| 1 |
/* String intrinsics helper functions. |
| 2 |
Copyright (C) 2002-2025 Free Software Foundation, Inc. |
| 3 |
|
| 4 |
This file is part of the GNU Fortran runtime library (libgfortran). |
| 5 |
|
| 6 |
Libgfortran is free software; you can redistribute it and/or |
| 7 |
modify it under the terms of the GNU General Public |
| 8 |
License as published by the Free Software Foundation; either |
| 9 |
version 3 of the License, or (at your option) any later version. |
| 10 |
|
| 11 |
Libgfortran is distributed in the hope that it will be useful, |
| 12 |
but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 13 |
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
| 14 |
GNU General Public License for more details. |
| 15 |
|
| 16 |
Under Section 7 of GPL version 3, you are granted additional |
| 17 |
permissions described in the GCC Runtime Library Exception, version |
| 18 |
3.1, as published by the Free Software Foundation. |
| 19 |
|
| 20 |
You should have received a copy of the GNU General Public License and |
| 21 |
a copy of the GCC Runtime Library Exception along with this program; |
| 22 |
see the files COPYING3 and COPYING.RUNTIME respectively. If not, see |
| 23 |
<http://www.gnu.org/licenses/>. */ |
| 24 |
|
| 25 |
|
| 26 |
/* Rename the functions. */ |
| 27 |
#define concat_string SUFFIX(concat_string) |
| 28 |
#define string_len_trim SUFFIX(string_len_trim) |
| 29 |
#define adjustl SUFFIX(adjustl) |
| 30 |
#define adjustr SUFFIX(adjustr) |
| 31 |
#define string_index SUFFIX(string_index) |
| 32 |
#define string_scan SUFFIX(string_scan) |
| 33 |
#define string_verify SUFFIX(string_verify) |
| 34 |
#define string_trim SUFFIX(string_trim) |
| 35 |
#define string_minmax SUFFIX(string_minmax) |
| 36 |
#define zero_length_string SUFFIX(zero_length_string) |
| 37 |
#define compare_string SUFFIX(compare_string) |
| 38 |
|
| 39 |
|
| 40 |
/* The prototypes. */ |
| 41 |
|
| 42 |
extern void concat_string (gfc_charlen_type, CHARTYPE *, |
| 43 |
gfc_charlen_type, const CHARTYPE *, |
| 44 |
gfc_charlen_type, const CHARTYPE *); |
| 45 |
export_proto(concat_string); |
| 46 |
|
| 47 |
extern void adjustl (CHARTYPE *, gfc_charlen_type, const CHARTYPE *); |
| 48 |
export_proto(adjustl); |
| 49 |
|
| 50 |
extern void adjustr (CHARTYPE *, gfc_charlen_type, const CHARTYPE *); |
| 51 |
export_proto(adjustr); |
| 52 |
|
| 53 |
extern gfc_charlen_type string_index (gfc_charlen_type, const CHARTYPE *, |
| 54 |
gfc_charlen_type, const CHARTYPE *, |
| 55 |
GFC_LOGICAL_4); |
| 56 |
export_proto(string_index); |
| 57 |
|
| 58 |
extern gfc_charlen_type string_scan (gfc_charlen_type, const CHARTYPE *, |
| 59 |
gfc_charlen_type, const CHARTYPE *, |
| 60 |
GFC_LOGICAL_4); |
| 61 |
export_proto(string_scan); |
| 62 |
