 | || prog_id < 0 |
64ef2921 SM |
1098 | || prog_id >= ASIZE (Vccl_program_table) |
1099 | || (slot = AREF (Vccl_program_table, prog_id), !VECTORP (slot)) | |
1100 | || !VECTORP (AREF (slot, 1))) | |
4ed46869 KH |
1101 | { |
1102 | if (stack_idx > 0) | |
1103 | { | |
1104 | ccl_prog = ccl_prog_stack_struct[0].ccl_prog; | |
1105 | ic = ccl_prog_stack_struct[0].ic; | |
1106 | } | |
1107 | CCL_INVALID_CMD; | |
1108 | } | |
1109 | ||
1110 | ccl_prog_stack_struct[stack_idx].ccl_prog = ccl_prog; | |
1111 | ccl_prog_stack_struct[stack_idx].ic = ic; | |
1112 | stack_idx++; | |
64ef2921 | 1113 | ccl_prog = XVECTOR (AREF (slot, 1))->contents; |
4ed46869 KH |
1114 | ic = CCL_HEADER_MAIN; |
1115 | } | |
1116 | break; | |
1117 | ||
1118 | case CCL_WriteConstString: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
1119 | if (!rrr) | |
1120 | CCL_WRITE_CHAR (field1); | |
1121 | else | |
1122 | { | |
1123 | CCL_WRITE_STRING (field1); | |
1124 | ic += (field1 + 2) / 3; | |
1125 | } | |
1126 | break; | |
1127 | ||
1128 | case CCL_WriteArray: /* CCCCCCCCCCCCCCCCCCCCrrrXXXXX */ | |
1129 | i = reg[rrr]; | |
1130 | if ((unsigned int) i < field1) | |
1131 | { | |
1132 | j = XINT (ccl_prog[ic + i]); | |
1133 | CCL_WRITE_CHAR (j); | |
1134 | } | |
1135 | ic += field1; | |
1136 | break; | |
1137 | ||
1138 | case CCL_End: /* 0000000000000000000000XXXXX */ | |
d3a478e2 | 1139 | if (stack_idx > 0) |
4ed46869 | 1140 | { |
d3a478e2 | 1141 | stack_idx--; |
4ed46869 KH |
1142 | ccl_prog = ccl_prog_stack_struct[stack_idx].ccl_prog; |
1143 | ic = ccl_prog_stack_struct[stack_idx].ic; | |
1144 | break; | |
1145 | } | |
ad3d1b1d KH |
1146 | if (src) |
1147 | src = src_end; | |
1148 | /* ccl->ic should points to this command code again to | |
1149 | suppress further processing. */ | |
1150 | ic--; | |
4ed46869 KH |
1151 | CCL_SUCCESS; |
1152 | ||
1153 | case CCL_ExprSelfConst: /* 00000OPERATION000000rrrXXXXX */ | |
1154 | i = XINT (ccl_prog[ic]); | |
1155 | ic++; | |
1156 | op = field1 >> 6; | |
1157 | goto ccl_expr_self; | |
1158 | ||
1159 | case CCL_ExprSelfReg: /* 00000OPERATION000RRRrrrXXXXX */ | |
1160 | i = reg[RRR]; | |
1161 | op = field1 >> 6; | |
1162 | ||
1163 | ccl_expr_self: | |
1164 | switch (op) | |
1165 | { | |
1166 | case CCL_PLUS: reg[rrr] += i; break; | |
1167 | case CCL_MINUS: reg[rrr] -= i; break; | |
1168 | case CCL_MUL: reg[rrr] *= i; break; | |
1169 | case CCL_DIV: reg[rrr] /= i; break; | |
1170 | case CCL_MOD: reg[rrr] %= i; break; | |
1171 | case CCL_AND: reg[rrr] &= i; break; | |
1172 | case CCL_OR: reg[rrr] |= i; break; | |
1173 | case CCL_XOR: reg[rrr] ^= i; break; | |
1174 | case CCL_LSH: reg[rrr] <<= i; break; | |
1175 | case CCL_RSH: reg[rrr] >>= i; break; | |
1176 | case CCL_LSH8: reg[rrr] <<= 8; reg[rrr] |= i; break; | |
1177 | case CCL_RSH8: reg[7] = reg[rrr] & 0xFF; reg[rrr] >>= 8; break; | |
1178 | case CCL_DIVMOD: reg[7] = reg[rrr] % i; reg[rrr] /= i; break; | |
1179 | case CCL_LS: reg[rrr] = reg[rrr] < i; break; | |
1180 | case CCL_GT: reg[rrr] = reg[rrr] > i; break; | |
1181 | case CCL_EQ: reg[rrr] = reg[rrr] == i; break; | |
1182 | case CCL_LE: reg[rrr] = reg[rrr] <= i; break; | |
1183 | case CCL_GE: reg[rrr] = reg[rrr] >= i; break; | |
1184 | case CCL_NE: reg[rrr] = reg[rrr] != i; break; | |
1185 | default: CCL_INVALID_CMD; | |
1186 | } | |
1187 | break; | |
1188 | ||
1189 | case CCL_SetExprConst: /* 00000OPERATION000RRRrrrXXXXX */ | |
1190 | i = reg[RRR]; | |
1191 | j = XINT (ccl_prog[ic]); | |
1192 | op = field1 >> 6; | |
1193 | jump_address = ++ic; | |
1194 | goto ccl_set_expr; | |
1195 | ||
1196 | case CCL_SetExprReg: /* 00000OPERATIONRrrRRRrrrXXXXX */ | |
1197 | i = reg[RRR]; | |
1198 | j = reg[Rrr]; | |
1199 | op = field1 >> 6; | |
1200 | jump_address = ic; | |
1201 | goto ccl_set_expr; | |
1202 | ||
1203 | case CCL_ReadJumpCondExprConst: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
1204 | CCL_READ_CHAR (reg[rrr]); | |
1205 | case CCL_JumpCondExprConst: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
1206 | i = reg[rrr]; | |
1207 | op = XINT (ccl_prog[ic]); | |
1208 | jump_address = ic++ + ADDR; | |
1209 | j = XINT (ccl_prog[ic]); | |
1210 | ic++; | |
1211 | rrr = 7; | |
1212 | goto ccl_set_expr; | |
1213 | ||
1214 | case CCL_ReadJumpCondExprReg: /* A--D--D--R--E--S--S-rrrXXXXX */ | |
1215 | CCL_READ_CHAR (reg[rrr]); | |
1216 | case CCL_JumpCondExprReg: | |
1217 | i = reg[rrr]; | |
1218 | op = XINT (ccl_prog[ic]); | |
1219 | jump_address = ic++ + ADDR; | |
1220 | j = reg[XINT (ccl_prog[ic])]; | |
1221 | ic++; | |
1222 | rrr = 7; | |
1223 | ||
1224 | ccl_set_expr: | |
1225 | switch (op) | |
1226 | { | |
1227 | case CCL_PLUS: reg[rrr] = i + j; break; | |
1228 | case CCL_MINUS: reg[rrr] = i - j; break; | |
1229 | case CCL_MUL: reg[rrr] = i * j; break; | |
1230 | case CCL_DIV: reg[rrr] = i / j; break; | |
1231 | case CCL_MOD: reg[rrr] = i % j; break; | |
1232 | case CCL_AND: reg[rrr] = i & j; break; | |
1233 | case CCL_OR: reg[rrr] = i | j; break; | |
1234 | case CCL_XOR: reg[rrr] = i ^ j;; break; | |
1235 | case CCL_LSH: reg[rrr] = i << j; break; | |
1236 | case CCL_RSH: reg[rrr] = i >> j; break; | |
1237 | case CCL_LSH8: reg[rrr] = (i << 8) | j; break; | |
1238 | case CCL_RSH8: reg[rrr] = i >> 8; reg[7] = i & 0xFF; break; | |
1239 | case CCL_DIVMOD: reg[rrr] = i / j; reg[7] = i % j; break; | |
1240 | case CCL_LS: reg[rrr] = i < j; break; | |
1241 | case CCL_GT: reg[rrr] = i > j; break; | |
1242 | case CCL_EQ: reg[rrr] = i == j; break; | |
1243 | case CCL_LE: reg[rrr] = i <= j; break; | |
1244 | case CCL_GE: reg[rrr] = i >= j; break; | |
1245 | case CCL_NE: reg[rrr] = i != j; break; | |
4ed46869 | 1246 | case CCL_DECODE_SJIS: DECODE_SJIS (i, j, reg[rrr], reg[7]); break; |
51520e8a | 1247 | case CCL_ENCODE_SJIS: ENCODE_SJIS (i, j, reg[rrr], reg[7]); break; |
4ed46869 KH |
1248 | default: CCL_INVALID_CMD; |
1249 | } | |
1250 | code &= 0x1F; | |
1251 | if (code == CCL_WriteExprConst || code == CCL_WriteExprRegister) | |
1252 | { | |
