ircd/ircd_string.c

   1 /*
   2  * IRC - Internet Relay Chat, ircd/ircd_string.c
   3  * Copyright (C) 1999 Thomas Helvey
   4  *
   5  * This program is free software; you can redistribute it and/or modify
   6  * it under the terms of the GNU General Public License as published by
   7  * the Free Software Foundation; either version 1, or (at your option)
   8  * any later version.
   9  *
  10  * This program is distributed in the hope that it will be useful,
  11  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  12  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  13  * GNU General Public License for more details.
  14  *
  15  * You should have received a copy of the GNU General Public License
  16  * along with this program; if not, write to the Free Software
  17  * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
  18  */
  19 /** @file
  20  * @brief Implementation of string operations.
  21  * @version $Id: ircd_string.c,v 1.24.2.2 2007/08/21 02:02:10 entrope Exp $
  22  */
  23 #include "config.h"
  24
  25 #include "ircd_string.h"
  26 #include "ircd_defs.h"
  27 #include "ircd_chattr.h"
  28 #include "ircd_log.h"
  29 #include "res.h"
  30
  31 /* #include <assert.h> -- Now using assert in ircd_log.h */
  32 #include <string.h>
  33 #include <sys/types.h>
  34 #include <netinet/in.h>
  35
  36 /*
  37  * include the character attribute tables here
  38  */
  39 #include "chattr.tab.c"
  40
  41 /** Check whether \a str contains wildcard characters.
  42  * @param[in] str String that might contain wildcards.
  43  * @return Non-zero if \a str contains naked (non-escaped) wildcards,
  44  * zero if there are none or if they are all escaped.
  45  */
  46 int string_has_wildcards(const char* str)
  47 {
  48   assert(0 != str);
  49   for ( ; *str; ++str) {
  50     if ('\\' == *str) {
  51       if ('\0' == *++str)
  52         break;
  53     }
  54     else if ('*' == *str || '?' == *str)
  55       return 1;
  56   }
  57   return 0;
  58 }
  59
  60 /** Split a string on certain delimiters.
  61  * This is a reentrant version of normal strtok().  The first call for
  62  * a particular input string must use a non-NULL \a str; *save will be
  63  * initialized based on that.  Later calls must use a NULL \a str;
  64  * *save will be updated.
  65  * @param[in,out] save Pointer to a position indicator.
  66  * @param[in] str Pointer to the input string, or NULL to continue.
  67  * @param[in] fs String that lists token delimiters.
  68  * @return Next token in input string, or NULL if no tokens remain.
  69  */
  70 char* ircd_strtok(char **save, char *str, char *fs)
  71 {
  72   char *pos = *save;            /* keep last position across calls */
  73   char *tmp;
  74
  75   if (str)
  76     pos = str;                  /* new string scan */
  77
  78   while (pos && *pos && strchr(fs, *pos) != NULL)
  79     pos++;                      /* skip leading separators */
  80
  81   if (!pos || !*pos)
  82     return (pos = *save = NULL);        /* string contains only sep's */
  83
  84   tmp = pos;                    /* now, keep position of the token */
  85
  86   while (*pos && strchr(fs, *pos) == NULL)
  87     pos++;                      /* skip content of the token */
  88
  89   if (*pos)
  90     *pos++ = '\0';              /* remove first sep after the token */
  91   else
  92     pos = NULL;                 /* end of string */
  93
  94   *save = pos;
  95   return (tmp);
  96 }
  97
  98 /** Rewrite a comma-delimited list of items to remove duplicates.
  99  * @param[in,out] buffer Comma-delimited list.
 100  * @return The input buffer \a buffer.
