ghostutils.c 19.7 KB
Newer Older
1 2 3 4 5 6 7 8
/* -*- mode: C; c-file-style: "gnu"; indent-tabs-mode: nil; -*- */

/* GLIB - Library of useful routines for C programming
 * Copyright (C) 2008 Red Hat, Inc.
 *
 * This library is free software; you can redistribute it and/or
 * modify it under the terms of the GNU Lesser General Public
 * License as published by the Free Software Foundation; either
9
 * version 2.1 of the License, or (at your option) any later version.
10 11 12 13 14 15 16
 *
 * This library is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
 * Lesser General Public License for more details.
 *
 * You should have received a copy of the GNU Lesser General
17
 * Public License along with this library; if not, see <http://www.gnu.org/licenses/>.
18 19 20 21 22 23
 */

#include "config.h"

#include <string.h>

Matthias Clasen's avatar
Matthias Clasen committed
24 25 26
#include "ghostutils.h"

#include "garray.h"
27
#include "gmem.h"
Matthias Clasen's avatar
Matthias Clasen committed
28 29 30 31
#include "gstring.h"
#include "gstrfuncs.h"
#include "glibintl.h"

32 33 34 35 36 37 38 39 40

/**
 * SECTION:ghostutils
 * @short_description: Internet hostname utilities
 *
 * Functions for manipulating internet hostnames; in particular, for
 * converting between Unicode and ASCII-encoded forms of
 * Internationalized Domain Names (IDNs).
 *
41 42 43
 * The
 * [Internationalized Domain Names for Applications (IDNA)](http://www.ietf.org/rfc/rfc3490.txt)
 * standards allow for the use
44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303
 * of Unicode domain names in applications, while providing
 * backward-compatibility with the old ASCII-only DNS, by defining an
 * ASCII-Compatible Encoding of any given Unicode name, which can be
 * used with non-IDN-aware applications and protocols. (For example,
 * "Παν語.org" maps to "xn--4wa8awb4637h.org".)
 **/

#define IDNA_ACE_PREFIX     "xn--"
#define IDNA_ACE_PREFIX_LEN 4

/* Punycode constants, from RFC 3492. */

#define PUNYCODE_BASE          36
#define PUNYCODE_TMIN           1
#define PUNYCODE_TMAX          26
#define PUNYCODE_SKEW          38
#define PUNYCODE_DAMP         700
#define PUNYCODE_INITIAL_BIAS  72
#define PUNYCODE_INITIAL_N   0x80

#define PUNYCODE_IS_BASIC(cp) ((guint)(cp) < 0x80)

/* Encode/decode a single base-36 digit */
static inline gchar
encode_digit (guint dig)
{
  if (dig < 26)
    return dig + 'a';
  else
    return dig - 26 + '0';
}

static inline guint
decode_digit (gchar dig)
{
  if (dig >= 'A' && dig <= 'Z')
    return dig - 'A';
  else if (dig >= 'a' && dig <= 'z')
    return dig - 'a';
  else if (dig >= '0' && dig <= '9')
    return dig - '0' + 26;
  else
    return G_MAXUINT;
}

/* Punycode bias adaptation algorithm, RFC 3492 section 6.1 */
static guint
adapt (guint    delta,
       guint    numpoints,
       gboolean firsttime)
{
  guint k;

  delta = firsttime ? delta / PUNYCODE_DAMP : delta / 2;
  delta += delta / numpoints;

  k = 0;
  while (delta > ((PUNYCODE_BASE - PUNYCODE_TMIN) * PUNYCODE_TMAX) / 2)
    {
      delta /= PUNYCODE_BASE - PUNYCODE_TMIN;
      k += PUNYCODE_BASE;
    }

  return k + ((PUNYCODE_BASE - PUNYCODE_TMIN + 1) * delta /
	      (delta + PUNYCODE_SKEW));
}

/* Punycode encoder, RFC 3492 section 6.3. The algorithm is
 * sufficiently bizarre that it's not really worth trying to explain
 * here.
 */
static gboolean
punycode_encode (const gchar *input_utf8,
                 gsize        input_utf8_length,
		 GString     *output)
{
  guint delta, handled_chars, num_basic_chars, bias, j, q, k, t, digit;
  gunichar n, m, *input;
  glong input_length;
  gboolean success = FALSE;