1253 | i = reg[rrr]; | |
1254 | CCL_WRITE_CHAR (i); | |
25660570 | 1255 | ic = jump_address; |
4ed46869 KH |
1256 | } |
1257 | else if (!reg[rrr]) | |
1258 | ic = jump_address; | |
1259 | break; | |
1260 | ||
450ed226 | 1261 | case CCL_Extension: |
e34b1164 KH |
1262 | switch (EXCMD) |
1263 | { | |
6ae21908 | 1264 | case CCL_ReadMultibyteChar2: |
e34b1164 KH |
1265 | if (!src) |
1266 | CCL_INVALID_CMD; | |
60768428 | 1267 | |
0ee1088b KH |
1268 | if (src >= src_end) |
1269 | { | |
1270 | src++; | |
1271 | goto ccl_read_multibyte_character_suspend; | |
1272 | } | |
e34b1164 | 1273 | |
38b9ed6a KH |
1274 | if (!ccl->multibyte) |
1275 | { | |
1276 | int bytes; | |
1277 | if (!UNIBYTE_STR_AS_MULTIBYTE_P (src, src_end - src, bytes)) | |
1278 | { | |
1279 | reg[RRR] = CHARSET_8_BIT_CONTROL; | |
1280 | reg[rrr] = *src++; | |
1281 | break; | |
1282 | } | |
1283 | } | |
0ee1088b KH |
1284 | i = *src++; |
1285 | if (i == '\n' && ccl->eol_type != CODING_EOL_LF) | |
1286 | { | |
1287 | /* We are encoding. */ | |
1288 | if (ccl->eol_type == CODING_EOL_CRLF) | |
1289 | { | |
1290 | if (ccl->cr_consumed) | |
1291 | ccl->cr_consumed = 0; | |
1292 | else | |
1293 | { | |
1294 | ccl->cr_consumed = 1; | |
1295 | i = '\r'; | |
1296 | src--; | |
1297 | } | |
1298 | } | |
1299 | else | |
1300 | i = '\r'; | |
1301 | reg[rrr] = i; | |
1302 | reg[RRR] = CHARSET_ASCII; | |
1303 | } | |
1304 | else if (i < 0x80) | |
1305 | { | |
1306 | /* ASCII */ | |
1307 | reg[rrr] = i; | |
1308 | reg[RRR] = CHARSET_ASCII; | |
1309 | } | |
0ee1088b KH |
1310 | else if (i <= MAX_CHARSET_OFFICIAL_DIMENSION2) |
1311 | { | |
0fc71a77 KH |
1312 | int dimension = BYTES_BY_CHAR_HEAD (i) - 1; |
1313 | ||
1314 | if (dimension == 0) | |
1315 | { | |
1316 | /* `i' is a leading code for an undefined charset. */ | |
1317 | reg[RRR] = CHARSET_8_BIT_GRAPHIC; | |
1318 | reg[rrr] = i; | |
1319 | } | |
1320 | else if (src + dimension > src_end) | |
0ee1088b | 1321 | goto ccl_read_multibyte_character_suspend; |
0fc71a77 KH |
1322 | else |
1323 | { | |
1324 | reg[RRR] = i; | |
1325 | i = (*src++ & 0x7F); | |
1326 | if (dimension == 1) | |
1327 | reg[rrr] = i; | |
1328 | else | |
1329 | reg[rrr] = ((i << 7) | (*src++ & 0x7F)); | |
1330 | } | |
0ee1088b KH |
1331 | } |
1332 | else if ((i == LEADING_CODE_PRIVATE_11) | |
1333 | || (i == LEADING_CODE_PRIVATE_12)) | |
1334 | { | |
1335 | if ((src + 1) >= src_end) | |
1336 | goto ccl_read_multibyte_character_suspend; | |
1337 | reg[RRR] = *src++; | |
1338 | reg[rrr] = (*src++ & 0x7F); | |
1339 | } | |
1340 | else if ((i == LEADING_CODE_PRIVATE_21) | |
1341 | || (i == LEADING_CODE_PRIVATE_22)) | |
1342 | { | |
1343 | if ((src + 2) >= src_end) | |
1344 | goto ccl_read_multibyte_character_suspend; | |
1345 | reg[RRR] = *src++; | |
1346 | i = (*src++ & 0x7F); | |
1347 | reg[rrr] = ((i << 7) | (*src & 0x7F)); | |
1348 | src++; | |
1349 | } | |
1350 | else if (i == LEADING_CODE_8_BIT_CONTROL) | |
1351 | { | |
1352 | if (src >= src_end) | |
1353 | goto ccl_read_multibyte_character_suspend; | |
1354 | reg[RRR] = CHARSET_8_BIT_CONTROL; | |
1355 | reg[rrr] = (*src++ - 0x20); | |
1356 | } | |
1357 | else if (i >= 0xA0) | |
1358 | { | |
1359 | reg[RRR] = CHARSET_8_BIT_GRAPHIC; | |
1360 | reg[rrr] = i; | |
1361 | } | |
1362 | else | |
1363 | { | |
1364 | /* INVALID CODE. Return a single byte character. */ | |
1365 | reg[RRR] = CHARSET_ASCII; | |
1366 | reg[rrr] = i; | |
1367 | } | |
e34b1164 KH |
1368 | break; |
1369 | ||
1370 | ccl_read_multibyte_character_suspend: | |
38b9ed6a KH |
1371 | if (src <= src_end && !ccl->multibyte && ccl->last_block) |
1372 | { | |
1373 | reg[RRR] = CHARSET_8_BIT_CONTROL; | |
1374 | reg[rrr] = i; | |
1375 | break; | |
1376 | } | |
e34b1164 KH |
1377 | src--; |
1378 | if (ccl->last_block) | |
1379 | { | |
1380 | ic = ccl->eof_ic; | |
0db078dc | 1381 | goto ccl_repeat; |
e34b1164 KH |
1382 | } |
1383 | else | |
1384 | CCL_SUSPEND (CCL_STAT_SUSPEND_BY_SRC); | |
1385 | ||
1386 | break; | |
1387 | ||
6ae21908 | 1388 | case CCL_WriteMultibyteChar2: |
e34b1164 | 1389 | i = reg[RRR]; /* charset */ |
5c464c4d KH |
1390 | if (i == CHARSET_ASCII |
1391 | || i == CHARSET_8_BIT_CONTROL | |
1392 | || i == CHARSET_8_BIT_GRAPHIC) | |
c13362d8 | 1393 | i = reg[rrr] & 0xFF; |
e34b1164 KH |
1394 | else if (CHARSET_DIMENSION (i) == 1) |
1395 | i = ((i - 0x70) << 7) | (reg[rrr] & 0x7F); | |
1396 | else if (i < MIN_CHARSET_PRIVATE_DIMENSION2) | |
1397 | i = ((i - 0x8F) << 14) | reg[rrr]; | |
1398 | else | |
1399 | i = ((i - 0xE0) << 14) | reg[rrr]; | |
1400 | ||
a8302ba3 | 1401 | CCL_WRITE_MULTIBYTE_CHAR (i); |
e34b1164 KH |
1402 | |
1403 | break; | |
1404 | ||
8146262a | 1405 | case CCL_TranslateCharacter: |
4ffd4870 | 1406 | CCL_MAKE_CHAR (reg[RRR], reg[rrr], i); |
8146262a KH |
1407 | op = translate_char (GET_TRANSLATION_TABLE (reg[Rrr]), |
1408 | i, -1, 0, 0); | |
e34b1164 KH |
1409 | SPLIT_CHAR (op, reg[RRR], i, j); |
1410 | if (j != -1) | |
1411 | i = (i << 7) | j; | |
1412 | ||
1413 | reg[rrr] = i; | |
1414 | break; | |
1415 | ||
8146262a | 1416 | case CCL_TranslateCharacterConstTbl: |
e34b1164 KH |
1417 | op = XINT (ccl_prog[ic]); /* table */ |
1418 | ic++; | |
4ffd4870 | 1419 | CCL_MAKE_CHAR (reg[RRR], reg[rrr], i); |
8146262a | 1420 | op = translate_char (GET_TRANSLATION_TABLE (op), i, -1, 0, 0); |
e34b1164 KH |
1421 | SPLIT_CHAR (op, reg[RRR], i, j); |
1422 | if (j != -1) | |
1423 | i = (i << 7) | j; | |
1424 | ||
1425 | reg[rrr] = i; | |
1426 | break; | |
1427 | ||
d80dc57e DL |
1428 | case CCL_LookupIntConstTbl: |
1429 | op = XINT (ccl_prog[ic]); /* table */ | |
1430 | ic++; | |
1431 | { | |
1432 | struct Lisp_Hash_Table *h = GET_HASH_TABLE (op); | |
1433 | ||
1434 | op = hash_lookup (h, make_number (reg[RRR]), NULL); | |
1435 | if (op >= 0) | |
1436 | { | |
1437 | op = HASH_VALUE (h, op); | |
1438 | if (!CHAR_VALID_P (op, 0)) | |
1439 | CCL_INVALID_CMD; | |
1440 | SPLIT_CHAR (XINT (op), reg[RRR], i, j); | |
1441 | if (j != -1) | |
1442 | i = (i << 7) | j; | |
