Mercurial > hg > xemacs-beta
annotate src/number-mp.c @ 4928:ea701c23ed84
change text_width method to take a window, in preparation for unicode-internal changes
-------------------- ChangeLog entries follow: --------------------
src/ChangeLog addition:
2010-01-19 Ben Wing <ben@xemacs.org>
* console-impl.h (struct console_methods):
* console-stream.c (stream_text_width):
* redisplay-msw.c (mswindows_output_string):
* redisplay-msw.c (mswindows_text_width):
* redisplay-tty.c (tty_text_width):
* redisplay-xlike-inc.c (XLIKE_text_width):
* redisplay-xlike-inc.c (XLIKE_output_string):
* redisplay.c:
* redisplay.c (redisplay_window_text_width_ichar_string):
* redisplay.c (redisplay_text_width_string):
Change the text_width method to take a window instead of a frame.
Needed for Unicode-internal.
| author | Ben Wing <ben@xemacs.org> |
|---|---|
| date | Tue, 19 Jan 2010 11:21:34 -0600 |
| parents | 2fc0e2f18322 |
| children | 16112448d484 |
| rev | line source |
|---|---|
| 1983 | 1 /* Numeric types for XEmacs using the MP library. |
| 2 Copyright (C) 2004 Jerry James. | |
| 3 | |
| 4 This file is part of XEmacs. | |
| 5 | |
| 6 XEmacs is free software; you can redistribute it and/or modify it | |
| 7 under the terms of the GNU General Public License as published by the | |
| 8 Free Software Foundation; either version 2, or (at your option) any | |
| 9 later version. | |
| 10 | |
| 11 XEmacs is distributed in the hope that it will be useful, but WITHOUT | |
| 12 ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or | |
| 13 FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License | |
| 14 for more details. | |
| 15 | |
| 16 You should have received a copy of the GNU General Public License | |
| 17 along with XEmacs; see the file COPYING. If not, write to | |
|
4802
2fc0e2f18322
Don't create any bignums before pdumping. Add bignum, ratio, and bigfloat
Jerry James <james@xemacs.org>
parents:
1993
diff
changeset
|
18 the Free Software Foundation, Inc., 51 Franklin St - Fifth Floor, |
|
2fc0e2f18322
Don't create any bignums before pdumping. Add bignum, ratio, and bigfloat
Jerry James <james@xemacs.org>
parents:
1993
diff
changeset
|
19 Boston, MA 02111-1301, USA. */ |
| 1983 | 20 |
| 21 /* Synched up with: Not in FSF. */ | |
| 22 | |
| 23 #include <config.h> | |
| 24 #include <limits.h> | |
| 25 #include <math.h> | |
| 26 #include "lisp.h" | |
| 27 | |
| 28 static MINT *bignum_bytesize, *bignum_long_sign_bit, *bignum_one, *bignum_two; | |
| 29 MINT *bignum_zero, *intern_bignum; | |
| 30 MINT *bignum_min_int, *bignum_max_int, *bignum_max_uint; | |
| 31 MINT *bignum_min_long, *bignum_max_long, *bignum_max_ulong; | |
| 32 short div_rem; | |
| 33 | |
| 34 char * | |
| 35 bignum_to_string (bignum b, int base) | |
| 36 { | |
| 37 REGISTER unsigned int i; | |
| 38 unsigned int bufsize = 128U, index = 0U; | |
| 39 int sign; | |
| 40 char *buffer = xnew_array (char, 128), *retval; | |
| 41 MINT *quo = MP_ITOM (0); | |
| 42 short rem; | |
