utils.c

   1 /* ################################################################### */
   2 /* Copyright 2015, Pierre Gentile (p.gen.progs@gmail.com)              */
   3 /*                                                                     */
   4 /* This Source Code Form is subject to the terms of the Mozilla Public */
   5 /* License, v. 2.0. If a copy of the MPL was not distributed with this */
   6 /* file, You can obtain one at https://mozilla.org/MPL/2.0/.           */
   7 /* ################################################################### */
   8
   9 /* ******************************** */
  10 /* Various small utility functions. */
  11 /* ******************************** */
  12
  13 #include "config.h"
  14 #include <stddef.h>
  15 #include <stdio.h>
  16 #include <stdlib.h>
  17 #include <limits.h>
  18 #include <string.h>
  19 #include <ctype.h>
  20 #include <stdarg.h>
  21 #include <wctype.h>
  22 #include "xmalloc.h"
  23 #include "wchar.h"
  24 #include "list.h"
  25 #include "utf8.h"
  26 #include "utils.h"
  27
  28 /* ******************* */
  29 /* Interval functions. */
  30 /* ******************* */
  31
  32 /* ======================= */
  33 /* Creates a new interval. */
  34 /* ======================= */
  35 interval_t *
  36 interval_new(void)
  37 {
  38   return xmalloc(sizeof(interval_t));
  39 }
  40
  41 /* ======================================= */
  42 /* Compares 2 intervals as integer couples */
  43 /* same return values as for strcmp.       */
  44 /* ======================================= */
  45 int
  46 interval_comp(void const *a, void const *b)
  47 {
  48   interval_t const *ia = (interval_t *)a;
  49   interval_t const *ib = (interval_t *)b;
  50
  51   if (ia->low < ib->low)
  52     /* ia: [...      */
  53     /* ib:      [... */
  54     return -1;
  55   if (ia->low > ib->low)
  56     /* ia:      [... */
  57     /* ib: [...      */
  58     return 1;
  59   if (ia->high < ib->high)
  60     /* ia: ...]      */
  61     /* ib:      ...] */
  62     return -1;
  63   if (ia->high > ib->high)
  64     /* ia:      ...] */
  65     /* ib: ...]      */
  66     return 1;
  67
  68   return 0;
  69 }
  70
  71 /* ================================== */
  72 /* Swaps the values of two intervals. */
  73 /* ================================== */
  74 void
  75 interval_swap(void **a, void **b)
  76 {
  77   interval_t *ia = (interval_t *)*a;
  78   interval_t *ib = (interval_t *)*b;
  79   long        tmp;
  80
  81   tmp     = ia->low;
  82   ia->low = ib->low;
  83   ib->low = tmp;
  84
  85   tmp      = ia->high;
  86   ia->high = ib->high;
  87   ib->high = tmp;
  88 }
  89
  90 /* ====================================================================== */
  91 /* Merges the intervals from an interval list in order to get the minimum */
  92 /* number of intervals to consider.                                       */
  93 /* ====================================================================== */
  94 void
  95 optimize_an_interval_list(ll_t *list)
  96 {
  97   ll_node_t  *node1, *node2;
  98   interval_t *data1, *data2;
  99
 100   if (!list || list->len < 2)
 101     return;
 102
 103   /* Step 1: sort the intervals list. */
 104   /* """""""""""""""""""""""""""""""" */
 105   ll_sort(list, interval_comp, interval_swap);
 106
 107   /* Step 2: merge the list by merging the consecutive intervals. */
 108   /* """""""""""""""""""""""""""""""""""""""""""""""""""""""""""" */
 109   node1 = list->head;
 110   node2 = node1->next;
 111
 112   while (node2)
 113   {
 114     data1 = (interval_t *)(node1->data);
 115     data2 = (interval_t *)(node2->data);
 116
 117     if (data1->high >= data2->low - 1)
 118     {
 119       /* Interval 1 overlaps interval 2. */
 120       /* ''''''''''''''''''''''''''''''' */
 121       if (data2->high >= data1->high)
 122         data1->high = data2->high;
 123       ll_delete(list, node2);
 124       free(data2);
 125       node2 = node1->next;
 126     }
 127     else
 128     {
 129       /* No overlap. */
 130       /* ''''''''''' */
 131       node1 = node2;
 132       node2 = node2->next;
 133     }
 134   }
 135 }
 136
 137 /* ***************** */
 138 /* String functions. */
 139 /* ***************** */
 140
 141 /* ========================================================================= */
 142 /* Allocates memory and safely concatenate strings. Stolen from a public     */
 143 /* domain implementation which can be found here:                            */
 144 /* http://openwall.info/wiki/people/solar/software/public-domain-source-code */
 145 /* ========================================================================= */
 146 char *
 147 concat(const char *s1, ...)