1443 | reg[rrr] = i; | |
1444 | reg[7] = 1; /* r7 true for success */ | |
1445 | } | |
1446 | else | |
1447 | reg[7] = 0; | |
1448 | } | |
1449 | break; | |
1450 | ||
1451 | case CCL_LookupCharConstTbl: | |
1452 | op = XINT (ccl_prog[ic]); /* table */ | |
1453 | ic++; | |
1454 | CCL_MAKE_CHAR (reg[RRR], reg[rrr], i); | |
1455 | { | |
1456 | struct Lisp_Hash_Table *h = GET_HASH_TABLE (op); | |
1457 | ||
1458 | op = hash_lookup (h, make_number (i), NULL); | |
1459 | if (op >= 0) | |
1460 | { | |
1461 | op = HASH_VALUE (h, op); | |
1462 | if (!INTEGERP (op)) | |
1463 | CCL_INVALID_CMD; | |
1464 | reg[RRR] = XINT (op); | |
1465 | reg[7] = 1; /* r7 true for success */ | |
1466 | } | |
1467 | else | |
1468 | reg[7] = 0; | |
1469 | } | |
1470 | break; | |
1471 | ||
e34b1164 KH |
1472 | case CCL_IterateMultipleMap: |
1473 | { | |
8146262a | 1474 | Lisp_Object map, content, attrib, value; |
e34b1164 KH |
1475 | int point, size, fin_ic; |
1476 | ||
8146262a | 1477 | j = XINT (ccl_prog[ic++]); /* number of maps. */ |
e34b1164 KH |
1478 | fin_ic = ic + j; |
1479 | op = reg[rrr]; | |
1480 | if ((j > reg[RRR]) && (j >= 0)) | |
1481 | { | |
1482 | ic += reg[RRR]; | |
1483 | i = reg[RRR]; | |
1484 | } | |
1485 | else | |
1486 | { | |
1487 | reg[RRR] = -1; | |
1488 | ic = fin_ic; | |
1489 | break; | |
1490 | } | |
1491 | ||
1492 | for (;i < j;i++) | |
1493 | { | |
1494 | ||
64ef2921 | 1495 | size = ASIZE (Vcode_conversion_map_vector); |
d387866a | 1496 | point = XINT (ccl_prog[ic++]); |
e34b1164 | 1497 | if (point >= size) continue; |
64ef2921 | 1498 | map = AREF (Vcode_conversion_map_vector, point); |
8146262a KH |
1499 | |
1500 | /* Check map varidity. */ | |
1501 | if (!CONSP (map)) continue; | |
03699b14 | 1502 | map = XCDR (map); |
8146262a | 1503 | if (!VECTORP (map)) continue; |
64ef2921 | 1504 | size = ASIZE (map); |
e34b1164 | 1505 | if (size <= 1) continue; |
6ae21908 | 1506 | |
64ef2921 | 1507 | content = AREF (map, 0); |
6ae21908 | 1508 | |
8146262a | 1509 | /* check map type, |
6ae21908 KH |
1510 | [STARTPOINT VAL1 VAL2 ...] or |
1511 | [t ELELMENT STARTPOINT ENDPOINT] */ | |
1512 | if (NUMBERP (content)) | |
1513 | { | |
1514 | point = XUINT (content); | |
1515 | point = op - point + 1; | |
1516 | if (!((point >= 1) && (point < size))) continue; | |
64ef2921 | 1517 | content = AREF (map, point); |
6ae21908 KH |
1518 | } |
1519 | else if (EQ (content, Qt)) | |
1520 | { | |
1521 | if (size != 4) continue; | |
64ef2921 SM |
1522 | if ((op >= XUINT (AREF (map, 2))) |
1523 | && (op < XUINT (AREF (map, 3)))) | |
1524 | content = AREF (map, 1); | |
6ae21908 KH |
1525 | else |
1526 | continue; | |
1527 | } | |
1528 | else | |
1529 | continue; | |
e34b1164 KH |
1530 | |
1531 | if (NILP (content)) | |
1532 | continue; | |
1533 | else if (NUMBERP (content)) | |
1534 | { | |
1535 | reg[RRR] = i; | |
6ae21908 | 1536 | reg[rrr] = XINT(content); |
e34b1164 KH |
1537 | break; |
1538 | } | |
1539 | else if (EQ (content, Qt) || EQ (content, Qlambda)) | |
1540 | { | |
1541 | reg[RRR] = i; | |
1542 | break; | |
1543 | } | |
1544 | else if (CONSP (content)) | |
1545 | { | |
03699b14 KR |
1546 | attrib = XCAR (content); |
1547 | value = XCDR (content); | |
e34b1164 KH |
1548 | if (!NUMBERP (attrib) || !NUMBERP (value)) |
1549 | continue; | |
1550 | reg[RRR] = i; | |
6ae21908 | 1551 | reg[rrr] = XUINT (value); |
e34b1164 KH |
1552 | break; |
1553 | } | |
54fa5bc1 KH |
1554 | else if (SYMBOLP (content)) |
1555 | CCL_CALL_FOR_MAP_INSTRUCTION (content, fin_ic); | |
1556 | else | |
1557 | CCL_INVALID_CMD; | |
e34b1164 KH |
1558 | } |
1559 | if (i == j) | |
1560 | reg[RRR] = -1; | |
1561 | ic = fin_ic; | |
1562 | } | |
1563 | break; | |
1564 | ||
8146262a | 1565 | case CCL_MapMultiple: |
e34b1164 | 1566 | { |
8146262a KH |
1567 | Lisp_Object map, content, attrib, value; |
1568 | int point, size, map_vector_size; | |
1569 | int map_set_rest_length, fin_ic; | |
54fa5bc1 KH |
1570 | int current_ic = this_ic; |
1571 | ||
1572 | /* inhibit recursive call on MapMultiple. */ | |
1573 | if (stack_idx_of_map_multiple > 0) | |
1574 | { | |
1575 | if (stack_idx_of_map_multiple <= stack_idx) | |
1576 | { | |
1577 | stack_idx_of_map_multiple = 0; | |
1578 | mapping_stack_pointer = mapping_stack; | |
1579 | CCL_INVALID_CMD; | |
1580 | } | |
1581 | } | |
1582 | else | |
1583 | mapping_stack_pointer = mapping_stack; | |
1584 | stack_idx_of_map_multiple = 0; | |
8146262a KH |
1585 | |
1586 | map_set_rest_length = | |
1587 | XINT (ccl_prog[ic++]); /* number of maps and separators. */ | |
1588 | fin_ic = ic + map_set_rest_length; | |
54fa5bc1 KH |
1589 | op = reg[rrr]; |
1590 | ||
8146262a | 1591 | if ((map_set_rest_length > reg[RRR]) && (reg[RRR] >= 0)) |
e34b1164 KH |
1592 | { |
1593 | ic += reg[RRR]; | |
1594 | i = reg[RRR]; | |
8146262a | 1595 | map_set_rest_length -= i; |
e34b1164 KH |
1596 | } |
1597 | else | |
1598 | { | |
1599 | ic = fin_ic; | |
1600 | reg[RRR] = -1; | |
54fa5bc1 | 1601 | mapping_stack_pointer = mapping_stack; |
e34b1164 KH |
1602 | break; |
1603 | } | |
6ae21908 | 1604 | |
54fa5bc1 KH |
1605 | if (mapping_stack_pointer <= (mapping_stack + 1)) |
1606 | { | |
1607 | /* Set up initial state. */ | |
1608 | mapping_stack_pointer = mapping_stack; | |
1609 | PUSH_MAPPING_STACK (0, op); | |
1610 | reg[RRR] = -1; | |
1611 | } | |
1612 | else | |
1613 | { | |
1614 | /* Recover after calling other ccl program. */ | |
1615 | int orig_op; | |
e34b1164 | 1616 | |
54fa5bc1 KH |
1617 | POP_MAPPING_STACK (map_set_rest_length, orig_op); |
1618 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1619 | switch (op) | |
e34b1164 | 1620 | { |
54fa5bc1 KH |
1621 | case -1: |
1622 | /* Regard it as Qnil. */ | |
1623 | op = orig_op; | |
1624 | i++; | |
1625 | ic++; | |
1626 | map_set_rest_length--; | |
1627 | break; | |
1628 | case -2: | |
1629 | /* Regard it as Qt. */ | |
e34b1164 | 1630 | op = reg[rrr]; |
54fa5bc1 KH |
1631 | i++; |
1632 | ic++; | |
1633 | map_set_rest_length--; | |