 101  */
 102 char* canonize(char* buffer)
 103 {
 104   static char cbuf[BUFSIZE];
 105   char*       s;
 106   char*       t;
 107   char*       cp = cbuf;
 108   int         l = 0;
 109   char*       p = NULL;
 110   char*       p2;
 111
 112   *cp = '\0';
 113
 114   for (s = ircd_strtok(&p, buffer, ","); s; s = ircd_strtok(&p, NULL, ","))
 115   {
 116     if (l)
 117     {
 118       p2 = NULL;
 119       for (t = ircd_strtok(&p2, cbuf, ","); t; t = ircd_strtok(&p2, NULL, ","))
 120         if (0 == ircd_strcmp(s, t))
 121           break;
 122         else if (p2)
 123           p2[-1] = ',';
 124     }
 125     else
 126       t = NULL;
 127     if (!t)
 128     {
 129       if (l)
 130         *(cp - 1) = ',';
 131       else
 132         l = 1;
 133       strcpy(cp, s);
 134       if (p)
 135         cp += (p - s);
 136     }
 137     else if (p2)
 138       p2[-1] = ',';
 139   }
 140   return cbuf;
 141 }
 142
 143 /** Copy one string to another, not to exceed a certain length.
 144  * @param[in] s1 Output buffer.
 145  * @param[in] s2 Source buffer.
 146  * @param[in] n Maximum number of bytes to write, plus one.
 147  * @return The original input buffer \a s1.
 148  */
 149 char* ircd_strncpy(char* s1, const char* s2, size_t n)
 150 {
 151   char* endp = s1 + n;
 152   char* s = s1;
 153
 154   assert(0 != s1);
 155   assert(0 != s2);
 156
 157   while (s < endp && (*s++ = *s2++))
 158     ;
 159   if (s == endp)
 160     *s = '\0';
 161   return s1;
 162 }
 163
 164
 165 #ifndef FORCEINLINE
 166 NTL_HDR_strChattr { NTL_SRC_strChattr }
 167 NTL_HDR_strCasediff { NTL_SRC_strCasediff }
 168 #endif /* !FORCEINLINE */
 169
 170 /*
 171  * Other functions visible externally
 172  */
 173
 174 /** Case insensitive string comparison.
 175  * @param[in] a First string to compare.
 176  * @param[in] b Second string to compare.
 177  * @return Less than, equal to, or greater than zero if \a a is lexicographically less than, equal to, or greater than \a b.
 178  */
 179 int ircd_strcmp(const char *a, const char *b)
 180 {
 181   const char* ra = a;
 182   const char* rb = b;
 183   while (ToLower(*ra) == ToLower(*rb)) {
 184     if (!*ra++)
 185       return 0;
 186     else
 187       ++rb;
 188   }
 189   return (ToLower(*ra) - ToLower(*rb));
 190 }
 191
 192 /** Case insensitive comparison of the starts of two strings.
 193  * @param[in] a First string to compare.
 194  * @param[in] b Second string to compare.
 195  * @param[in] n Maximum number of characters to compare.
 196  * @return Less than, equal to, or greater than zero if \a a is
 197  * lexicographically less than, equal to, or greater than \a b.
 198  */
 199 int ircd_strncmp(const char *a, const char *b, size_t n)
 200 {
 201   const char* ra = a;
 202   const char* rb = b;
 203   int left = n;
 204   if (!left--)
 205     return 0;
 206   while (ToLower(*ra) == ToLower(*rb)) {
 207     if (!*ra++ || !left--)
 208       return 0;
 209     else
 210       ++rb;
 211   }
 212   return (ToLower(*ra) - ToLower(*rb));
 213 }
 214
 215 /** Fill a vector of distinct names from a delimited input list.
 216  * Empty tokens (when \a token occurs at the start or end of \a list,
 217  * or when \a token occurs adjacent to itself) are ignored.  When
 218  * \a size tokens have been written to \a vector, the rest of the
 219  * string is ignored.
 220  * Unlike token_vector(), if a token repeats an earlier token, it is
 221  * skipped.
 222  * @param[in,out] names Input buffer.
 223  * @param[in] token Delimiter used to split \a list.
 224  * @param[out] vector Output vector.
 225  * @param[in] size Maximum number of elements to put in \a vector.
 226  * @return Number of elements written to \a vector.