  /* Convert from UTF-8 to Unicode code points */
  input = g_utf8_to_ucs4 (input_utf8, input_utf8_length, NULL,
			  &input_length, NULL);
  if (!input)
    return FALSE;

  /* Copy basic chars */
  for (j = num_basic_chars = 0; j < input_length; j++)
    {
      if (PUNYCODE_IS_BASIC (input[j]))
	{
	  g_string_append_c (output, g_ascii_tolower (input[j]));
	  num_basic_chars++;
	}
    }
  if (num_basic_chars)
    g_string_append_c (output, '-');

  handled_chars = num_basic_chars;

  /* Encode non-basic chars */
  delta = 0;
  bias = PUNYCODE_INITIAL_BIAS;
  n = PUNYCODE_INITIAL_N;
  while (handled_chars < input_length)
    {
      /* let m = the minimum {non-basic} code point >= n in the input */
      for (m = G_MAXUINT, j = 0; j < input_length; j++)
	{
	  if (input[j] >= n && input[j] < m)
	    m = input[j];
	}

      if (m - n > (G_MAXUINT - delta) / (handled_chars + 1))
	goto fail;
      delta += (m - n) * (handled_chars + 1);
      n = m;

      for (j = 0; j < input_length; j++)
	{
	  if (input[j] < n)
	    {
	      if (++delta == 0)
		goto fail;
	    }
	  else if (input[j] == n)
	    {
	      q = delta;
	      for (k = PUNYCODE_BASE; ; k += PUNYCODE_BASE)
		{
		  if (k <= bias)
		    t = PUNYCODE_TMIN;
		  else if (k >= bias + PUNYCODE_TMAX)
		    t = PUNYCODE_TMAX;
		  else
		    t = k - bias;
		  if (q < t)
		    break;
		  digit = t + (q - t) % (PUNYCODE_BASE - t);
		  g_string_append_c (output, encode_digit (digit));
		  q = (q - t) / (PUNYCODE_BASE - t);
		}

	      g_string_append_c (output, encode_digit (q));
	      bias = adapt (delta, handled_chars + 1, handled_chars == num_basic_chars);
	      delta = 0;
	      handled_chars++;
	    }
	}

      delta++;
      n++;
    }

  success = TRUE;

 fail:
  g_free (input);
  return success;
}

/* From RFC 3454, Table B.1 */
#define idna_is_junk(ch) ((ch) == 0x00AD || (ch) == 0x1806 || (ch) == 0x200B || (ch) == 0x2060 || (ch) == 0xFEFF || (ch) == 0x034F || (ch) == 0x180B || (ch) == 0x180C || (ch) == 0x180D || (ch) == 0x200C || (ch) == 0x200D || ((ch) >= 0xFE00 && (ch) <= 0xFE0F))

/* Scan @str for "junk" and return a cleaned-up string if any junk
 * is found. Else return %NULL.
 */
static gchar *
remove_junk (const gchar *str,
             gint         len)
{
  GString *cleaned = NULL;
  const gchar *p;
  gunichar ch;

  for (p = str; len == -1 ? *p : p < str + len; p = g_utf8_next_char (p))
    {
      ch = g_utf8_get_char (p);
      if (idna_is_junk (ch))
	{
	  if (!cleaned)
	    {
	      cleaned = g_string_new (NULL);
	      g_string_append_len (cleaned, str, p - str);
	    }
	}
      else if (cleaned)
	g_string_append_unichar (cleaned, ch);
    }

  if (cleaned)
    return g_string_free (cleaned, FALSE);
  else
    return NULL;
}

static inline gboolean
contains_uppercase_letters (const gchar *str,
                            gint         len)
{
  const gchar *p;

  for (p = str; len == -1 ? *p : p < str + len; p = g_utf8_next_char (p))
    {
      if (g_unichar_isupper (g_utf8_get_char (p)))
	return TRUE;
    }
  return FALSE;
}

static inline gboolean
contains_non_ascii (const gchar *str,
                    gint         len)
{
  const gchar *p;

  for (p = str; len == -1 ? *p : p < str + len; p++)
    {
      if ((guchar)*p > 0x80)
	return TRUE;
    }
  return FALSE;
}