| 43 | |
| 44 /* FIXME: signal something if base is < 2 or doesn't fit into a short. */ | |
| 45 | |
| 46 /* Save the sign for later */ | |
| 47 sign = MP_MCMP (b, bignum_zero); | |
| 48 | |
| 49 if (sign == 0) | |
| 50 { | |
| 51 XREALLOC_ARRAY (buffer, char, 2); | |
| 52 buffer[0] = '0'; | |
| 53 buffer[1] = '\0'; | |
| 54 return buffer; | |
| 55 } | |
| 56 /* Copy abs(b) into quo for destructive modification */ | |
| 57 else if (sign < 0) | |
| 58 MP_MSUB (bignum_zero, b, quo); | |
| 59 else | |
| 60 MP_MOVE (b, quo); | |
| 61 | |
| 62 quo = MP_ITOM (0); | |
| 63 | |
| 64 /* Loop over the digits of b (in BASE) and place each one into buffer */ | |
| 65 for (i = 0U; MP_MCMP(quo, bignum_zero) > 0; i++) | |
| 66 { | |
| 67 MP_SDIV (quo, base, quo, &rem); | |
| 68 if (index == bufsize) | |
| 69 { | |
| 70 bufsize <<= 1; | |
| 71 XREALLOC_ARRAY (buffer, char, bufsize); | |
| 72 } | |
| 73 buffer[index++] = rem < 10 ? rem + '0' : rem - 10 + 'a'; | |
| 74 } | |
| 75 MP_MFREE (quo); | |
| 76 | |
| 77 /* Reverse the digits, maybe add a minus sign, and add a null terminator */ | |
| 78 bufsize = index + (sign < 0 ? 1 : 0) + 1; | |
| 79 retval = xnew_array (char, bufsize); | |
| 80 if (sign < 0) | |
| 81 { | |
| 82 retval[0] = '-'; | |
| 83 i = 1; | |
| 84 } | |
| 85 else | |
| 86 i = 0; | |
| 87 for (; i < bufsize - 1; i++) | |
| 88 retval[i] = buffer[--index]; | |
| 89 retval[bufsize - 1] = '\0'; | |
| 90 xfree (buffer, char *); | |
| 91 return retval; | |
| 92 } | |
| 93 | |
| 94 #define BIGNUM_TO_TYPE(type,accumtype) do { \ | |
| 95 MP_MULT (b, quo, quo); \ | |
| 96 for (i = 0U; i < sizeof(type); i++) \ | |
| 97 { \ | |
| 98 MP_SDIV (quo, 256, quo, &rem); \ | |
| 99 retval |= ((accumtype) rem) << (8 * i); \ | |
| 100 } \ | |
| 101 MP_MFREE (quo); \ | |
| 102 } while (0) | |
| 103 | |
| 104 int | |
| 105 bignum_to_int (bignum b) | |
| 106 { | |
| 107 short rem, sign; | |
| 108 unsigned int retval = 0; | |
| 109 REGISTER unsigned int i; | |
| 110 MINT *quo; | |
| 111 | |
| 112 sign = MP_MCMP (b, bignum_zero) < 0 ? -1 : 1; | |
| 113 quo = MP_ITOM (sign); | |
| 114 BIGNUM_TO_TYPE (int, unsigned int); | |
| 115 return ((int) retval) * sign; | |
| 116 } | |
| 117 | |
| 118 unsigned int | |
| 119 bignum_to_uint (bignum b) | |
| 120 { | |
| 121 short rem; | |
| 122 unsigned int retval = 0U; | |
| 123 REGISTER unsigned int i; | |
| 124 MINT *quo; | |
| 125 | |
| 126 quo = MP_ITOM (MP_MCMP (b, bignum_zero) < 0 ? -1 : 1); | |
| 127 BIGNUM_TO_TYPE (unsigned int, unsigned int); | |
| 128 return retval; | |
| 129 } | |
| 130 | |
| 131 long | |
| 132 bignum_to_long (bignum b) | |
| 133 { | |
| 134 short rem, sign; | |
| 135 unsigned long retval = 0L; | |
| 136 REGISTER unsigned int i; | |
| 137 MINT *quo; | |
| 138 | |
| 139 sign = MP_MCMP (b, bignum_zero) < 0 ? -1 : 1; | |
| 140 quo = MP_ITOM (sign); | |
| 141 BIGNUM_TO_TYPE (long, unsigned long); | |
| 142 return ((long) retval) * sign; | |
| 143 } | |
| 144 | |
| 145 unsigned long | |