 148 {
 149   va_list     args;
 150   const char *s;
 151   char       *p, *result;
 152   size_t      l, m, n;
 153
 154   m = n = strlen(s1);
 155   va_start(args, s1);
 156   while ((s = va_arg(args, char *)))
 157   {
 158     l = strlen(s);
 159     if ((m += l) < l)
 160       break;
 161   }
 162   va_end(args);
 163   if (s || m >= INT_MAX)
 164     return NULL;
 165
 166   result = (char *)xmalloc(m + 1);
 167
 168   memcpy(p = result, s1, n);
 169   p += n;
 170   va_start(args, s1);
 171   while ((s = va_arg(args, char *)))
 172   {
 173     l = strlen(s);
 174     if ((n += l) < l || n > m)
 175       break;
 176     memcpy(p, s, l);
 177     p += l;
 178   }
 179   va_end(args);
 180   if (s || m != n || p != result + n)
 181   {
 182     free(result);
 183     return NULL;
 184   }
 185
 186   *p = 0;
 187   return result;
 188 }
 189
 190 /* =============================================== */
 191 /* Is the string str2 a prefix of the string str1? */
 192 /* Returns 1 if true, else 0.                      */
 193 /* =============================================== */
 194 int
 195 strprefix(char *str1, char *str2)
 196 {
 197   while (*str1 != '\0' && *str1 == *str2)
 198   {
 199     str1++;
 200     str2++;
 201   }
 202
 203   return *str2 == '\0';
 204 }
 205
 206 /* ========================= */
 207 /* Trims leading characters. */
 208 /* ========================= */
 209 void
 210 ltrim(char *str, const char *trim_str)
 211 {
 212   size_t len   = strlen(str);
 213   size_t begin = strspn(str, trim_str);
 214
 215   if (begin > 0)
 216     for (size_t i = begin; i <= len; ++i)
 217       str[i - begin] = str[i];
 218 }
 219
 220 /* ==================================================================== */
 221 /* Trims trailing characters.                                           */
 222 /* All (ASCII) characters in trim_str will be removed.                  */
 223 /* The min argument guarantees that the length of the resulting string  */
 224 /* will not be smaller than this size if it was larger before, 0 is the */
 225 /* usual value here.                                                    */
 226 /* Note that when min is greater than 0, tail characters intended to be */
 227 /* deleted may remain.                                                  */
 228 /* ==================================================================== */
 229 void
 230 rtrim(char *str, const char *trim_str, size_t min)
 231 {
 232   size_t len = strlen(str);
 233   while (len > min && strchr(trim_str, str[len - 1]))
 234     str[--len] = '\0';
 235 }
 236
 237 /* ========================================= */
 238 /* Case insensitive strcmp.                  */
 239 /* from http://c.snippets.org/code/stricmp.c */
 240 /* ========================================= */
 241 int
 242 my_strcasecmp(const char *str1, const char *str2)
 243 {
 244 #ifdef HAVE_STRCASECMP
 245   return strcasecmp(str1, str2);
 246 #else
 247   int retval = 0;
 248
 249   while (1)
 250   {
 251     retval = tolower(*str1++) - tolower(*str2++);
 252
 253     if (retval)
 254       break;
 255
 256     if (*str1 && *str2)
 257       continue;
 258     else
 259       break;
 260   }
 261   return retval;
 262 #endif
 263 }
 264
 265 /* ============================================= */
 266 /* memmove based strcpy (tolerates overlapping). */
 267 /* ============================================= */
 268 char *
 269 my_strcpy(char *str1, char *str2)
 270 {
 271   if (str1 == NULL || str2 == NULL)
 272     return NULL;
 273
 274   memmove(str1, str2, strlen(str2) + 1);
 275
 276   return str1;
 277 }
 278
 279 /* ================================ */
 280 /* 7 bits aware version of isprint. */
 281 /* ================================ */
 282 int
 283 isprint7(int i)
 284 {
 285   return i >= 0x20 && i <= 0x7e;
 286 }
 287
 288 /* ================================ */
 289 /* 8 bits aware version of isprint. */
 290 /* ================================ */
 291 int
 292 isprint8(int i)
 293 {
 294   unsigned char c = i & (unsigned char)0xff;
 295
 296   return (c >= 0x20 && c < 0x7f) || (c >= (unsigned char)0xa0);
 297 }
 298
 299 /* ==================================================== */
 300 /* Private implementation of wcscasecmp missing in c99. */
 301 /* ==================================================== */
 302 int
 303 my_wcscasecmp(const wchar_t *s1, const wchar_t *s2)
 304 {
 305   wchar_t c1, c2;
 306
 307   while (*s1)
 308   {
 309     c1 = towlower(*s1);
 310     c2 = towlower(*s2);
 311
 312     if (c1 != c2)
 313       return (int)(c1 - c2);
 314
 315     s1++;
 316     s2++;
 317   }
 318   return -*s2;
 319 }
 320
 321 /* ================================================================ */
 322 /* Returns 1 if s can be converted into an int otherwise returns 0. */
 323 /* ================================================================ */
 324 int
 325 is_integer(const char * const s)
 326 {
 327   long int n;
