| File: | url/./../bnf/sofia-sip/bnf.h |
| Warning: | line 266, column 10 Array subscript is undefined |
Press '?' to see keyboard shortcuts
Keyboard shortcuts:
| 1 | /* | |||
| 2 | * This file is part of the Sofia-SIP package | |||
| 3 | * | |||
| 4 | * Copyright (C) 2005 Nokia Corporation. | |||
| 5 | * | |||
| 6 | * Contact: Pekka Pessi <pekka.pessi@nokia.com> | |||
| 7 | * | |||
| 8 | * This library is free software; you can redistribute it and/or | |||
| 9 | * modify it under the terms of the GNU Lesser General Public License | |||
| 10 | * as published by the Free Software Foundation; either version 2.1 of | |||
| 11 | * the License, or (at your option) any later version. | |||
| 12 | * | |||
| 13 | * This library is distributed in the hope that it will be useful, but | |||
| 14 | * WITHOUT ANY WARRANTY; without even the implied warranty of | |||
| 15 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |||
| 16 | * Lesser General Public License for more details. | |||
| 17 | * | |||
| 18 | * You should have received a copy of the GNU Lesser General Public | |||
| 19 | * License along with this library; if not, write to the Free Software | |||
| 20 | * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA | |||
| 21 | * 02110-1301 USA | |||
| 22 | * | |||
| 23 | */ | |||
| 24 | ||||
| 25 | /**@CFILE url.c | |||
| 26 | * | |||
| 27 | * Implementation of basic URL parsing and handling. | |||
| 28 | * | |||
| 29 | * @author Pekka Pessi <Pekka.Pessi@nokia.com> | |||
| 30 | * | |||
| 31 | * @date Created: Thu Jun 29 22:44:37 2000 ppessi | |||
| 32 | */ | |||
| 33 | ||||
| 34 | #include "config.h" | |||
| 35 | ||||
| 36 | #include <sofia-sip/su_alloc.h> | |||
| 37 | #include <sofia-sip/bnf.h> | |||
| 38 | #include <sofia-sip/hostdomain.h> | |||
| 39 | #include <sofia-sip/url.h> | |||
| 40 | ||||
| 41 | #include <sofia-sip/string0.h> | |||
| 42 | ||||
| 43 | #include <stdio.h> | |||
| 44 | #include <string.h> | |||
| 45 | #include <stdlib.h> | |||
| 46 | #include <assert.h> | |||
| 47 | #include <ctype.h> | |||
| 48 | #include <limits.h> | |||
| 49 | ||||
| 50 | /**@def URL_PRINT_FORMAT | |||
| 51 | * Format string used when printing url with printf(). | |||
| 52 | * | |||
| 53 | * The macro URL_PRINT_FORMAT is used in format string of printf() or | |||
| 54 | * similar printing functions. A URL can be printed like this: | |||
| 55 | * @code | |||
| 56 | * printf("%s received URL " URL_PRINT_FORMAT "\n", | |||
| 57 | * my_name, URL_PRINT_ARGS(url)); | |||
| 58 | * @endcode | |||
| 59 | */ | |||
| 60 | ||||
| 61 | /** @def URL_PRINT_ARGS(u) | |||
| 62 | * Argument list used when printing url with printf(). | |||
| 63 | * | |||
| 64 | * The macro URL_PRINT_ARGS() is used to create a stdarg list for printf() | |||
| 65 | * or similar printing functions. Using it, a URL can be printed like this: | |||
| 66 | * | |||
| 67 | * @code | |||
| 68 | * printf("%s received URL " URL_PRINT_FORMAT "\n", | |||
| 69 | * my_name, URL_PRINT_ARGS(url)); | |||
| 70 | * @endcode | |||
| 71 | */ | |||
| 72 | ||||
| 73 | #define RESERVED";/?:@&=+$," ";/?:@&=+$," | |||
| 74 | #define DELIMS"<>#%\"" "<>#%\"" | |||
| 75 | #define UNWISE"{}|\\^[]`" "{}|\\^[]`" | |||
| 76 | ||||
| 77 | #define EXCLUDED";/?:@&=+$," "<>#%\"" "{}|\\^[]`" RESERVED";/?:@&=+$," DELIMS"<>#%\"" UNWISE"{}|\\^[]`" | |||
| 78 | ||||
| 79 | #define UNRESERVED"ABCDEFGHIJKLMNOPQRSTUVWXYZ" "abcdefghijklmnopqrstuvwxyz" "0123456789" "-_.!~*'()" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" \ | |||
| 80 | "abcdefghijklmnopqrstuvwxyz" \ | |||
| 81 | "0123456789" \ | |||
| 82 | "-_.!~*'()" | |||
| 83 | ||||
| 84 | #define IS_EXCLUDED(u, m32, m64, m96)(u <= ' ' || u >= '\177' || (u < 64 ? (m32 & (1 << (63 - u))) : (u < 96 ? (m64 & (1 << (95 - u))) : (m96 & (1 << (127 - u))))) != 0) \ | |||
| 85 | (u <= ' ' \ | |||
| 86 | || u >= '\177' \ | |||
| 87 | || (u < 64 ? (m32 & (1 << (63 - u))) \ | |||
| 88 | : (u < 96 ? (m64 & (1 << (95 - u))) \ | |||
| 89 | : /*u < 128*/ (m96 & (1 << (127 - u))))) != 0) | |||