| 146 bignum_to_ulong (bignum b) | |
| 147 { | |
| 148 short rem; | |
| 149 unsigned long retval = 0UL; | |
| 150 REGISTER unsigned int i; | |
| 151 MINT *quo; | |
| 152 | |
| 153 quo = MP_ITOM (MP_MCMP (b, bignum_zero) < 0 ? -1 : 1); | |
| 154 BIGNUM_TO_TYPE (unsigned long, unsigned long); | |
| 155 return retval; | |
| 156 } | |
| 157 | |
| 158 double | |
| 159 bignum_to_double (bignum b) | |
| 160 { | |
| 161 short rem, sign; | |
| 1990 | 162 double retval = 0.0, factor = 1.0; |
| 1983 | 163 REGISTER unsigned int i; |
| 164 MINT *quo; | |
| 165 | |
| 166 sign = MP_MCMP (b, bignum_zero) < 0 ? -1 : 1; | |
| 167 quo = MP_ITOM (sign); | |
| 168 MP_MULT (b, quo, quo); | |
| 1990 | 169 for (i = 0U; MP_MCMP (quo, bignum_zero) > 0; i++) |
| 1983 | 170 { |
| 171 MP_SDIV (quo, 256, quo, &rem); | |
| 1990 | 172 retval += rem * factor; |
| 173 factor *= 256.0; | |
| 1983 | 174 } |
| 175 MP_MFREE (quo); | |
| 176 return retval * sign; | |
| 177 } | |
| 178 | |
| 179 static short | |
| 180 char_to_number (char c) | |
| 181 { | |
| 182 if (c >= '0' && c <= '9') | |
| 183 return c - '0'; | |
| 184 if (c >= 'a' && c <= 'z') | |
| 185 return c - 'a' + 10; | |
| 186 if (c >= 'A' && c <= 'Z') | |
| 187 return c - 'A' + 10; | |
| 188 return -1; | |
| 189 } | |
| 190 | |
| 191 int | |
| 192 bignum_set_string (bignum b, const char *s, int base) | |
| 193 { | |
| 194 MINT *mbase; | |
| 195 short digit; | |
| 1993 | 196 int neg = 0; |
| 1983 | 197 |
| 198 if (base == 0) | |
| 199 { | |
| 200 if (s[0] == '0' && (s[1] == 'x' || s[1] == 'X')) | |
| 201 { | |
| 202 base = 16; | |
| 203 s += 2; | |
| 204 } | |
| 205 else if (*s == '0') | |
| 206 { | |
| 207 base = 8; | |
| 208 s++; | |
| 209 } | |
| 210 else | |
| 211 base = 10; | |
| 212 } | |
| 213 | |
| 214 /* FIXME: signal something if base is < 2 or doesn't fit into a short. */ | |
| 215 | |
| 1993 | 216 if (*s == '-') |
| 217 { | |
| 218 s++; | |
| 219 neg = 1; | |
| 220 } | |
| 221 | |
| 1983 | 222 mbase = MP_ITOM ((short) base); |
| 223 MP_MOVE (bignum_zero, b); | |
|
4802
2fc0e2f18322
Don't create any bignums before pdumping. Add bignum, ratio, and bigfloat
Jerry James <james@xemacs.org>
parents:
1993
diff
changeset
|
224 |
| 1983 | 225 for (digit = char_to_number (*s); digit >= 0 && digit < base; |
| 226 digit = char_to_number (*++s)) | |
| 227 { | |
| 228 MINT *temp; | |
| 229 | |
| 230 MP_MULT (b, mbase, b); | |
| 231 temp = MP_ITOM (digit); | |
| 232 MP_MADD (b, temp, b); | |
| 233 MP_MFREE (temp); | |
| 234 } | |
| 235 | |
| 1993 | 236 if (neg) |
| 237 MP_MSUB (bignum_zero, b, b); | |
| 238 | |
| 1983 | 239 return (digit >= 0) ? -1 : 0; |
| 240 } | |
| 241 | |
| 242 void | |
| 243 bignum_set_long (MINT *b, long l) | |
| 244 { | |
| 245 /* Negative l is hard, not least because -LONG_MIN == LONG_MIN. We pretend | |
| 246 that l is unsigned, then subtract off the amount equal to the sign bit. */ | |
| 247 bignum_set_ulong (b, (unsigned long) l); | |
| 248 if (l < 0L) | |