1634 | break; | |
1635 | case -3: | |
1636 | /* Regard it as Qlambda. */ | |
1637 | op = orig_op; | |
1638 | i += map_set_rest_length; | |
1639 | ic += map_set_rest_length; | |
1640 | map_set_rest_length = 0; | |
1641 | break; | |
1642 | default: | |
1643 | /* Regard it as normal mapping. */ | |
8146262a | 1644 | i += map_set_rest_length; |
54fa5bc1 | 1645 | ic += map_set_rest_length; |
8146262a | 1646 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); |
6ae21908 KH |
1647 | break; |
1648 | } | |
e34b1164 | 1649 | } |
64ef2921 | 1650 | map_vector_size = ASIZE (Vcode_conversion_map_vector); |
54fa5bc1 KH |
1651 | |
1652 | do { | |
1653 | for (;map_set_rest_length > 0;i++, ic++, map_set_rest_length--) | |
1654 | { | |
1655 | point = XINT(ccl_prog[ic]); | |
1656 | if (point < 0) | |
1657 | { | |
1658 | /* +1 is for including separator. */ | |
1659 | point = -point + 1; | |
1660 | if (mapping_stack_pointer | |
1661 | >= &mapping_stack[MAX_MAP_SET_LEVEL]) | |
1662 | CCL_INVALID_CMD; | |
1663 | PUSH_MAPPING_STACK (map_set_rest_length - point, | |
1664 | reg[rrr]); | |
1665 | map_set_rest_length = point; | |
1666 | reg[rrr] = op; | |
1667 | continue; | |
1668 | } | |
1669 | ||
1670 | if (point >= map_vector_size) continue; | |
64ef2921 | 1671 | map = AREF (Vcode_conversion_map_vector, point); |
54fa5bc1 KH |
1672 | |
1673 | /* Check map varidity. */ | |
1674 | if (!CONSP (map)) continue; | |
1675 | map = XCDR (map); | |
1676 | if (!VECTORP (map)) continue; | |
64ef2921 | 1677 | size = ASIZE (map); |
54fa5bc1 KH |
1678 | if (size <= 1) continue; |
1679 | ||
64ef2921 | 1680 | content = AREF (map, 0); |
54fa5bc1 KH |
1681 | |
1682 | /* check map type, | |
1683 | [STARTPOINT VAL1 VAL2 ...] or | |
1684 | [t ELEMENT STARTPOINT ENDPOINT] */ | |
1685 | if (NUMBERP (content)) | |
1686 | { | |
1687 | point = XUINT (content); | |
1688 | point = op - point + 1; | |
1689 | if (!((point >= 1) && (point < size))) continue; | |
64ef2921 | 1690 | content = AREF (map, point); |
54fa5bc1 KH |
1691 | } |
1692 | else if (EQ (content, Qt)) | |
1693 | { | |
1694 | if (size != 4) continue; | |
64ef2921 SM |
1695 | if ((op >= XUINT (AREF (map, 2))) && |
1696 | (op < XUINT (AREF (map, 3)))) | |
1697 | content = AREF (map, 1); | |
54fa5bc1 KH |
1698 | else |
1699 | continue; | |
1700 | } | |
1701 | else | |
1702 | continue; | |
1703 | ||
1704 | if (NILP (content)) | |
1705 | continue; | |
1706 | ||
1707 | reg[RRR] = i; | |
1708 | if (NUMBERP (content)) | |
1709 | { | |
1710 | op = XINT (content); | |
1711 | i += map_set_rest_length - 1; | |
1712 | ic += map_set_rest_length - 1; | |
1713 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1714 | map_set_rest_length++; | |
1715 | } | |
1716 | else if (CONSP (content)) | |
1717 | { | |
1718 | attrib = XCAR (content); | |
1719 | value = XCDR (content); | |
1720 | if (!NUMBERP (attrib) || !NUMBERP (value)) | |
1721 | continue; | |
1722 | op = XUINT (value); | |
1723 | i += map_set_rest_length - 1; | |
1724 | ic += map_set_rest_length - 1; | |
1725 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1726 | map_set_rest_length++; | |
1727 | } | |
1728 | else if (EQ (content, Qt)) | |
1729 | { | |
1730 | op = reg[rrr]; | |
1731 | } | |
1732 | else if (EQ (content, Qlambda)) | |
1733 | { | |
1734 | i += map_set_rest_length; | |
1735 | ic += map_set_rest_length; | |
1736 | break; | |
1737 | } | |
1738 | else if (SYMBOLP (content)) | |
1739 | { | |
1740 | if (mapping_stack_pointer | |
1741 | >= &mapping_stack[MAX_MAP_SET_LEVEL]) | |
1742 | CCL_INVALID_CMD; | |
1743 | PUSH_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1744 | PUSH_MAPPING_STACK (map_set_rest_length, op); | |
1745 | stack_idx_of_map_multiple = stack_idx + 1; | |
1746 | CCL_CALL_FOR_MAP_INSTRUCTION (content, current_ic); | |
1747 | } | |
1748 | else | |
1749 | CCL_INVALID_CMD; | |
1750 | } | |
1751 | if (mapping_stack_pointer <= (mapping_stack + 1)) | |
1752 | break; | |
1753 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1754 | i += map_set_rest_length; | |
1755 | ic += map_set_rest_length; | |
1756 | POP_MAPPING_STACK (map_set_rest_length, reg[rrr]); | |
1757 | } while (1); | |
1758 | ||
e34b1164 KH |
1759 | ic = fin_ic; |
1760 | } | |
1761 | reg[rrr] = op; | |
1762 | break; | |
1763 | ||
8146262a | 1764 | case CCL_MapSingle: |
e34b1164 | 1765 | { |
8146262a | 1766 | Lisp_Object map, attrib, value, content; |
e34b1164 | 1767 | int size, point; |
8146262a | 1768 | j = XINT (ccl_prog[ic++]); /* map_id */ |
e34b1164 | 1769 | op = reg[rrr]; |
64ef2921 | 1770 | if (j >= ASIZE (Vcode_conversion_map_vector)) |
e34b1164 KH |
1771 | { |
1772 | reg[RRR] = -1; | |
1773 | break; | |
1774 | } | |
64ef2921 | 1775 | map = AREF (Vcode_conversion_map_vector, j); |
8146262a | 1776 | if (!CONSP (map)) |
e34b1164 KH |
1777 | { |
1778 | reg[RRR] = -1; | |
1779 | break; | |
1780 | } | |
03699b14 | 1781 | map = XCDR (map); |
8146262a | 1782 | if (!VECTORP (map)) |
e34b1164 KH |
1783 | { |
1784 | reg[RRR] = -1; | |
1785 | break; | |
1786 | } | |
64ef2921 SM |
1787 | size = ASIZE (map); |
1788 | point = XUINT (AREF (map, 0)); | |
e34b1164 KH |
1789 | point = op - point + 1; |
1790 | reg[RRR] = 0; | |
1791 | if ((size <= 1) || | |
1792 | (!((point >= 1) && (point < size)))) | |
1793 | reg[RRR] = -1; | |
1794 | else | |
1795 | { | |
b1cab202 | 1796 | reg[RRR] = 0; |
64ef2921 | 1797 | content = AREF (map, point); |
e34b1164 KH |
1798 | if (NILP (content)) |
1799 | reg[RRR] = -1; | |
1800 | else if (NUMBERP (content)) | |
6ae21908 | 1801 | reg[rrr] = XINT (content); |
b1cab202 | 1802 | else if (EQ (content, Qt)); |
e34b1164 KH |
1803 | else if (CONSP (content)) |
1804 | { | |
03699b14 KR |
1805 | attrib = XCAR (content); |
1806 | value = XCDR (content); | |
e34b1164 KH |
1807 | if (!NUMBERP (attrib) || !NUMBERP (value)) |
1808 | continue; | |
1809 | reg[rrr] = XUINT(value); | |
1810 | break; | |
1811 | } | |
54fa5bc1 KH |
1812 | else if (SYMBOLP (content)) |
1813 | CCL_CALL_FOR_MAP_INSTRUCTION (content, ic); | |