 328   char    *endptr;
 329
 330   n = strtol(s, &endptr, 10);
 331
 332   if (errno != ERANGE && n >= INT_MIN && n <= INT_MAX && *endptr == '\0')
 333     return 1;
 334
 335   return 0;
 336 }
 337
 338 /* ===================================================== */
 339 /* Exchanges the start and end part of a string.         */
 340 /* The first part goes from char 0 to size-1.            */
 341 /* The second part goes from char size to the end of *s. */
 342 /* Returns 1 on success.                                 */
 343 /* ===================================================== */
 344 int
 345 swap_string_parts(char **s, size_t first)
 346 {
 347   char  *tmp;
 348   size_t size;
 349
 350   if (*s == NULL || **s == '\0')
 351     return 0;
 352
 353   tmp  = xmalloc(strlen(*s) * 2 + 1);
 354   size = strlen(*s);
 355
 356   if (first > size)
 357     return 0;
 358
 359   strcpy(tmp, *s);
 360   strcat(tmp, *s);
 361   strncpy(*s, tmp + first, size);
 362
 363   free(tmp);
 364   return 1;
 365 }
 366
 367 /* ================================================================ */
 368 /* Substitute all the characters c1 by c2 in the string s in place. */
 369 /* ================================================================ */
 370 void
 371 strrep(char *s, const char c1, const char c2)
 372 {
 373   if (s != NULL)
 374     while (*s)
 375     {
 376       if (*s == c1)
 377         *s = c2;
 378       s++;
 379     }
 380 }
 381
 382 /* ================================================================== */
 383 /* Allocates and returns a string similar to s but with non printable */
 384 /* character changed by their ASCII hexadecimal notation.             */
 385 /* ================================================================== */
 386 char *
 387 strprint(char const *s)
 388 {
 389   size_t l  = strlen(s);
 390   char *new = xcalloc(1, 4 * l + 1);
 391   char *p   = new;
 392
 393   while (*s)
 394   {
 395     if (isprint(*s))
 396       *(p++) = *s++;
 397     else
 398     {
 399       sprintf(p, "\\x%02X", (unsigned char)*s++);
 400       p += 4;
 401     }
 402   }
 403
 404   if (p - new > (ptrdiff_t)l)
 405     new = xrealloc(new, p - new + 1);
 406
 407   return new;
 408 }
 409
 410 /* =============================================== */
 411 /* Hexadecimal dump of part of a buffer to a file. */
 412 /*                                                 */
 413 /* buf   : buffer to dump.                         */
 414 /* fp    : file to dump to.                        */
 415 /* prefix: string to be printed before each line.  */
 416 /* size  : length of the buffer to consider.       */
 417 /* =============================================== */
 418 void
 419 hexdump(const char *buf, FILE *fp, const char *prefix, size_t size)
 420 {
 421   unsigned int  b;
 422   unsigned char d[17];
 423   unsigned int  o, mo;
 424   size_t        l;
 425
 426   o = mo = 0;
 427   l      = strlen(prefix);
 428
 429   memset(d, '\0', 17);
 430   for (b = 0; b < size; b++)
 431   {
 432
 433     d[b % 16] = isprint(buf[b]) ? (unsigned char)buf[b] : '.';
 434
 435     if ((b % 16) == 0)
 436     {
 437       o = l + 7;
 438       if (o > mo)
 439         mo = o;
 440       fprintf(fp, "%s: %04x:", prefix, b);
 441     }
 442
 443     o += 3;
 444     if (o > mo)
 445       mo = o;
 446     fprintf(fp, " %02x", (unsigned char)buf[b]);
 447
 448     if ((b % 16) == 15)
 449     {
 450       mo = o;
 451       o  = 0;
 452       fprintf(fp, " |%s|", d);
 453       memset(d, '\0', 17);
 454       fprintf(fp, "\n");
 455     }
 456   }
 457   if ((b % 16) != 0)
 458   {
 459     for (unsigned int i = 0; i < mo - o; i++)
 460       fprintf(fp, "%c", ' ');
 461
 462     fprintf(fp, " |%s", d);
 463     if (mo > o)
 464       for (unsigned int i = 0; i < 16 - strlen((char *)d); i++)
 465         fprintf(fp, "%c", ' ');
 466     fprintf(fp, "%c", '|');
 467     memset(d, '\0', 17);
 468     fprintf(fp, "\n");
 469   }
 470 }
 471
 472 /* ===================================================================== */
 473 /* Version of wcswidth which tries to support extended grapheme clusters */
 474 /* by taking into zero width characters.                                 */
 475 /* ===================================================================== */
 476 int
 477 my_wcswidth(const wchar_t *s, size_t n)
 478 {
 479   int len = 0;
 480   int l   = 0;
 481   int m   = 0;
 482
 483   if (s == NULL || *s == L'\0')
 484     return 0;
 485
 486   while (*s && m < n)
 487   {
 488     if ((l = wcwidth(*s)) >= 0)
 489     {
 490       /* Do not count zero-width-length glyphs. */
 491       /* """""""""""""""""""""""""""""""""""""" */
 492       if (*s != L'\x200d' && *(s + 1) != L'\x200d' && *(s + 1) != L'\xfe0f'
 493           && *(s + 1) != L'\x20e3')
 494         len += l;
 495     }
 496     else
 497       return -1; /* wcwidth returned -1. */
 498
 499     s++;
 500     m++;
 501   }
 502
 503   return len;
 504 }