/* RFC 3454, Appendix C. ish. */
static inline gboolean
idna_is_prohibited (gunichar ch)
{
  switch (g_unichar_type (ch))
    {
    case G_UNICODE_CONTROL:
    case G_UNICODE_FORMAT:
    case G_UNICODE_UNASSIGNED:
    case G_UNICODE_PRIVATE_USE:
    case G_UNICODE_SURROGATE:
    case G_UNICODE_LINE_SEPARATOR:
    case G_UNICODE_PARAGRAPH_SEPARATOR:
    case G_UNICODE_SPACE_SEPARATOR:
      return TRUE;

    case G_UNICODE_OTHER_SYMBOL:
      if (ch == 0xFFFC || ch == 0xFFFD ||
	  (ch >= 0x2FF0 && ch <= 0x2FFB))
	return TRUE;
      return FALSE;

    case G_UNICODE_NON_SPACING_MARK:
      if (ch == 0x0340 || ch == 0x0341)
	return TRUE;
      return FALSE;

    default:
      return FALSE;
    }
}

/* RFC 3491 IDN cleanup algorithm. */
static gchar *
nameprep (const gchar *hostname,
304 305
          gint         len,
          gboolean    *is_unicode)
306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337
{
  gchar *name, *tmp = NULL, *p;

  /* It would be nice if we could do this without repeatedly
   * allocating strings and converting back and forth between
   * gunichars and UTF-8... The code does at least avoid doing most of
   * the sub-operations when they would just be equivalent to a
   * g_strdup().
   */

  /* Remove presentation-only characters */
  name = remove_junk (hostname, len);
  if (name)
    {
      tmp = name;
      len = -1;
    }
  else
    name = (gchar *)hostname;

  /* Convert to lowercase */
  if (contains_uppercase_letters (name, len))
    {
      name = g_utf8_strdown (name, len);
      g_free (tmp);
      tmp = name;
      len = -1;
    }

  /* If there are no UTF8 characters, we're done. */
  if (!contains_non_ascii (name, len))
    {
338
      *is_unicode = FALSE;
339 340 341 342 343 344
      if (name == (gchar *)hostname)
        return len == -1 ? g_strdup (hostname) : g_strndup (hostname, len);
      else
        return name;
    }

345 346
  *is_unicode = TRUE;

347 348 349 350 351
  /* Normalize */
  name = g_utf8_normalize (name, len, G_NORMALIZE_NFKC);
  g_free (tmp);
  tmp = name;

352 353 354
  if (!name)
    return NULL;

355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387
  /* KC normalization may have created more capital letters (eg,
   * angstrom -> capital A with ring). So we have to lowercasify a
   * second time. (This is more-or-less how the nameprep algorithm
   * does it. If tolower(nfkc(tolower(X))) is guaranteed to be the
   * same as tolower(nfkc(X)), then we could skip the first tolower,
   * but I'm not sure it is.)
   */
  if (contains_uppercase_letters (name, -1))
    {
      name = g_utf8_strdown (name, -1);
      g_free (tmp);
      tmp = name;
    }

  /* Check for prohibited characters */
  for (p = name; *p; p = g_utf8_next_char (p))
    {
      if (idna_is_prohibited (g_utf8_get_char (p)))
	{
	  name = NULL;
          g_free (tmp);
	  goto done;
	}
    }