 227  */
 228 int unique_name_vector(char* names, char token, char** vector, int size)
 229 {
 230   int   i;
 231   int   count = 0;
 232   char* start = names;
 233   char* end;
 234
 235   assert(0 != names);
 236   assert(0 != vector);
 237   assert(0 < size);
 238
 239   /*
 240    * ignore spurious tokens
 241    */
 242   while (token == *start)
 243     ++start;
 244
 245   for (end = strchr(start, token); end; end = strchr(start, token)) {
 246     *end++ = '\0';
 247     /*
 248      * ignore spurious tokens
 249      */
 250     while (token == *end)
 251       ++end;
 252     for (i = 0; i < count; ++i) {
 253       if (0 == ircd_strcmp(vector[i], start))
 254         break;
 255     }
 256     if (i == count) {
 257       vector[count++] = start;
 258       if (count == size)
 259         return count;
 260     }
 261     start = end;
 262   }
 263   if (*start) {
 264     for (i = 0; i < count; ++i)
 265       if (0 == ircd_strcmp(vector[i], start))
 266         return count;
 267     vector[count++] = start;
 268   }
 269   return count;
 270 }
 271
 272 /** Fill a vector of tokens from a delimited input list.
 273  * Empty tokens (when \a token occurs at the start or end of \a list,
 274  * or when \a token occurs adjacent to itself) are ignored.  When
 275  * \a size tokens have been written to \a vector, the rest of the
 276  * string is ignored.
 277  * @param[in,out] names Input buffer.
 278  * @param[in] token Delimiter used to split \a list.
 279  * @param[out] vector Output vector.
 280  * @param[in] size Maximum number of elements to put in \a vector.
 281  * @return Number of elements written to \a vector.
 282  */
 283 int token_vector(char* names, char token, char** vector, int size)
 284 {
 285   int   count = 0;
 286   char* start = names;
 287   char* end;
 288
 289   assert(0 != names);
 290   assert(0 != vector);
 291   assert(1 < size);
 292
 293   vector[count++] = start;
 294   for (end = strchr(start, token); end; end = strchr(start, token)) {
 295     *end++ = '\0';
 296     start = end;
 297     if (*start) {
 298       vector[count++] = start;
 299       if (count < size)
 300         continue;
 301     }
 302     break;
 303   }
 304   return count;
 305 }
 306
 307 /** Copy all or part of the hostname in a string to another string.
 308  * If \a userhost contains an '\@', the remaining portion is used;
 309  * otherwise, the whole \a userhost is used.
 310  * @param[out] buf Output buffer.
 311  * @param[in] userhost user\@hostname or hostname string.
 312  * @param[in] len Maximum number of bytes to write to \a host.
 313  * @return The output buffer \a buf.
 314  */
 315 char* host_from_uh(char* buf, const char* userhost, size_t len)
 316 {
 317   const char* s;
 318
 319   assert(0 != buf);
 320   assert(0 != userhost);
 321
 322   if ((s = strchr(userhost, '@')))
 323     ++s;
 324   else
 325     s = userhost;
 326   ircd_strncpy(buf, s, len);
 327   buf[len] = '\0';
 328   return buf;
 329 }
 330
 331 /*
 332  * this new faster inet_ntoa was ripped from:
 333  * From: Thomas Helvey <tomh@inxpress.net>
 334  */
 335 /** Array of text strings for dotted quads. */
 336 static const char* IpQuadTab[] =
 337 {
 338     "0",   "1",   "2",   "3",   "4",   "5",   "6",   "7",   "8",   "9",
 339    "10",  "11",  "12",  "13",  "14",  "15",  "16",  "17",  "18",  "19",
 340    "20",  "21",  "22",  "23",  "24",  "25",  "26",  "27",  "28",  "29",
 341    "30",  "31",  "32",  "33",  "34",  "35",  "36",  "37",  "38",  "39",
 342    "40",  "41",  "42",  "43",  "44",  "45",  "46",  "47",  "48",  "49",
 343    "50",  "51",  "52",  "53",  "54",  "55",  "56",  "57",  "58",  "59",
 344    "60",  "61",  "62",  "63",  "64",  "65",  "66",  "67",  "68",  "69",
 345    "70",  "71",  "72",  "73",  "74",  "75",  "76",  "77",  "78",  "79",
 346    "80",  "81",  "82",  "83",  "84",  "85",  "86",  "87",  "88",  "89",
 347    "90",  "91",  "92",  "93",  "94",  "95",  "96",  "97",  "98",  "99",
 348   "100", "101", "102", "103", "104", "105", "106", "107", "108", "109",
 349   "110", "111", "112", "113", "114", "115", "116", "117", "118", "119",