e34b1164 KH |
1814 | else |
1815 | reg[RRR] = -1; | |
1816 | } | |
1817 | } | |
1818 | break; | |
1819 | ||
1820 | default: | |
1821 | CCL_INVALID_CMD; | |
1822 | } | |
1823 | break; | |
1824 | ||
4ed46869 KH |
1825 | default: |
1826 | CCL_INVALID_CMD; | |
1827 | } | |
1828 | } | |
1829 | ||
1830 | ccl_error_handler: | |
0fb94c7f EZ |
1831 | /* The suppress_error member is set when e.g. a CCL-based coding |
1832 | system is used for terminal output. */ | |
1833 | if (!ccl->suppress_error && destination) | |
4ed46869 KH |
1834 | { |
1835 | /* We can insert an error message only if DESTINATION is | |
1836 | specified and we still have a room to store the message | |
1837 | there. */ | |
1838 | char msg[256]; | |
1839 | int msglen; | |
1840 | ||
12abd7d1 KH |
1841 | if (!dst) |
1842 | dst = destination; | |
1843 | ||
4ed46869 KH |
1844 | switch (ccl->status) |
1845 | { | |
1846 | case CCL_STAT_INVALID_CMD: | |
1847 | sprintf(msg, "\nCCL: Invalid command %x (ccl_code = %x) at %d.", | |
519bf146 | 1848 | code & 0x1F, code, this_ic); |
4ed46869 KH |
1849 | #ifdef CCL_DEBUG |
1850 | { | |
1851 | int i = ccl_backtrace_idx - 1; | |
1852 | int j; | |
1853 | ||
1854 | msglen = strlen (msg); | |
12abd7d1 | 1855 | if (dst + msglen <= (dst_bytes ? dst_end : src)) |
4ed46869 KH |
1856 | { |
1857 | bcopy (msg, dst, msglen); | |
1858 | dst += msglen; | |
1859 | } | |
1860 | ||
1861 | for (j = 0; j < CCL_DEBUG_BACKTRACE_LEN; j++, i--) | |
1862 | { | |
1863 | if (i < 0) i = CCL_DEBUG_BACKTRACE_LEN - 1; | |
1864 | if (ccl_backtrace_table[i] == 0) | |
1865 | break; | |
1866 | sprintf(msg, " %d", ccl_backtrace_table[i]); | |
1867 | msglen = strlen (msg); | |
12abd7d1 | 1868 | if (dst + msglen > (dst_bytes ? dst_end : src)) |
4ed46869 KH |
1869 | break; |
1870 | bcopy (msg, dst, msglen); | |
1871 | dst += msglen; | |
1872 | } | |
12abd7d1 | 1873 | goto ccl_finish; |
4ed46869 | 1874 | } |
4ed46869 | 1875 | #endif |
12abd7d1 | 1876 | break; |
4ed46869 KH |
1877 | |
1878 | case CCL_STAT_QUIT: | |
1879 | sprintf(msg, "\nCCL: Quited."); | |
1880 | break; | |
1881 | ||
1882 | default: | |
1883 | sprintf(msg, "\nCCL: Unknown error type (%d).", ccl->status); | |
1884 | } | |
1885 | ||
1886 | msglen = strlen (msg); | |
12abd7d1 | 1887 | if (dst + msglen <= (dst_bytes ? dst_end : src)) |
4ed46869 KH |
1888 | { |
1889 | bcopy (msg, dst, msglen); | |
1890 | dst += msglen; | |
1891 | } | |
8a1ae4dd | 1892 | |
31165028 KH |
1893 | if (ccl->status == CCL_STAT_INVALID_CMD) |
1894 | { | |
8a1ae4dd GM |
1895 | #if 0 /* If the remaining bytes contain 0x80..0x9F, copying them |
1896 | results in an invalid multibyte sequence. */ | |
1897 | ||
31165028 KH |
1898 | /* Copy the remaining source data. */ |
1899 | int i = src_end - src; | |
1900 | if (dst_bytes && (dst_end - dst) < i) | |
1901 | i = dst_end - dst; | |
1902 | bcopy (src, dst, i); | |
1903 | src += i; | |
1904 | dst += i; | |
8a1ae4dd GM |
1905 | #else |
1906 | /* Signal that we've consumed everything. */ | |
1907 | src = src_end; | |
1908 | #endif | |
31165028 | 1909 | } |
4ed46869 KH |
1910 | } |
1911 | ||
1912 | ccl_finish: | |
1913 | ccl->ic = ic; | |
c13362d8 KH |
1914 | ccl->stack_idx = stack_idx; |
1915 | ccl->prog = ccl_prog; | |
a8302ba3 | 1916 | ccl->eight_bit_control = (extra_bytes > 0); |
8a1ae4dd GM |
1917 | if (consumed) |
1918 | *consumed = src - source; | |
12abd7d1 | 1919 | return (dst ? dst - destination : 0); |
4ed46869 KH |
1920 | } |
1921 | ||
5232fa7b KH |
1922 | /* Resolve symbols in the specified CCL code (Lisp vector). This |
1923 | function converts symbols of code conversion maps and character | |
1924 | translation tables embeded in the CCL code into their ID numbers. | |
1925 | ||
1926 | The return value is a vector (CCL itself or a new vector in which | |
1927 | all symbols are resolved), Qt if resolving of some symbol failed, | |
1928 | or nil if CCL contains invalid data. */ | |
1929 | ||
1930 | static Lisp_Object | |
1931 | resolve_symbol_ccl_program (ccl) | |
1932 | Lisp_Object ccl; | |
1933 | { | |
1934 | int i, veclen, unresolved = 0; | |
1935 | Lisp_Object result, contents, val; | |
1936 | ||
1937 | result = ccl; | |
64ef2921 | 1938 | veclen = ASIZE (result); |
5232fa7b KH |
1939 | |
1940 | for (i = 0; i < veclen; i++) | |
1941 | { | |
64ef2921 | 1942 | contents = AREF (result, i); |
5232fa7b KH |
1943 | if (INTEGERP (contents)) |
1944 | continue; | |
1945 | else if (CONSP (contents) | |
03699b14 KR |
1946 | && SYMBOLP (XCAR (contents)) |
1947 | && SYMBOLP (XCDR (contents))) | |
5232fa7b KH |
1948 | { |
1949 | /* This is the new style for embedding symbols. The form is | |
1950 | (SYMBOL . PROPERTY). (get SYMBOL PROPERTY) should give | |
1951 | an index number. */ | |
1952 | ||
1953 | if (EQ (result, ccl)) | |
1954 | result = Fcopy_sequence (ccl); | |
1955 | ||
03699b14 | 1956 | val = Fget (XCAR (contents), XCDR (contents)); |
5232fa7b | 1957 | if (NATNUMP (val)) |
64ef2921 | 1958 | AREF (result, i) = val; |
5232fa7b KH |
1959 | else |
1960 | unresolved = 1; | |
1961 | continue; | |
1962 | } | |
1963 | else if (SYMBOLP (contents)) | |
1964 | { | |
1965 | /* This is the old style for embedding symbols. This style | |
1966 | may lead to a bug if, for instance, a translation table | |
1967 | and a code conversion map have the same name. */ | |
1968 | if (EQ (result, ccl)) | |
1969 | result = Fcopy_sequence (ccl); | |
1970 | ||
1971 | val = Fget (contents, Qtranslation_table_id); | |
1972 | if (NATNUMP (val)) | |
64ef2921 | 1973 | AREF (result, i) = val; |
5232fa7b KH |
1974 | else |
1975 | { | |
1976 | val = Fget (contents, Qcode_conversion_map_id); | |
1977 | if (NATNUMP (val)) | |
64ef2921 | 1978 | AREF (result, i) = val; |
5232fa7b KH |
1979 | else |
1980 | { | |
1981 | val = Fget (contents, Qccl_program_idx); | |
1982 | if (NATNUMP (val)) | |
64ef2921 | 1983 | AREF (result, i) = val; |
5232fa7b KH |
1984 | else |
1985 | unresolved = 1; | |
1986 | } | |
1987 | } | |
1988 | continue; | |