|
| 63 |
extern gfc_charlen_type string_verify (gfc_charlen_type, const CHARTYPE *, |
| 64 |
gfc_charlen_type, const CHARTYPE *, |
| 65 |
GFC_LOGICAL_4); |
| 66 |
export_proto(string_verify); |
| 67 |
|
| 68 |
extern void string_trim (gfc_charlen_type *, CHARTYPE **, gfc_charlen_type, |
| 69 |
const CHARTYPE *); |
| 70 |
export_proto(string_trim); |
| 71 |
|
| 72 |
extern void string_minmax (gfc_charlen_type *, CHARTYPE **, int, int, ...); |
| 73 |
export_proto(string_minmax); |
| 74 |
|
| 75 |
|
| 76 |
/* Use for functions which can return a zero-length string. */ |
| 77 |
static CHARTYPE zero_length_string = 0; |
| 78 |
|
| 79 |
|
| 80 |
/* Strings of unequal length are extended with pad characters. */ |
| 81 |
|
| 82 |
int |
| 83 |
compare_string (gfc_charlen_type len1, const CHARTYPE *s1, |
| 84 |
gfc_charlen_type len2, const CHARTYPE *s2) |
| 85 |
{ |
| 86 |
const UCHARTYPE *s; |
| 87 |
gfc_charlen_type len; |
| 88 |
int res; |
| 89 |
|
| 90 |
/* Placate the sanitizer. */ |
| 91 |
if (!s1 && !s2) |
| 92 |
return 0; |
| 93 |
if (!s1) |
| 94 |
return -1; |
| 95 |
if (!s2) |
| 96 |
return 1; |
| 97 |
|
| 98 |
res = MEMCMP (s1, s2, ((len1 < len2) ? len1 : len2)); |
| 99 |
if (res != 0) |
| 100 |
return res; |
| 101 |
|
| 102 |
if (len1 == len2) |
| 103 |
return 0; |
| 104 |
|
| 105 |
if (len1 < len2) |
| 106 |
{ |
| 107 |
len = len2 - len1; |
| 108 |
s = (UCHARTYPE *) &s2[len1]; |
| 109 |
res = -1; |
| 110 |
} |
| 111 |
else |
| 112 |
{ |
| 113 |
len = len1 - len2; |
| 114 |
s = (UCHARTYPE *) &s1[len2]; |
| 115 |
res = 1; |
| 116 |
} |
| 117 |
|
| 118 |
while (len--) |
| 119 |
{ |
| 120 |
if (*s != ' ') |
| 121 |
{ |
| 122 |
if (*s > ' ') |
| 123 |
return res; |
| 124 |
else |
| 125 |
return -res; |
| 126 |
} |
| 127 |
s++; |
| 128 |
} |
| 129 |
|
| 130 |
return 0; |
| 131 |
} |
| 132 |
iexport(compare_string); |
| 133 |
|
| 134 |
|
| 135 |
/* The destination and source should not overlap. */ |
| 136 |
|
| 137 |
void |
| 138 |
concat_string (gfc_charlen_type destlen, CHARTYPE * dest, |
| 139 |
gfc_charlen_type len1, const CHARTYPE * s1, |
| 140 |
gfc_charlen_type len2, const CHARTYPE * s2) |
| 141 |
{ |
| 142 |
if (len1 >= destlen) |
| 143 |
{ |
| 144 |
memcpy (dest, s1, destlen * sizeof (CHARTYPE)); |
| 145 |
return; |
| 146 |
} |
| 147 |
memcpy (dest, s1, len1 * sizeof (CHARTYPE)); |
| 148 |
dest += len1; |
| 149 |
destlen -= len1; |
| 150 |
|
| 151 |
if (len2 >= destlen) |
| 152 |
{ |
| 153 |
memcpy (dest, s2, destlen * sizeof (CHARTYPE)); |
| 154 |
return; |
| 155 |
} |
| 156 |
|
| 157 |
memcpy (dest, s2, len2 * sizeof (CHARTYPE)); |
| 158 |
MEMSET (&dest[len2], ' ', destlen - len2); |
| 159 |
} |
| 160 |
|
| 161 |
|
| 162 |
/* Return string with all trailing blanks removed. */ |
| 163 |
|
| 164 |