 350   "120", "121", "122", "123", "124", "125", "126", "127", "128", "129",
 351   "130", "131", "132", "133", "134", "135", "136", "137", "138", "139",
 352   "140", "141", "142", "143", "144", "145", "146", "147", "148", "149",
 353   "150", "151", "152", "153", "154", "155", "156", "157", "158", "159",
 354   "160", "161", "162", "163", "164", "165", "166", "167", "168", "169",
 355   "170", "171", "172", "173", "174", "175", "176", "177", "178", "179",
 356   "180", "181", "182", "183", "184", "185", "186", "187", "188", "189",
 357   "190", "191", "192", "193", "194", "195", "196", "197", "198", "199",
 358   "200", "201", "202", "203", "204", "205", "206", "207", "208", "209",
 359   "210", "211", "212", "213", "214", "215", "216", "217", "218", "219",
 360   "220", "221", "222", "223", "224", "225", "226", "227", "228", "229",
 361   "230", "231", "232", "233", "234", "235", "236", "237", "238", "239",
 362   "240", "241", "242", "243", "244", "245", "246", "247", "248", "249",
 363   "250", "251", "252", "253", "254", "255"
 364 };
 365
 366 /** Convert an IP address to printable ASCII form.
 367  * This is generally deprecated in favor of ircd_ntoa_r().
 368  * @param[in] in Address to convert.
 369  * @return Pointer to a static buffer containing the readable form.
 370  */
 371 const char* ircd_ntoa(const struct irc_in_addr* in)
 372 {
 373   static char buf[SOCKIPLEN];
 374   return ircd_ntoa_r(buf, in);
 375 }
 376
 377 /** Convert an IP address to printable ASCII form.
 378  * @param[out] buf Output buffer to write to.
 379  * @param[in] in Address to format.
 380  * @return Pointer to the output buffer \a buf.
 381  */
 382 const char* ircd_ntoa_r(char* buf, const struct irc_in_addr* in)
 383 {
 384     assert(buf != NULL);
 385     assert(in != NULL);
 386
 387     if (irc_in_addr_is_ipv4(in)) {
 388       unsigned int pos, len;
 389       unsigned char *pch;
 390
 391       pch = (unsigned char*)&in->in6_16[6];
 392       len = strlen(IpQuadTab[*pch]);
 393       memcpy(buf, IpQuadTab[*pch++], len);
 394       pos = len;
 395       buf[pos++] = '.';
 396       len = strlen(IpQuadTab[*pch]);
 397       memcpy(buf+pos, IpQuadTab[*pch++], len);
 398       pos += len;
 399       buf[pos++] = '.';
 400       len = strlen(IpQuadTab[*pch]);
 401       memcpy(buf+pos, IpQuadTab[*pch++], len);
 402       pos += len;
 403       buf[pos++] = '.';
 404       len = strlen(IpQuadTab[*pch]);
 405       memcpy(buf+pos, IpQuadTab[*pch++], len);
 406       buf[pos + len] = '\0';
 407       return buf;
 408     } else {
 409       static const char hexdigits[] = "0123456789abcdef";
 410       unsigned int pos, part, max_start, max_zeros, curr_zeros, ii;
 411
 412       /* Find longest run of zeros. */
 413       for (max_start = ii = 1, max_zeros = curr_zeros = 0; ii < 8; ++ii) {
 414         if (!in->in6_16[ii])
 415           curr_zeros++;
 416         else if (curr_zeros > max_zeros) {
 417           max_start = ii - curr_zeros;
 418           max_zeros = curr_zeros;
 419           curr_zeros = 0;
 420         }
 421       }
 422       if (curr_zeros > max_zeros) {
 423         max_start = ii - curr_zeros;
 424         max_zeros = curr_zeros;
 425       }
 426
 427       /* Print out address. */
 428 /** Append \a CH to the output buffer. */
 429 #define APPEND(CH) do { buf[pos++] = (CH); } while (0)
 430       for (pos = ii = 0; (ii < 8); ++ii) {
 431         if ((max_zeros > 0) && (ii == max_start)) {
 432           APPEND(':');
 433           ii += max_zeros - 1;
 434           continue;
 435         }
 436         part = ntohs(in->in6_16[ii]);
 437         if (part >= 0x1000)
 438           APPEND(hexdigits[part >> 12]);
 439         if (part >= 0x100)
 440           APPEND(hexdigits[(part >> 8) & 15]);
 441         if (part >= 0x10)
 442           APPEND(hexdigits[(part >> 4) & 15]);
 443         APPEND(hexdigits[part & 15]);
 444         if (ii < 7)
 445           APPEND(':');
 446       }
 447 #undef APPEND
 448
 449       /* Nul terminate and return number of characters used. */
 450       buf[pos++] = '\0';
 451       return buf;
 452     }
 453 }
 454
 455 /** Attempt to parse an IPv4 address into a network-endian form.