1989 | } | |
1990 | return Qnil; | |
1991 | } | |
1992 | ||
1993 | return (unresolved ? Qt : result); | |
1994 | } | |
1995 | ||
1996 | /* Return the compiled code (vector) of CCL program CCL_PROG. | |
1997 | CCL_PROG is a name (symbol) of the program or already compiled | |
1998 | code. If necessary, resolve symbols in the compiled code to index | |
1999 | numbers. If we failed to get the compiled code or to resolve | |
2000 | symbols, return Qnil. */ | |
2001 | ||
2002 | static Lisp_Object | |
2003 | ccl_get_compiled_code (ccl_prog) | |
2004 | Lisp_Object ccl_prog; | |
2005 | { | |
2006 | Lisp_Object val, slot; | |
2007 | ||
2008 | if (VECTORP (ccl_prog)) | |
2009 | { | |
2010 | val = resolve_symbol_ccl_program (ccl_prog); | |
2011 | return (VECTORP (val) ? val : Qnil); | |
2012 | } | |
2013 | if (!SYMBOLP (ccl_prog)) | |
2014 | return Qnil; | |
2015 | ||
2016 | val = Fget (ccl_prog, Qccl_program_idx); | |
2017 | if (! NATNUMP (val) | |
64ef2921 | 2018 | || XINT (val) >= ASIZE (Vccl_program_table)) |
5232fa7b | 2019 | return Qnil; |
64ef2921 | 2020 | slot = AREF (Vccl_program_table, XINT (val)); |
5232fa7b | 2021 | if (! VECTORP (slot) |
64ef2921 SM |
2022 | || ASIZE (slot) != 3 |
2023 | || ! VECTORP (AREF (slot, 1))) | |
5232fa7b | 2024 | return Qnil; |
64ef2921 | 2025 | if (NILP (AREF (slot, 2))) |
5232fa7b | 2026 | { |
64ef2921 | 2027 | val = resolve_symbol_ccl_program (AREF (slot, 1)); |
5232fa7b KH |
2028 | if (! VECTORP (val)) |
2029 | return Qnil; | |
64ef2921 SM |
2030 | AREF (slot, 1) = val; |
2031 | AREF (slot, 2) = Qt; | |
5232fa7b | 2032 | } |
64ef2921 | 2033 | return AREF (slot, 1); |
5232fa7b KH |
2034 | } |
2035 | ||
4ed46869 | 2036 | /* Setup fields of the structure pointed by CCL appropriately for the |
5232fa7b KH |
2037 | execution of CCL program CCL_PROG. CCL_PROG is the name (symbol) |
2038 | of the CCL program or the already compiled code (vector). | |
2039 | Return 0 if we succeed this setup, else return -1. | |
2040 | ||
2041 | If CCL_PROG is nil, we just reset the structure pointed by CCL. */ | |
2042 | int | |
2043 | setup_ccl_program (ccl, ccl_prog) | |
4ed46869 | 2044 | struct ccl_program *ccl; |
5232fa7b | 2045 | Lisp_Object ccl_prog; |
4ed46869 KH |
2046 | { |
2047 | int i; | |
2048 | ||
5232fa7b | 2049 | if (! NILP (ccl_prog)) |
ad3d1b1d | 2050 | { |
5232fa7b | 2051 | struct Lisp_Vector *vp; |
ad3d1b1d | 2052 | |
5232fa7b KH |
2053 | ccl_prog = ccl_get_compiled_code (ccl_prog); |
2054 | if (! VECTORP (ccl_prog)) | |
2055 | return -1; | |
2056 | vp = XVECTOR (ccl_prog); | |
ad3d1b1d KH |
2057 | ccl->size = vp->size; |
2058 | ccl->prog = vp->contents; | |
2059 | ccl->eof_ic = XINT (vp->contents[CCL_HEADER_EOF]); | |
2060 | ccl->buf_magnification = XINT (vp->contents[CCL_HEADER_BUF_MAG]); | |
2061 | } | |
4ed46869 | 2062 | ccl->ic = CCL_HEADER_MAIN; |
4ed46869 KH |
2063 | for (i = 0; i < 8; i++) |
2064 | ccl->reg[i] = 0; | |
2065 | ccl->last_block = 0; | |
e34b1164 | 2066 | ccl->private_state = 0; |
4ed46869 | 2067 | ccl->status = 0; |
c13362d8 | 2068 | ccl->stack_idx = 0; |
5b8ca822 | 2069 | ccl->eol_type = CODING_EOL_LF; |
ae08ba36 | 2070 | ccl->suppress_error = 0; |
5232fa7b | 2071 | return 0; |
4ed46869 KH |
2072 | } |
2073 | ||
5232fa7b | 2074 | #ifdef emacs |
6ae21908 | 2075 | |
5232fa7b | 2076 | DEFUN ("ccl-program-p", Fccl_program_p, Sccl_program_p, 1, 1, 0, |
fdb82f93 PJ |
2077 | doc: /* Return t if OBJECT is a CCL program name or a compiled CCL program code. |
2078 | See the documentation of `define-ccl-program' for the detail of CCL program. */) | |
2079 | (object) | |
5232fa7b | 2080 | Lisp_Object object; |
6ae21908 | 2081 | { |
5232fa7b | 2082 | Lisp_Object val; |
6ae21908 | 2083 | |
5232fa7b | 2084 | if (VECTORP (object)) |
6ae21908 | 2085 | { |
5232fa7b KH |
2086 | val = resolve_symbol_ccl_program (object); |
2087 | return (VECTORP (val) ? Qt : Qnil); | |
6ae21908 | 2088 | } |
5232fa7b KH |
2089 | if (!SYMBOLP (object)) |
2090 | return Qnil; | |
6ae21908 | 2091 | |
5232fa7b KH |
2092 | val = Fget (object, Qccl_program_idx); |
2093 | return ((! NATNUMP (val) | |
64ef2921 | 2094 | || XINT (val) >= ASIZE (Vccl_program_table)) |
5232fa7b | 2095 | ? Qnil : Qt); |
6ae21908 KH |
2096 | } |
2097 | ||
4ed46869 | 2098 | DEFUN ("ccl-execute", Fccl_execute, Sccl_execute, 2, 2, 0, |
fdb82f93 PJ |
2099 | doc: /* Execute CCL-PROGRAM with registers initialized by REGISTERS. |
2100 | ||
2101 | CCL-PROGRAM is a CCL program name (symbol) | |
2102 | or compiled code generated by `ccl-compile' (for backward compatibility. | |
2103 | In the latter case, the execution overhead is bigger than in the former). | |
2104 | No I/O commands should appear in CCL-PROGRAM. | |
2105 | ||
2106 | REGISTERS is a vector of [R0 R1 ... R7] where RN is an initial value | |
2107 | for the Nth register. | |
2108 | ||
2109 | As side effect, each element of REGISTERS holds the value of | |
2110 | the corresponding register after the execution. | |
2111 | ||
2112 | See the documentation of `define-ccl-program' for a definition of CCL | |
2113 | programs. */) | |
2114 | (ccl_prog, reg) | |
4ed46869 KH |
2115 | Lisp_Object ccl_prog, reg; |
2116 | { | |
2117 | struct ccl_program ccl; | |
2118 | int i; | |
2119 | ||
5232fa7b KH |
2120 | if (setup_ccl_program (&ccl, ccl_prog) < 0) |
2121 | error ("Invalid CCL program"); | |
6ae21908 | 2122 | |
b7826503 | 2123 | CHECK_VECTOR (reg); |
64ef2921 | 2124 | if (ASIZE (reg) != 8) |
d7e1fe1f | 2125 | error ("Length of vector REGISTERS is not 8"); |
4ed46869 | 2126 | |
4ed46869 | 2127 | for (i = 0; i < 8; i++) |
64ef2921 SM |
2128 | ccl.reg[i] = (INTEGERP (AREF (reg, i)) |
2129 | ? XINT (AREF (reg, i)) | |
4ed46869 KH |
2130 | : 0); |
2131 | ||
b428fdfd | 2132 | ccl_driver (&ccl, (unsigned char *)0, (unsigned char *)0, 0, 0, (int *)0); |
4ed46869 KH |
2133 | QUIT; |
2134 | if (ccl.status != CCL_STAT_SUCCESS) | |
2135 | error ("Error in CCL program at %dth code", ccl.ic); | |
2136 | ||
2137 | for (i = 0; i < 8; i++) | |