  /* FIXME: We're supposed to verify certain constraints on bidi
   * characters, but glib does not appear to have that information.
   */

 done:
  return name;
}

388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407
/* RFC 3490, section 3.1 says '.', 0x3002, 0xFF0E, and 0xFF61 count as
 * label-separating dots. @str must be '\0'-terminated.
 */
#define idna_is_dot(str) ( \
  ((guchar)(str)[0] == '.') ||                                                 \
  ((guchar)(str)[0] == 0xE3 && (guchar)(str)[1] == 0x80 && (guchar)(str)[2] == 0x82) || \
  ((guchar)(str)[0] == 0xEF && (guchar)(str)[1] == 0xBC && (guchar)(str)[2] == 0x8E) || \
  ((guchar)(str)[0] == 0xEF && (guchar)(str)[1] == 0xBD && (guchar)(str)[2] == 0xA1) )

static const gchar *
idna_end_of_label (const gchar *str)
{
  for (; *str; str = g_utf8_next_char (str))
    {
      if (idna_is_dot (str))
        return str;
    }
  return str;
}

408 409 410 411 412 413 414 415
/**
 * g_hostname_to_ascii:
 * @hostname: a valid UTF-8 or ASCII hostname
 *
 * Converts @hostname to its canonical ASCII form; an ASCII-only
 * string containing no uppercase letters and not ending with a
 * trailing dot.
 *
416
 * Returns: an ASCII hostname, which must be freed, or %NULL if
417 418 419 420 421 422 423 424 425 426 427 428
 * @hostname is in some way invalid.
 *
 * Since: 2.22
 **/
gchar *
g_hostname_to_ascii (const gchar *hostname)
{
  gchar *name, *label, *p;
  GString *out;
  gssize llen, oldlen;
  gboolean unicode;

429 430 431
  label = name = nameprep (hostname, -1, &unicode);
  if (!name || !unicode)
    return name;
432 433

  out = g_string_new (NULL);
434 435 436 437

  do
    {
      unicode = FALSE;
438
      for (p = label; *p && !idna_is_dot (p); p++)
439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461
	{
	  if ((guchar)*p > 0x80)
	    unicode = TRUE;
	}

      oldlen = out->len;
      llen = p - label;
      if (unicode)
	{
          if (!strncmp (label, IDNA_ACE_PREFIX, IDNA_ACE_PREFIX_LEN))
            goto fail;

	  g_string_append (out, IDNA_ACE_PREFIX);
	  if (!punycode_encode (label, llen, out))
	    goto fail;
	}
      else
        g_string_append_len (out, label, llen);

      if (out->len - oldlen > 63)
	goto fail;

      label += llen;
462 463 464
      if (*label)
        label = g_utf8_next_char (label);
      if (*label)
465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489
        g_string_append_c (out, '.');
    }
  while (*label);

  g_free (name);
  return g_string_free (out, FALSE);

 fail:
  g_free (name);
  g_string_free (out, TRUE);
  return NULL;
}

/**
 * g_hostname_is_non_ascii:
 * @hostname: a hostname
 *
 * Tests if @hostname contains Unicode characters. If this returns
 * %TRUE, you need to encode the hostname with g_hostname_to_ascii()
 * before using it in non-IDN-aware contexts.
 *
 * Note that a hostname might contain a mix of encoded and unencoded
 * segments, and so it is possible for g_hostname_is_non_ascii() and
 * g_hostname_is_ascii_encoded() to both return %TRUE for a name.
 *
490
 * Returns: %TRUE if @hostname contains any non-ASCII characters
491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596
 *
 * Since: 2.22
 **/
gboolean
g_hostname_is_non_ascii (const gchar *hostname)
{
  return contains_non_ascii (hostname, -1);
}

/* Punycode decoder, RFC 3492 section 6.2. As with punycode_encode(),
 * read the RFC if you want to understand what this is actually doing.
 */
static gboolean
punycode_decode (const gchar *input,
                 gsize        input_length,
                 GString     *output)
{
  GArray *output_chars;
  gunichar n;
  guint i, bias;
  guint oldi, w, k, digit, t;
  const gchar *split;

  n = PUNYCODE_INITIAL_N;
  i = 0;
  bias = PUNYCODE_INITIAL_BIAS;

  split = input + input_length - 1;
  while (split > input && *split != '-')
    split--;
  if (split > input)
    {
      output_chars = g_array_sized_new (FALSE, FALSE, sizeof (gunichar),
					split - input);
      input_length -= (split - input) + 1;
      while (input < split)
	{
	  gunichar ch = (gunichar)*input++;
	  if (!PUNYCODE_IS_BASIC (ch))
	    goto fail;
	  g_array_append_val (output_chars, ch);
	}
      input++;
    }
  else
    output_chars = g_array_new (FALSE, FALSE, sizeof (gunichar));