void |
| 165 |
string_trim (gfc_charlen_type *len, CHARTYPE **dest, gfc_charlen_type slen, |
| 166 |
const CHARTYPE *src) |
| 167 |
{ |
| 168 |
*len = string_len_trim (slen, src); |
| 169 |
|
| 170 |
if (*len == 0) |
| 171 |
*dest = &zero_length_string; |
| 172 |
else |
| 173 |
{ |
| 174 |
/* Allocate space for result string. */ |
| 175 |
*dest = xmallocarray (*len, sizeof (CHARTYPE)); |
| 176 |
|
| 177 |
/* Copy string if necessary. */ |
| 178 |
memcpy (*dest, src, *len * sizeof (CHARTYPE)); |
| 179 |
} |
| 180 |
} |
| 181 |
|
| 182 |
|
| 183 |
/* The length of a string not including trailing blanks. */ |
| 184 |
|
| 185 |
gfc_charlen_type |
| 186 |
string_len_trim (gfc_charlen_type len, const CHARTYPE *s) |
| 187 |
{ |
| 188 |
if (len <= 0) |
| 189 |
return 0; |
| 190 |
|
| 191 |
const size_t long_len = sizeof (unsigned long); |
| 192 |
|
| 193 |
size_t i = len - 1; |
| 194 |
|
| 195 |
/* If we've got the standard (KIND=1) character type, we scan the string in |
| 196 |
long word chunks to speed it up (until a long word is hit that does not |
| 197 |
consist of ' 's). */ |
| 198 |
if (sizeof (CHARTYPE) == 1 && i >= long_len) |
| 199 |
{ |
| 200 |
size_t starting; |
| 201 |
unsigned long blank_longword; |
| 202 |
|
| 203 |
/* Handle the first characters until we're aligned on a long word |
| 204 |
boundary. Actually, s + i + 1 must be properly aligned, because |
| 205 |
s + i will be the last byte of a long word read. */ |
| 206 |
starting = ( |
| 207 |
#ifdef __INTPTR_TYPE__ |
| 208 |
(__INTPTR_TYPE__) |
| 209 |
#endif |
| 210 |
(s + i + 1)) % long_len; |
| 211 |
i -= starting; |
| 212 |
for (; starting > 0; --starting) |
| 213 |
if (s[i + starting] != ' ') |
| 214 |
return i + starting + 1; |
| 215 |
|
| 216 |
/* Handle the others in a batch until first non-blank long word is |
| 217 |
found. Here again, s + i is the last byte of the current chunk, |
| 218 |
to it starts at s + i - sizeof (long) + 1. */ |
| 219 |
|
| 220 |
#if __SIZEOF_LONG__ == 4 |
| 221 |
blank_longword = 0x20202020L; |
| 222 |
#elif __SIZEOF_LONG__ == 8 |
| 223 |
blank_longword = 0x2020202020202020L; |
| 224 |
#else |
| 225 |
#error Invalid size of long! |
| 226 |
#endif |
| 227 |
|
| 228 |
while (i >= long_len) |
| 229 |
{ |
| 230 |
i -= long_len; |
| 231 |
if (*((unsigned long*) (s + i + 1)) != blank_longword) |
| 232 |
{ |
| 233 |
i += long_len; |
| 234 |
break; |
| 235 |
} |
| 236 |
} |
| 237 |
} |
| 238 |
|
| 239 |
/* Simply look for the first non-blank character. */ |
| 240 |
while (s[i] == ' ') |
| 241 |
{ |
| 242 |
if (i == 0) |
| 243 |
return 0; |
| 244 |
--i; |
| 245 |
} |
| 246 |
return i + 1; |
| 247 |
} |
| 248 |
|
| 249 |
|
| 250 |
/* Find a substring within a string. */ |
| 251 |
|
| 252 |
gfc_charlen_type |