64ef2921 | 2138 | XSETINT (AREF (reg, i), ccl.reg[i]); |
4ed46869 KH |
2139 | return Qnil; |
2140 | } | |
2141 | ||
2142 | DEFUN ("ccl-execute-on-string", Fccl_execute_on_string, Sccl_execute_on_string, | |
39a68837 | 2143 | 3, 5, 0, |
fdb82f93 PJ |
2144 | doc: /* Execute CCL-PROGRAM with initial STATUS on STRING. |
2145 | ||
2146 | CCL-PROGRAM is a symbol registered by register-ccl-program, | |
2147 | or a compiled code generated by `ccl-compile' (for backward compatibility, | |
2148 | in this case, the execution is slower). | |
2149 | ||
2150 | Read buffer is set to STRING, and write buffer is allocated automatically. | |
2151 | ||
2152 | STATUS is a vector of [R0 R1 ... R7 IC], where | |
2153 | R0..R7 are initial values of corresponding registers, | |
2154 | IC is the instruction counter specifying from where to start the program. | |
2155 | If R0..R7 are nil, they are initialized to 0. | |
2156 | If IC is nil, it is initialized to head of the CCL program. | |
2157 | ||
2158 | If optional 4th arg CONTINUE is non-nil, keep IC on read operation | |
2159 | when read buffer is exausted, else, IC is always set to the end of | |
2160 | CCL-PROGRAM on exit. | |
2161 | ||
2162 | It returns the contents of write buffer as a string, | |
2163 | and as side effect, STATUS is updated. | |
2164 | If the optional 5th arg UNIBYTE-P is non-nil, the returned string | |
2165 | is a unibyte string. By default it is a multibyte string. | |
2166 | ||
2167 | See the documentation of `define-ccl-program' for the detail of CCL program. */) | |
2168 | (ccl_prog, status, str, contin, unibyte_p) | |
39a68837 | 2169 | Lisp_Object ccl_prog, status, str, contin, unibyte_p; |
4ed46869 KH |
2170 | { |
2171 | Lisp_Object val; | |
2172 | struct ccl_program ccl; | |
2173 | int i, produced; | |
2174 | int outbufsize; | |
2175 | char *outbuf; | |
5232fa7b | 2176 | struct gcpro gcpro1, gcpro2; |
6ae21908 | 2177 | |
5232fa7b KH |
2178 | if (setup_ccl_program (&ccl, ccl_prog) < 0) |
2179 | error ("Invalid CCL program"); | |
4ed46869 | 2180 | |
b7826503 | 2181 | CHECK_VECTOR (status); |
64ef2921 | 2182 | if (ASIZE (status) != 9) |
5232fa7b | 2183 | error ("Length of vector STATUS is not 9"); |
b7826503 | 2184 | CHECK_STRING (str); |
4ed46869 | 2185 | |
5232fa7b KH |
2186 | GCPRO2 (status, str); |
2187 | ||
4ed46869 KH |
2188 | for (i = 0; i < 8; i++) |
2189 | { | |
64ef2921 SM |
2190 | if (NILP (AREF (status, i))) |
2191 | XSETINT (AREF (status, i), 0); | |
2192 | if (INTEGERP (AREF (status, i))) | |
2193 | ccl.reg[i] = XINT (AREF (status, i)); | |
4ed46869 | 2194 | } |
64ef2921 | 2195 | if (INTEGERP (AREF (status, i))) |
4ed46869 | 2196 | { |
64ef2921 | 2197 | i = XFASTINT (AREF (status, 8)); |
4ed46869 KH |
2198 | if (ccl.ic < i && i < ccl.size) |
2199 | ccl.ic = i; | |
2200 | } | |
d5db4077 | 2201 | outbufsize = SBYTES (str) * ccl.buf_magnification + 256; |
4ed46869 | 2202 | outbuf = (char *) xmalloc (outbufsize); |
cb5373dd | 2203 | ccl.last_block = NILP (contin); |
7a837c89 | 2204 | ccl.multibyte = STRING_MULTIBYTE (str); |
d5db4077 KR |
2205 | produced = ccl_driver (&ccl, SDATA (str), outbuf, |
2206 | SBYTES (str), outbufsize, (int *) 0); | |
4ed46869 | 2207 | for (i = 0; i < 8; i++) |
64ef2921 SM |
2208 | XSET (AREF (status, i), Lisp_Int, ccl.reg[i]); |
2209 | XSETINT (AREF (status, 8), ccl.ic); | |
4ed46869 KH |
2210 | UNGCPRO; |
2211 | ||
39a68837 | 2212 | if (NILP (unibyte_p)) |
a3d8fcf2 KH |
2213 | { |
2214 | int nchars; | |
2215 | ||
2216 | produced = str_as_multibyte (outbuf, outbufsize, produced, &nchars); | |
2217 | val = make_multibyte_string (outbuf, nchars, produced); | |
2218 | } | |
39a68837 KH |
2219 | else |
2220 | val = make_unibyte_string (outbuf, produced); | |
157f852b | 2221 | xfree (outbuf); |
4ed46869 | 2222 | QUIT; |
a3d8fcf2 KH |
2223 | if (ccl.status == CCL_STAT_SUSPEND_BY_DST) |
2224 | error ("Output buffer for the CCL programs overflow"); | |
4ed46869 | 2225 | if (ccl.status != CCL_STAT_SUCCESS |
a3d8fcf2 | 2226 | && ccl.status != CCL_STAT_SUSPEND_BY_SRC) |
4ed46869 KH |
2227 | error ("Error in CCL program at %dth code", ccl.ic); |
2228 | ||
2229 | return val; | |
2230 | } | |
2231 | ||
2232 | DEFUN ("register-ccl-program", Fregister_ccl_program, Sregister_ccl_program, | |
2233 | 2, 2, 0, | |
fdb82f93 PJ |
2234 | doc: /* Register CCL program CCL_PROG as NAME in `ccl-program-table'. |
2235 | CCL_PROG should be a compiled CCL program (vector), or nil. | |
2236 | If it is nil, just reserve NAME as a CCL program name. | |
2237 | Return index number of the registered CCL program. */) | |
2238 | (name, ccl_prog) | |
4ed46869 KH |
2239 | Lisp_Object name, ccl_prog; |
2240 | { | |
64ef2921 | 2241 | int len = ASIZE (Vccl_program_table); |
5232fa7b KH |
2242 | int idx; |
2243 | Lisp_Object resolved; | |
4ed46869 | 2244 | |
b7826503 | 2245 | CHECK_SYMBOL (name); |
5232fa7b | 2246 | resolved = Qnil; |
4ed46869 | 2247 | if (!NILP (ccl_prog)) |
6ae21908 | 2248 | { |
b7826503 | 2249 | CHECK_VECTOR (ccl_prog); |
5232fa7b | 2250 | resolved = resolve_symbol_ccl_program (ccl_prog); |
4d247a1f KH |
2251 | if (NILP (resolved)) |
2252 | error ("Error in CCL program"); | |
2253 | if (VECTORP (resolved)) | |
5232fa7b KH |
2254 | { |
2255 | ccl_prog = resolved; | |
2256 | resolved = Qt; | |
2257 | } | |
4d247a1f KH |
2258 | else |
2259 | resolved = Qnil; | |
6ae21908 | 2260 | } |
5232fa7b KH |
2261 | |
2262 | for (idx = 0; idx < len; idx++) | |
4ed46869 | 2263 | { |
5232fa7b | 2264 | Lisp_Object slot; |
4ed46869 | 2265 | |
64ef2921 | 2266 | slot = AREF (Vccl_program_table, idx); |
5232fa7b KH |
2267 | if (!VECTORP (slot)) |
2268 | /* This is the first unsed slot. Register NAME here. */ | |
4ed46869 KH |
2269 | break; |
2270 | ||
64ef2921 | 2271 | if (EQ (name, AREF (slot, 0))) |
4ed46869 | 2272 | { |
5232fa7b | 2273 | /* Update this slot. */ |
64ef2921 SM |
2274 | AREF (slot, 1) = ccl_prog; |
2275 | AREF (slot, 2) = resolved; | |
5232fa7b | 2276 | return make_number (idx); |