 456  * @param[in] input Input string.
 457  * @param[out] output Network-endian representation of the address.
 458  * @param[out] pbits Number of bits found in pbits.
 459  * @return Number of characters used from \a input, or 0 if the parse failed.
 460  */
 461 static unsigned int
 462 ircd_aton_ip4(const char *input, unsigned int *output, unsigned char *pbits)
 463 {
 464   unsigned int dots = 0, pos = 0, part = 0, ip = 0, bits;
 465
 466   /* Intentionally no support for bizarre IPv4 formats (plain
 467    * integers, octal or hex components) -- only vanilla dotted
 468    * decimal quads.
 469    */
 470   if (input[0] == '.')
 471     return 0;
 472   bits = 32;
 473   while (1) switch (input[pos]) {
 474   case '\0':
 475     if (dots < 3)
 476       return 0;
 477   out:
 478     ip |= part << (24 - 8 * dots);
 479     *output = htonl(ip);
 480     if (pbits)
 481       *pbits = bits;
 482     return pos;
 483   case '.':
 484     if (input[++pos] == '.')
 485       return 0;
 486     ip |= part << (24 - 8 * dots++);
 487     part = 0;
 488     if (input[pos] == '*') {
 489       while (input[++pos] == '*' || input[pos] == '.') ;
 490       if (input[pos] != '\0')
 491         return 0;
 492       if (pbits)
 493         *pbits = dots * 8;
 494       *output = htonl(ip);
 495       return pos;
 496     }
 497     break;
 498   case '/':
 499     if (!pbits || !IsDigit(input[pos + 1]))
 500       return 0;
 501     for (bits = 0; IsDigit(input[++pos]); )
 502       bits = bits * 10 + input[pos] - '0';
 503     if (bits > 32)
 504       return 0;
 505     goto out;
 506   case '0': case '1': case '2': case '3': case '4':
 507   case '5': case '6': case '7': case '8': case '9':
 508     part = part * 10 + input[pos++] - '0';
 509     if (part > 255)
 510       return 0;
 511     break;
 512   default:
 513     return 0;
 514   }
 515 }
 516
 517 /** Parse a numeric IPv4 or IPv6 address into an irc_in_addr.
 518  * @param[in] input Input buffer.
 519  * @param[out] ip Receives parsed IP address.
 520  * @param[out] pbits If non-NULL, receives number of bits specified in address mask.
 521  * @return Number of characters used from \a input, or 0 if the
 522  * address was unparseable or malformed.
 523  */
 524 int
 525 ipmask_parse(const char *input, struct irc_in_addr *ip, unsigned char *pbits)
 526 {
 527   char *colon;
 528   char *dot;
 529
 530   assert(ip);
 531   assert(input);
 532   memset(ip, 0, sizeof(*ip));
 533   colon = strchr(input, ':');
 534   dot = strchr(input, '.');
 535
 536   if (colon && (!dot || (dot > colon))) {
 537     unsigned int part = 0, pos = 0, ii = 0, colon = 8;
 538     const char *part_start = NULL;
 539
 540     /* Parse IPv6, possibly like ::127.0.0.1.