| 249 MP_MSUB (b, bignum_long_sign_bit, b); | |
| 250 } | |
| 251 | |
| 252 void | |
| 253 bignum_set_ulong (bignum b, unsigned long l) | |
| 254 { | |
| 255 REGISTER unsigned int i; | |
| 256 MINT *multiplier = MP_ITOM (1); | |
| 257 | |
| 258 MP_MOVE (bignum_zero, b); | |
| 259 for (i = 0UL; l > 0UL; l >>= 8, i++) | |
| 260 { | |
| 261 MINT *temp = MP_ITOM ((short) (l & 255)); | |
| 262 MP_MULT (multiplier, temp, temp); | |
| 263 MP_MADD (b, temp, b); | |
| 264 MP_MULT (multiplier, bignum_bytesize, multiplier); | |
| 265 MP_MFREE (temp); | |
| 266 } | |
| 267 MP_MFREE (multiplier); | |
| 268 } | |
| 269 | |
| 270 void | |
| 271 bignum_set_double (bignum b, double d) | |
| 272 { | |
| 273 REGISTER unsigned int i; | |
| 274 int negative = (d < 0) ? 1 : 0; | |
| 275 MINT *multiplier = MP_ITOM (1); | |
| 276 | |
| 277 MP_MOVE (bignum_zero, b); | |
| 278 if (negative) | |
| 279 d = -d; | |
| 280 for (i = 0UL; d > 0.0; d /= 256, i++) | |
| 281 { | |
| 282 MINT *temp = MP_ITOM ((short) fmod (d, 256.0)); | |
| 283 MP_MULT (multiplier, temp, temp); | |
| 284 MP_MADD (b, temp, b); | |
| 285 MP_MULT (multiplier, bignum_bytesize, multiplier); | |
| 286 MP_MFREE (temp); | |
| 287 } | |
| 288 MP_MFREE (multiplier); | |
| 289 if (negative) | |
| 290 MP_MSUB (bignum_zero, b, b); | |
| 291 } | |
| 292 | |
| 293 /* Return nonzero if b1 is exactly divisible by b2 */ | |
| 294 int | |
| 295 bignum_divisible_p (bignum b1, bignum b2) | |
| 296 { | |
| 297 int retval; | |
| 298 MINT *rem = MP_ITOM (0); | |
| 299 MP_MDIV (b1, b2, intern_bignum, rem); | |
| 300 retval = (MP_MCMP (rem, bignum_zero) == 0); | |
| 301 MP_MFREE (rem); | |
| 302 return retval; | |
| 303 } | |
| 304 | |
| 305 void bignum_ceil (bignum quotient, bignum N, bignum D) | |
| 306 { | |
| 307 MP_MDIV (N, D, quotient, intern_bignum); | |
| 308 if (MP_MCMP (intern_bignum, bignum_zero) > 0 && | |
| 309 MP_MCMP (quotient, bignum_zero) > 0) | |
| 310 MP_MADD (quotient, bignum_one, quotient); | |
| 311 } | |
| 312 | |
| 313 void bignum_floor (bignum quotient, bignum N, bignum D) | |
| 314 { | |
| 315 MP_MDIV (N, D, quotient, intern_bignum); | |
| 316 if (MP_MCMP (intern_bignum, bignum_zero) > 0 && | |
| 317 MP_MCMP (quotient, bignum_zero) < 0) | |
| 318 MP_MSUB (quotient, bignum_one, quotient); | |
| 319 } | |
| 320 | |
| 321 /* RESULT = N to the POWth power */ | |
| 322 void | |
| 323 bignum_pow (bignum result, bignum n, unsigned long pow) | |
| 324 { | |
| 325 MP_MOVE (bignum_one, result); | |
| 326 for ( ; pow > 0UL; pow--) | |
| 327 MP_MULT (result, n, result); | |
| 328 } | |
| 329 | |
| 330 /* lcm(b1,b2) = b1 * b2 / gcd(b1, b2) */ | |
| 331 void | |
| 332 bignum_lcm (bignum result, bignum b1, bignum b2) | |
| 333 { | |
| 334 MP_MULT (b1, b2, result); | |
| 335 MP_GCD (b1, b2, intern_bignum); | |
| 336 MP_MDIV (result, intern_bignum, result, intern_bignum); | |
| 337 } | |
| 338 | |
| 339 /* FIXME: We can't handle negative args, so right now we just make them | |