| 253 |
string_index (gfc_charlen_type slen, const CHARTYPE *str, |
| 254 |
gfc_charlen_type sslen, const CHARTYPE *sstr, |
| 255 |
GFC_LOGICAL_4 back) |
| 256 |
{ |
| 257 |
gfc_charlen_type start, last, delta, i; |
| 258 |
|
| 259 |
if (sslen == 0) |
| 260 |
return back ? (slen + 1) : 1; |
| 261 |
|
| 262 |
if (sslen > slen) |
| 263 |
return 0; |
| 264 |
|
| 265 |
if (!back) |
| 266 |
{ |
| 267 |
last = slen + 1 - sslen; |
| 268 |
start = 0; |
| 269 |
delta = 1; |
| 270 |
} |
| 271 |
else |
| 272 |
{ |
| 273 |
last = -1; |
| 274 |
start = slen - sslen; |
| 275 |
delta = -1; |
| 276 |
} |
| 277 |
|
| 278 |
for (; start != last; start+= delta) |
| 279 |
{ |
| 280 |
for (i = 0; i < sslen; i++) |
| 281 |
{ |
| 282 |
if (str[start + i] != sstr[i]) |
| 283 |
break; |
| 284 |
} |
| 285 |
if (i == sslen) |
| 286 |
return (start + 1); |
| 287 |
} |
| 288 |
return 0; |
| 289 |
} |
| 290 |
|
| 291 |
|
| 292 |
/* Remove leading blanks from a string, padding at end. The src and dest |
| 293 |
should not overlap. */ |
| 294 |
|
| 295 |
void |
| 296 |
adjustl (CHARTYPE *dest, gfc_charlen_type len, const CHARTYPE *src) |
| 297 |
{ |
| 298 |
gfc_charlen_type i; |
| 299 |
|
| 300 |
i = 0; |
| 301 |
while (i < len && src[i] == ' ') |
| 302 |
i++; |
| 303 |
|
| 304 |
if (i < len) |
| 305 |
memcpy (dest, &src[i], (len - i) * sizeof (CHARTYPE)); |
| 306 |
if (i > 0) |
| 307 |
MEMSET (&dest[len - i], ' ', i); |
| 308 |
} |
| 309 |
|
| 310 |
|
| 311 |
/* Remove trailing blanks from a string. */ |
| 312 |
|
| 313 |
void |
| 314 |
adjustr (CHARTYPE *dest, gfc_charlen_type len, const CHARTYPE *src) |
| 315 |
{ |
| 316 |
gfc_charlen_type i; |
| 317 |
|
| 318 |
i = len; |
| 319 |
while (i > 0 && src[i - 1] == ' ') |
| 320 |
i--; |
| 321 |
|
| 322 |
if (i < len) |
| 323 |
MEMSET (dest, ' ', len - i); |
| 324 |
memcpy (&dest[len - i], src, i * sizeof (CHARTYPE)); |
| 325 |
} |
| 326 |
|
| 327 |
|
| 328 |
/* Scan a string for any one of the characters in a set of characters. */ |
| 329 |
|
| 330 |
gfc_charlen_type |
| 331 |
string_scan (gfc_charlen_type slen, const CHARTYPE *str, |
| 332 |
gfc_charlen_type setlen, const CHARTYPE *set, GFC_LOGICAL_4 back) |
| 333 |
{ |
| 334 |
gfc_charlen_type i, j; |
| 335 |
|
| 336 |
if (slen == 0 || setlen == 0) |
| 337 |
return 0; |
| 338 |
|
| 339 |
if (back) |
| 340 |
{ |
| 341 |
for (i = slen; i != 0; i--) |
| 342 |
{ |
| 343 |
for (j = 0; j < setlen; j++) |
| 344 |
{ |
| 345 |
if (str[i - 1] == set[j]) |
| 346 |
return i; |
| 347 |
} |
| 348 |
} |
| 349 |
} |
| 350 |
else |
| 351 |
{ |
| 352 |
for (i = 0; i < slen; i++) |
| 353 |
{ |
| 354 |
for (j = 0; j < setlen; j++) |
| 355 |
{ |
| 356 |
if (str[i] == set[j]) |
| 357 |
return (i + 1); |
| 358 |
} |
| 359 |
} |
| 360 |
} |
| 361 |
|
| 362 |