 541      * This is pretty straightforward; the only trick is borrowed
 542      * from Paul Vixie (BIND): when it sees a "::" continue as if
 543      * it were a single ":", but note where it happened, and fill
 544      * with zeros afterward.
 545      */
 546     if (input[pos] == ':') {
 547       if ((input[pos+1] != ':') || (input[pos+2] == ':'))
 548         return 0;
 549       colon = 0;
 550       pos += 2;
 551       part_start = input + pos;
 552     }
 553     while (ii < 8) switch (input[pos]) {
 554       unsigned char chval;
 555     case '0': case '1': case '2': case '3': case '4':
 556     case '5': case '6': case '7': case '8': case '9':
 557       chval = input[pos] - '0';
 558     use_chval:
 559       part = (part << 4) | chval;
 560       if (part > 0xffff)
 561         return 0;
 562       pos++;
 563       break;
 564     case 'A': case 'B': case 'C': case 'D': case 'E': case 'F':
 565       chval = input[pos] - 'A' + 10;
 566       goto use_chval;
 567     case 'a': case 'b': case 'c': case 'd': case 'e': case 'f':
 568       chval = input[pos] - 'a' + 10;
 569       goto use_chval;
 570     case ':':
 571       part_start = input + ++pos;
 572       if (input[pos] == '.')
 573         return 0;
 574       ip->in6_16[ii++] = htons(part);
 575       part = 0;
 576       if (input[pos] == ':') {
 577         if (colon < 8)
 578           return 0;
 579         colon = ii;
 580         pos++;
 581       }
 582       break;
 583     case '.': {
 584       uint32_t ip4;
 585       unsigned int len;
 586       len = ircd_aton_ip4(part_start, &ip4, pbits);
 587       if (!len || (ii > 6))
 588         return 0;
 589       ip->in6_16[ii++] = htons(ntohl(ip4) >> 16);
 590       ip->in6_16[ii++] = htons(ntohl(ip4) & 65535);
 591       if (pbits)
 592         *pbits += 96;
 593       pos = part_start + len - input;
 594       goto finish;
 595     }
 596     case '/':
 597       if (!pbits || !IsDigit(input[pos + 1]))
 598         return 0;
 599       ip->in6_16[ii++] = htons(part);
 600       for (part = 0; IsDigit(input[++pos]); )
 601         part = part * 10 + input[pos] - '0';
 602       if (part > 128)
 603         return 0;
 604       *pbits = part;
 605       goto finish;
 606     case '*':
 607       while (input[++pos] == '*' || input[pos] == ':') ;
 608       if (input[pos] != '\0' || colon < 8)
 609         return 0;
 610       if (pbits)
 611         *pbits = ii * 16;
 612       return pos;
 613     case '\0':
 614       ip->in6_16[ii++] = htons(part);
 615       if (colon == 8 && ii < 8)
 616         return 0;
 617       if (pbits)
 618         *pbits = 128;
 619       goto finish;
 620     default:
 621       return 0;
 622     }
 623   finish:
 624     if (colon < 8) {
 625       unsigned int jj;
 626       /* Shift stuff after "::" up and fill middle with zeros. */
 627       for (jj = 0; jj < ii - colon; jj++)
 628         ip->in6_16[7 - jj] = ip->in6_16[ii - jj - 1];
 629       for (jj = 0; jj < 8 - ii; jj++)
 630         ip->in6_16[colon + jj] = 0;
 631     }
 632     return pos;
 633   } else if (dot || strchr(input, '/')) {
 634     unsigned int addr;
 635     int len = ircd_aton_ip4(input, &addr, pbits);
 636     if (len) {
 637       ip->in6_16[5] = htons(65535);
 638       ip->in6_16[6] = htons(ntohl(addr) >> 16);
 639       ip->in6_16[7] = htons(ntohl(addr) & 65535);
 640       if (pbits)
 641         *pbits += 96;
 642     }
 643     return len;
 644   } else if (input[0] == '*') {
 645     unsigned int pos = 0;
 646     while (input[++pos] == '*') ;
 647     if (input[pos] != '\0')
 648       return 0;
 649     if (pbits)
 650       *pbits = 0;
 651     return pos;
 652   } else return 0; /* parse failed */
 653 }