4ed46869 KH |
2277 | } |
2278 | } | |
2279 | ||
5232fa7b | 2280 | if (idx == len) |
4ed46869 | 2281 | { |
5232fa7b KH |
2282 | /* Extend the table. */ |
2283 | Lisp_Object new_table; | |
4ed46869 KH |
2284 | int j; |
2285 | ||
5232fa7b | 2286 | new_table = Fmake_vector (make_number (len * 2), Qnil); |
4ed46869 | 2287 | for (j = 0; j < len; j++) |
64ef2921 SM |
2288 | AREF (new_table, j) |
2289 | = AREF (Vccl_program_table, j); | |
4ed46869 KH |
2290 | Vccl_program_table = new_table; |
2291 | } | |
2292 | ||
5232fa7b KH |
2293 | { |
2294 | Lisp_Object elt; | |
2295 | ||
2296 | elt = Fmake_vector (make_number (3), Qnil); | |
64ef2921 SM |
2297 | AREF (elt, 0) = name; |
2298 | AREF (elt, 1) = ccl_prog; | |
2299 | AREF (elt, 2) = resolved; | |
2300 | AREF (Vccl_program_table, idx) = elt; | |
5232fa7b KH |
2301 | } |
2302 | ||
2303 | Fput (name, Qccl_program_idx, make_number (idx)); | |
2304 | return make_number (idx); | |
4ed46869 KH |
2305 | } |
2306 | ||
8146262a KH |
2307 | /* Register code conversion map. |
2308 | A code conversion map consists of numbers, Qt, Qnil, and Qlambda. | |
d617f6df DL |
2309 | The first element is the start code point. |
2310 | The other elements are mapped numbers. | |
8146262a KH |
2311 | Symbol t means to map to an original number before mapping. |
2312 | Symbol nil means that the corresponding element is empty. | |
d617f6df | 2313 | Symbol lambda means to terminate mapping here. |
e34b1164 KH |
2314 | */ |
2315 | ||
8146262a KH |
2316 | DEFUN ("register-code-conversion-map", Fregister_code_conversion_map, |
2317 | Sregister_code_conversion_map, | |
e34b1164 | 2318 | 2, 2, 0, |
fdb82f93 PJ |
2319 | doc: /* Register SYMBOL as code conversion map MAP. |
2320 | Return index number of the registered map. */) | |
2321 | (symbol, map) | |
8146262a | 2322 | Lisp_Object symbol, map; |
e34b1164 | 2323 | { |
64ef2921 | 2324 | int len = ASIZE (Vcode_conversion_map_vector); |
e34b1164 KH |
2325 | int i; |
2326 | Lisp_Object index; | |
2327 | ||
b7826503 PJ |
2328 | CHECK_SYMBOL (symbol); |
2329 | CHECK_VECTOR (map); | |
e34b1164 KH |
2330 | |
2331 | for (i = 0; i < len; i++) | |
2332 | { | |
64ef2921 | 2333 | Lisp_Object slot = AREF (Vcode_conversion_map_vector, i); |
e34b1164 KH |
2334 | |
2335 | if (!CONSP (slot)) | |
2336 | break; | |
2337 | ||
03699b14 | 2338 | if (EQ (symbol, XCAR (slot))) |
e34b1164 KH |
2339 | { |
2340 | index = make_number (i); | |
f3fbd155 | 2341 | XSETCDR (slot, map); |
8146262a KH |
2342 | Fput (symbol, Qcode_conversion_map, map); |
2343 | Fput (symbol, Qcode_conversion_map_id, index); | |
e34b1164 KH |
2344 | return index; |
2345 | } | |
2346 | } | |
2347 | ||
2348 | if (i == len) | |
2349 | { | |
2350 | Lisp_Object new_vector = Fmake_vector (make_number (len * 2), Qnil); | |
2351 | int j; | |
2352 | ||
2353 | for (j = 0; j < len; j++) | |
64ef2921 SM |
2354 | AREF (new_vector, j) |
2355 | = AREF (Vcode_conversion_map_vector, j); | |
8146262a | 2356 | Vcode_conversion_map_vector = new_vector; |
e34b1164 KH |
2357 | } |
2358 | ||
2359 | index = make_number (i); | |
8146262a KH |
2360 | Fput (symbol, Qcode_conversion_map, map); |
2361 | Fput (symbol, Qcode_conversion_map_id, index); | |
64ef2921 | 2362 | AREF (Vcode_conversion_map_vector, i) = Fcons (symbol, map); |
e34b1164 KH |
2363 | return index; |
2364 | } | |
2365 | ||
2366 | ||
dfcf069d | 2367 | void |
4ed46869 KH |
2368 | syms_of_ccl () |
2369 | { | |
2370 | staticpro (&Vccl_program_table); | |
6703ac4f | 2371 | Vccl_program_table = Fmake_vector (make_number (32), Qnil); |
4ed46869 | 2372 | |
6ae21908 KH |
2373 | Qccl_program = intern ("ccl-program"); |
2374 | staticpro (&Qccl_program); | |
2375 | ||
2376 | Qccl_program_idx = intern ("ccl-program-idx"); | |
2377 | staticpro (&Qccl_program_idx); | |
e34b1164 | 2378 | |
8146262a KH |
2379 | Qcode_conversion_map = intern ("code-conversion-map"); |
2380 | staticpro (&Qcode_conversion_map); | |
6ae21908 | 2381 | |
8146262a KH |
2382 | Qcode_conversion_map_id = intern ("code-conversion-map-id"); |
2383 | staticpro (&Qcode_conversion_map_id); | |
6ae21908 | 2384 | |
8146262a | 2385 | DEFVAR_LISP ("code-conversion-map-vector", &Vcode_conversion_map_vector, |
fdb82f93 | 2386 | doc: /* Vector of code conversion maps. */); |
8146262a | 2387 | Vcode_conversion_map_vector = Fmake_vector (make_number (16), Qnil); |
e34b1164 | 2388 | |
4ed46869 | 2389 | DEFVAR_LISP ("font-ccl-encoder-alist", &Vfont_ccl_encoder_alist, |
fdb82f93 PJ |
2390 | doc: /* Alist of fontname patterns vs corresponding CCL program. |
2391 | Each element looks like (REGEXP . CCL-CODE), | |
2392 | where CCL-CODE is a compiled CCL program. | |
2393 | When a font whose name matches REGEXP is used for displaying a character, | |
2394 | CCL-CODE is executed to calculate the code point in the font | |
2395 | from the charset number and position code(s) of the character which are set | |
2396 | in CCL registers R0, R1, and R2 before the execution. | |
2397 | The code point in the font is set in CCL registers R1 and R2 | |
2398 | when the execution terminated. | |
2399 | If the font is single-byte font, the register R2 is not used. */); | |
4ed46869 KH |
2400 | Vfont_ccl_encoder_alist = Qnil; |
2401 | ||
d80dc57e DL |
2402 | DEFVAR_LISP ("translation-hash-table-vector", &Vtranslation_hash_table_vector, |
2403 | doc: /* Vector containing all translation hash tables ever defined. | |
2404 | Comprises pairs (SYMBOL . TABLE) where SYMBOL and TABLE were set up by calls | |
2405 | to `define-translation-hash-table'. The vector is indexed by the table id | |
2406 | used by CCL. */); | |
2407 | Vtranslation_hash_table_vector = Qnil; | |
2408 | ||
5232fa7b | 2409 | defsubr (&Sccl_program_p); |
4ed46869 KH |
2410 | defsubr (&Sccl_execute); |
2411 | defsubr (&Sccl_execute_on_string); | |
2412 | defsubr (&Sregister_ccl_program); | |
8146262a | 2413 | defsubr (&Sregister_code_conversion_map); |
4ed46869 KH |
2414 | } |
2415 | ||
2416 | #endif /* emacs */ |