return 0; |
| 363 |
} |
| 364 |
|
| 365 |
|
| 366 |
/* Verify that a set of characters contains all the characters in a |
| 367 |
string by identifying the position of the first character in a |
| 368 |
characters that does not appear in a given set of characters. */ |
| 369 |
|
| 370 |
gfc_charlen_type |
| 371 |
string_verify (gfc_charlen_type slen, const CHARTYPE *str, |
| 372 |
gfc_charlen_type setlen, const CHARTYPE *set, |
| 373 |
GFC_LOGICAL_4 back) |
| 374 |
{ |
| 375 |
gfc_charlen_type start, last, delta, i; |
| 376 |
|
| 377 |
if (slen == 0) |
| 378 |
return 0; |
| 379 |
|
| 380 |
if (back) |
| 381 |
{ |
| 382 |
last = -1; |
| 383 |
start = slen - 1; |
| 384 |
delta = -1; |
| 385 |
} |
| 386 |
else |
| 387 |
{ |
| 388 |
last = slen; |
| 389 |
start = 0; |
| 390 |
delta = 1; |
| 391 |
} |
| 392 |
for (; start != last; start += delta) |
| 393 |
{ |
| 394 |
for (i = 0; i < setlen; i++) |
| 395 |
{ |
| 396 |
if (str[start] == set[i]) |
| 397 |
break; |
| 398 |
} |
| 399 |
if (i == setlen) |
| 400 |
return (start + 1); |
| 401 |
} |
| 402 |
|
| 403 |
return 0; |
| 404 |
} |
| 405 |
|
| 406 |
|
| 407 |
/* MIN and MAX intrinsics for strings. The front-end makes sure that |
| 408 |
nargs is at least 2. */ |
| 409 |
|
| 410 |
void |
| 411 |
string_minmax (gfc_charlen_type *rlen, CHARTYPE **dest, int op, int nargs, ...) |
| 412 |
{ |
| 413 |
va_list ap; |
| 414 |
int i; |
| 415 |
CHARTYPE *next, *res; |
| 416 |
gfc_charlen_type nextlen, reslen; |
| 417 |
|
| 418 |
va_start (ap, nargs); |
| 419 |
reslen = va_arg (ap, gfc_charlen_type); |
| 420 |
res = va_arg (ap, CHARTYPE *); |
| 421 |
*rlen = reslen; |
| 422 |
|
| 423 |
if (res == NULL) |
| 424 |
runtime_error ("First argument of '%s' intrinsic should be present", |
| 425 |
op > 0 ? "MAX" : "MIN"); |
| 426 |
|
| 427 |
for (i = 1; i < nargs; i++) |
| 428 |
{ |
| 429 |
nextlen = va_arg (ap, gfc_charlen_type); |
| 430 |
next = va_arg (ap, CHARTYPE *); |
| 431 |
|
| 432 |
if (next == NULL) |
| 433 |
{ |
| 434 |
if (i == 1) |
| 435 |
runtime_error ("Second argument of '%s' intrinsic should be " |
| 436 |
"present", op > 0 ? "MAX" : "MIN"); |
| 437 |
else |
| 438 |
continue; |
| 439 |
} |
| 440 |
|
| 441 |
if (nextlen > *rlen) |
| 442 |
*rlen = nextlen; |
| 443 |
|
| 444 |
if (op * compare_string (reslen, res, nextlen, next) < 0) |
| 445 |
{ |
| 446 |
reslen = nextlen; |
| 447 |
res = next; |
| 448 |
} |
| 449 |
} |
| 450 |
va_end (ap); |
| 451 |
|
| 452 |
if (*rlen == 0) |
| 453 |
*dest = &zero_length_string; |
| 454 |
else |
| 455 |
{ |
| 456 |
CHARTYPE *tmp = xmallocarray (*rlen, sizeof (CHARTYPE)); |
| 457 |
memcpy (tmp, res, reslen * sizeof (CHARTYPE)); |
| 458 |
MEMSET (&tmp[reslen], ' ', *rlen - reslen); |
| 459 |
*dest = tmp; |
| 460 |
} |
| 461 |
} |