| 340 positive before doing anything else. How should we really handle negative | |
| 341 args? */ | |
| 342 #define bignum_bit_op(result, b1, b2, op) \ | |
| 343 REGISTER unsigned int i; \ | |
| 344 MINT *multiplier = MP_ITOM (1), *n1 = MP_ITOM (0), *n2 = MP_ITOM (0); \ | |
| 345 \ | |
| 346 if (MP_MCMP (bignum_zero, b1) > 0) \ | |
| 347 MP_MSUB (bignum_zero, b1, n1); \ | |
| 348 else \ | |
| 349 MP_MOVE (b1, n1); \ | |
| 350 if (MP_MCMP (bignum_zero, b2) > 0) \ | |
| 351 MP_MSUB (bignum_zero, b2, n2); \ | |
| 352 else \ | |
| 353 MP_MOVE (b2, n2); \ | |
| 354 \ | |
| 355 MP_MOVE (bignum_zero, result); \ | |
| 356 \ | |
| 357 for (i = 0UL; MP_MCMP (bignum_zero, n1) < 0 && \ | |
| 358 MP_MCMP (bignum_zero, n2) < 0; i++) \ | |
| 359 { \ | |
| 360 short byte1, byte2; \ | |
| 361 MINT *temp; \ | |
| 362 \ | |
| 363 MP_SDIV (n1, 256, n1, &byte1); \ | |
| 364 MP_SDIV (n2, 256, n2, &byte2); \ | |
| 365 temp = MP_ITOM (byte1 op byte2); \ | |
| 366 MP_MULT (multiplier, temp, temp); \ | |
| 367 MP_MADD (result, temp, result); \ | |
| 368 MP_MULT (multiplier, bignum_bytesize, multiplier); \ | |
| 369 MP_MFREE (temp); \ | |
| 370 } \ | |
| 371 MP_MFREE (n2); \ | |
| 372 MP_MFREE (n1); \ | |
| 373 MP_MFREE (multiplier) | |
| 374 | |
| 375 void | |
| 376 bignum_and (bignum result, bignum b1, bignum b2) | |
| 377 { | |
| 378 bignum_bit_op (result, b1, b2, &); | |
| 379 } | |
| 380 | |
| 381 void | |
| 382 bignum_ior (bignum result, bignum b1, bignum b2) | |
| 383 { | |
| 384 bignum_bit_op (result, b1, b2, |); | |
| 385 } | |
| 386 | |
| 387 void | |
| 388 bignum_xor (bignum result, bignum b1, bignum b2) | |
| 389 { | |
| 390 bignum_bit_op (result, b1, b2, ^); | |
| 391 } | |
| 392 | |
| 393 /* NOT is not well-defined for bignums ... where do you stop flipping bits? | |
| 394 We just flip until we see the last one. This is probably a bad idea. */ | |
| 395 void | |
| 396 bignum_not (bignum result, bignum b) | |
| 397 { | |
| 398 REGISTER unsigned int i; | |
| 399 MINT *multiplier = MP_ITOM (1), *n = MP_ITOM (0); | |
| 400 | |
| 401 if (MP_MCMP (bignum_zero, b) > 0) | |
| 402 MP_MSUB (bignum_zero, b, n); | |
| 403 else | |
| 404 MP_MOVE (b, n); | |
| 405 | |
| 406 MP_MOVE (bignum_zero, result); | |
| 407 | |
| 408 for (i = 0UL; MP_MCMP (bignum_zero, n) < 0; i++) | |
| 409 { | |
| 410 short byte; | |
| 411 MINT *temp; | |
| 412 | |
| 413 MP_SDIV (n, 256, n, &byte); | |
| 414 temp = MP_ITOM (~byte); | |
| 415 MP_MULT (multiplier, temp, temp); | |
| 416 MP_MADD (result, temp, result); | |
| 417 MP_MULT (multiplier, bignum_bytesize, multiplier); | |
| 418 MP_MFREE (temp); | |
| 419 } | |
| 420 MP_MFREE (n); | |
| 421 MP_MFREE (multiplier); | |
| 422 } | |
| 423 | |
| 424 void | |
| 425 bignum_setbit (bignum b, unsigned long bit) | |
| 426 { | |
| 427 bignum_pow (intern_bignum, bignum_two, bit); | |
| 428 bignum_ior (b, b, intern_bignum); | |
| 429 } | |
| 430 | |
| 431 /* This is so evil, even I feel queasy. */ | |
| 432 void | |
| 433 bignum_clrbit (bignum b, unsigned long bit) | |