  while (input_length)
    {
      oldi = i;
      w = 1;
      for (k = PUNYCODE_BASE; ; k += PUNYCODE_BASE)
	{
	  if (!input_length--)
	    goto fail;
	  digit = decode_digit (*input++);
	  if (digit >= PUNYCODE_BASE)
	    goto fail;
	  if (digit > (G_MAXUINT - i) / w)
	    goto fail;
	  i += digit * w;
	  if (k <= bias)
	    t = PUNYCODE_TMIN;
	  else if (k >= bias + PUNYCODE_TMAX)
	    t = PUNYCODE_TMAX;
	  else
	    t = k - bias;
	  if (digit < t)
	    break;
	  if (w > G_MAXUINT / (PUNYCODE_BASE - t))
	    goto fail;
	  w *= (PUNYCODE_BASE - t);
	}

      bias = adapt (i - oldi, output_chars->len + 1, oldi == 0);

      if (i / (output_chars->len + 1) > G_MAXUINT - n)
	goto fail;
      n += i / (output_chars->len + 1);
      i %= (output_chars->len + 1);

      g_array_insert_val (output_chars, i++, n);
    }

  for (i = 0; i < output_chars->len; i++)
    g_string_append_unichar (output, g_array_index (output_chars, gunichar, i));
  g_array_free (output_chars, TRUE);
  return TRUE;

 fail:
  g_array_free (output_chars, TRUE);
  return FALSE;
}

/**
 * g_hostname_to_unicode:
 * @hostname: a valid UTF-8 or ASCII hostname
 *
 * Converts @hostname to its canonical presentation form; a UTF-8
 * string in Unicode normalization form C, containing no uppercase
 * letters, no forbidden characters, and no ASCII-encoded segments,
 * and not ending with a trailing dot.
 *
 * Of course if @hostname is not an internationalized hostname, then
 * the canonical presentation form will be entirely ASCII.
 *
597
 * Returns: a UTF-8 hostname, which must be freed, or %NULL if
598 599 600 601 602 603 604 605 606 607 608 609 610 611
 * @hostname is in some way invalid.
 *
 * Since: 2.22
 **/
gchar *
g_hostname_to_unicode (const gchar *hostname)
{
  GString *out;
  gssize llen;

  out = g_string_new (NULL);

  do
    {
612
      llen = idna_end_of_label (hostname) - hostname;
613 614 615 616 617 618 619 620 621 622 623 624
      if (!g_ascii_strncasecmp (hostname, IDNA_ACE_PREFIX, IDNA_ACE_PREFIX_LEN))
	{
	  hostname += IDNA_ACE_PREFIX_LEN;
	  llen -= IDNA_ACE_PREFIX_LEN;
	  if (!punycode_decode (hostname, llen, out))
	    {
	      g_string_free (out, TRUE);
	      return NULL;
	    }
	}
      else
        {
625 626
          gboolean unicode;
          gchar *canonicalized = nameprep (hostname, llen, &unicode);
627

628 629 630 631 632
          if (!canonicalized)
            {
              g_string_free (out, TRUE);
              return NULL;
            }
633 634 635 636 637
          g_string_append (out, canonicalized);
          g_free (canonicalized);
        }

      hostname += llen;
638 639 640
      if (*hostname)
        hostname = g_utf8_next_char (hostname);
      if (*hostname)
641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660
        g_string_append_c (out, '.');
    }
  while (*hostname);

  return g_string_free (out, FALSE);
}

/**
 * g_hostname_is_ascii_encoded:
 * @hostname: a hostname
 *
 * Tests if @hostname contains segments with an ASCII-compatible
 * encoding of an Internationalized Domain Name. If this returns
 * %TRUE, you should decode the hostname with g_hostname_to_unicode()
 * before displaying it to the user.
 *
 * Note that a hostname might contain a mix of encoded and unencoded
 * segments, and so it is possible for g_hostname_is_non_ascii() and
 * g_hostname_is_ascii_encoded() to both return %TRUE for a name.
 *
661
 * Returns: %TRUE if @hostname contains any ASCII-encoded
662 663 664 665 666 667 668 669 670 671 672
 * segments.
 *
 * Since: 2.22
 **/
gboolean
g_hostname_is_ascii_encoded (const gchar *hostname)
{
  while (1)
    {
      if (!g_ascii_strncasecmp (hostname, IDNA_ACE_PREFIX, IDNA_ACE_PREFIX_LEN))
	return TRUE;
673 674 675 676
      hostname = idna_end_of_label (hostname);
      if (*hostname)
        hostname = g_utf8_next_char (hostname);
      if (!*hostname)
677 678 679 680 681 682 683 684 685 686 687
	return FALSE;
    }
}

/**
 * g_hostname_is_ip_address:
 * @hostname: a hostname (or IP address in string form)
 *
 * Tests if @hostname is the string form of an IPv4 or IPv6 address.
 * (Eg, "192.168.0.1".)
 *
688
 * Returns: %TRUE if @hostname is an IP address
689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787
 *
 * Since: 2.22
 **/
gboolean
g_hostname_is_ip_address (const gchar *hostname)
{
  gchar *p, *end;
  gint nsegments, octet;

  /* On Linux we could implement this using inet_pton, but the Windows
   * equivalent of that requires linking against winsock, so we just
   * figure this out ourselves. Tested by tests/hostutils.c.
   */

  p = (char *)hostname;

  if (strchr (p, ':'))
    {
      gboolean skipped;

      /* If it contains a ':', it's an IPv6 address (assuming it's an
       * IP address at all). This consists of eight ':'-separated
       * segments, each containing a 1-4 digit hex number, except that
       * optionally: (a) the last two segments can be replaced by an
       * IPv4 address, and (b) a single span of 1 to 8 "0000" segments
       * can be replaced with just "::".
       */

      nsegments = 0;
      skipped = FALSE;
      while (*p && nsegments < 8)
        {
          /* Each segment after the first must be preceded by a ':'.
           * (We also handle half of the "string starts with ::" case
           * here.)
           */
          if (p != (char *)hostname || (p[0] == ':' && p[1] == ':'))
            {
              if (*p != ':')
                return FALSE;
              p++;
            }

          /* If there's another ':', it means we're skipping some segments */
          if (*p == ':' && !skipped)
            {
              skipped = TRUE;
              nsegments++;

              /* Handle the "string ends with ::" case */
              if (!p[1])
                p++;

              continue;
            }

          /* Read the segment, make sure it's valid. */
          for (end = p; g_ascii_isxdigit (*end); end++)
            ;
          if (end == p || end > p + 4)
            return FALSE;

          if (*end == '.')
            {
              if ((nsegments == 6 && !skipped) || (nsegments <= 6 && skipped))
                goto parse_ipv4;
              else
                return FALSE;
            }

          nsegments++;
          p = end;
        }

      return !*p && (nsegments == 8 || skipped);
    }

 parse_ipv4:

  /* Parse IPv4: N.N.N.N, where each N <= 255 and doesn't have leading 0s. */
  for (nsegments = 0; nsegments < 4; nsegments++)
    {
      if (nsegments != 0)
        {
          if (*p != '.')
            return FALSE;
          p++;
        }

      /* Check the segment; a little tricker than the IPv6 case since
       * we can't allow extra leading 0s, and we can't assume that all
       * strings of valid length are within range.
       */
      octet = 0;
      if (*p == '0')
        end = p + 1;
      else
        {
          for (end = p; g_ascii_isdigit (*end); end++)
788 789 790 791 792 793
            {
              octet = 10 * octet + (*end - '0');

              if (octet > 255)
                break;
            }
794 795 796 797 798 799 800 801 802 803
        }
      if (end == p || end > p + 3 || octet > 255)
        return FALSE;

      p = end;
    }

  /* If there's nothing left to parse, then it's ok. */
  return !*p;
}