| 434 { | |
| 435 MINT *num = MP_ITOM (0); | |
| 436 | |
| 437 /* See if the bit is already set, and subtract it off if not */ | |
| 438 MP_MOVE (b, intern_bignum); | |
| 439 bignum_pow (num, bignum_two, bit); | |
| 440 bignum_ior (intern_bignum, intern_bignum, num); | |
| 441 if (MP_MCMP (b, intern_bignum) == 0) | |
| 442 MP_MSUB (b, num, b); | |
| 443 MP_MFREE (num); | |
| 444 } | |
| 445 | |
| 446 int | |
| 447 bignum_testbit (bignum b, unsigned long bit) | |
| 448 { | |
| 449 bignum_pow (intern_bignum, bignum_two, bit); | |
| 450 bignum_and (intern_bignum, b, intern_bignum); | |
| 451 return MP_MCMP (intern_bignum, bignum_zero); | |
| 452 } | |
| 453 | |
| 454 void | |
| 455 bignum_lshift (bignum result, bignum b, unsigned long bits) | |
| 456 { | |
| 457 bignum_pow (intern_bignum, bignum_two, bits); | |
| 458 MP_MULT (b, intern_bignum, result); | |
| 459 } | |
| 460 | |
| 461 void | |
| 462 bignum_rshift (bignum result, bignum b, unsigned long bits) | |
| 463 { | |
| 464 bignum_pow (intern_bignum, bignum_two, bits); | |
| 465 MP_MDIV (b, intern_bignum, result, intern_bignum); | |
| 466 } | |
| 467 | |
| 468 void bignum_random_seed(unsigned long seed) | |
| 469 { | |
| 470 /* FIXME: Implement me */ | |
| 471 } | |
| 472 | |
| 473 void bignum_random(bignum result, bignum limit) | |
| 474 { | |
| 475 /* FIXME: Implement me */ | |
| 476 MP_MOVE (bignum_zero, result); | |
| 477 } | |
| 478 | |
| 479 void | |
| 480 init_number_mp () | |
| 481 { | |
| 482 REGISTER unsigned int i; | |
| 483 | |
| 484 bignum_zero = MP_ITOM (0); | |
| 485 bignum_one = MP_ITOM (1); | |
| 486 bignum_two = MP_ITOM (2); | |
| 487 | |
| 488 /* intern_bignum holds throwaway values from macro expansions in | |
| 489 number-mp.h. Its value is immaterial. */ | |
| 490 intern_bignum = MP_ITOM (0); | |
| 491 | |
| 492 /* bignum_bytesize holds the number of bits in a byte. */ | |
| 493 bignum_bytesize = MP_ITOM (256); | |
| 494 | |
| 495 /* bignum_long_sign_bit holds an adjustment for negative longs. */ | |
| 496 bignum_long_sign_bit = MP_ITOM (256); | |
| 497 for (i = 1UL; i < sizeof (long); i++) | |
| 498 MP_MULT (bignum_bytesize, bignum_long_sign_bit, bignum_long_sign_bit); | |
| 499 | |
| 500 /* The MP interface only supports turning short ints into MINTs, so we have | |
| 501 to set these the hard way. */ | |
| 502 | |
| 503 bignum_min_int = MP_ITOM (0); | |
| 504 bignum_set_long (bignum_min_int, INT_MIN); | |
| 505 | |
| 506 bignum_max_int = MP_ITOM (0); | |
| 507 bignum_set_long (bignum_max_int, INT_MAX); | |
| 508 | |
| 509 bignum_max_uint = MP_ITOM (0); | |
| 510 bignum_set_ulong (bignum_max_uint, UINT_MAX); | |
| 511 | |
| 512 bignum_min_long = MP_ITOM (0); | |
| 513 bignum_set_long (bignum_min_long, LONG_MIN); | |
| 514 | |
| 515 bignum_max_long = MP_ITOM (0); | |
| 516 bignum_set_long (bignum_max_long, LONG_MAX); | |
| 517 | |
| 518 bignum_max_ulong = MP_ITOM (0); | |
| 519 bignum_set_ulong (bignum_max_ulong, ULONG_MAX); | |
| 520 } |