| 90 | ||||
| 91 | #define MASKS_WITH_RESERVED(reserved, m32, m64, m96)if (reserved == ((void*)0)) { m32 = 0xbe19003f, m64 = 0x8000001e , m96 = 0x8000001d; } else do { m32 = 0xb400000a, m64 = 0x0000001e , m96 = 0x8000001d; for (;reserved[0]; reserved++) { unsigned r = reserved[0]; if (r < 32) ; else if (r < 64) m32 |= 1U << (63 - r); else if (r < 96) m64 |= 1U << (95 - r); else if (r < 128) m96 |= 1U << (127 - r); } } while (0) \ | |||
| 92 | if (reserved == NULL((void*)0)) { \ | |||
| 93 | m32 = 0xbe19003f, m64 = 0x8000001e, m96 = 0x8000001d; \ | |||
| 94 | } else do { \ | |||
| 95 | m32 = 0xb400000a, m64 = 0x0000001e, m96 = 0x8000001d; \ | |||
| 96 | \ | |||
| 97 | for (;reserved[0]; reserved++) { \ | |||
| 98 | unsigned r = reserved[0]; \ | |||
| 99 | RESERVE(r, m32, m64, m96)if (r < 32) ; else if (r < 64) m32 |= 1U << (63 - r); else if (r < 96) m64 |= 1U << (95 - r); else if (r < 128) m96 |= 1U << (127 - r); \ | |||
| 100 | } \ | |||
| 101 | } while (0) | |||
| 102 | ||||
| 103 | #define RESERVE(reserved, m32, m64, m96)if (r < 32) ; else if (r < 64) m32 |= 1U << (63 - r); else if (r < 96) m64 |= 1U << (95 - r); else if (r < 128) m96 |= 1U << (127 - r) \ | |||
| 104 | if (r < 32) \ | |||
| 105 | ; \ | |||
| 106 | else if (r < 64) \ | |||
| 107 | m32 |= 1U << (63 - r); \ | |||
| 108 | else if (r < 96) \ | |||
| 109 | m64 |= 1U << (95 - r); \ | |||
| 110 | else if (r < 128) \ | |||
| 111 | m96 |= 1U << (127 - r) | |||
| 112 | ||||
| 113 | #define MASKS_WITH_ALLOWED(allowed, mask32, mask64, mask96)do { if (allowed) { for (;allowed[0]; allowed++) { unsigned a = allowed[0]; if (a < 32) ; else if (a < 64) mask32 &= ~(1U << (63 - a)); else if (a < 96) mask64 &= ~ (1U << (95 - a)); else if (a < 128) mask96 &= ~( 1U << (127 - a)); } } } while (0) \ | |||
| 114 | do { \ | |||
| 115 | if (allowed) { \ | |||
| 116 | for (;allowed[0]; allowed++) { \ | |||
| 117 | unsigned a = allowed[0]; \ | |||
| 118 | ALLOW(a, mask32, mask64, mask96)if (a < 32) ; else if (a < 64) mask32 &= ~(1U << (63 - a)); else if (a < 96) mask64 &= ~(1U << ( 95 - a)); else if (a < 128) mask96 &= ~(1U << (127 - a)); \ | |||
| 119 | } \ | |||
| 120 | } \ | |||
| 121 | } while (0) | |||
| 122 | ||||
| 123 | #define ALLOW(a, mask32, mask64, mask96)if (a < 32) ; else if (a < 64) mask32 &= ~(1U << (63 - a)); else if (a < 96) mask64 &= ~(1U << ( 95 - a)); else if (a < 128) mask96 &= ~(1U << (127 - a)) \ | |||
| 124 | if (a < 32) \ | |||
| 125 | ; \ | |||
| 126 | else if (a < 64) \ | |||
| 127 | mask32 &= ~(1U << (63 - a)); \ | |||
| 128 | else if (a < 96) \ | |||
| 129 | mask64 &= ~(1U << (95 - a)); \ | |||
| 130 | else if (a < 128) \ | |||
| 131 | mask96 &= ~(1U << (127 - a)) | |||
| 132 | ||||
| 133 | #define NUL'\0' '\0' | |||
| 134 | #define NULNULNUL'\0', '\0', '\0' '\0', '\0', '\0' | |||
| 135 | ||||
| 136 | #define RMASK10xbe19003f 0xbe19003f | |||
| 137 | #define RMASK20x8000001e 0x8000001e | |||
| 138 | #define RMASK30x8000001d 0x8000001d | |||
| 139 | ||||
| 140 | #define RESERVED_MASK0xbe19003f, 0x8000001e, 0x8000001d 0xbe19003f, 0x8000001e, 0x8000001d | |||
| 141 | #define URIC_MASK0xb400000a, 0x0000001e, 0x8000001d 0xb400000a, 0x0000001e, 0x8000001d | |||
| 142 | ||||
| 143 | #define IS_EXCLUDED_MASK(u, m)IS_EXCLUDED IS_EXCLUDED(u, m) | |||
| 144 | ||||
| 145 | /* Internal prototypes */ | |||
| 146 | static char *url_canonize(char *d, char const *s, size_t n, | |||
| 147 | unsigned syn33, | |||
| 148 | char const allowed[]); | |||
| 149 | static char *url_canonize2(char *d, char const *s, size_t n, | |||
| 150 | unsigned syn33, | |||
| 151 | unsigned m32, unsigned m64, unsigned m96); | |||
| 152 | static int url_tel_cmp_numbers(char const *A, char const *B); | |||
| 153 | ||||
| 154 | /**Test if string contains excluded or url-reserved characters. | |||
| 155 | * | |||
| 156 | * | |||
| 157 | * | |||
| 158 | * @param s string to be searched | |||
| 159 | * | |||
| 160 | * @retval 0 if no reserved characters were found. | |||
| 161 | * @retval l if a reserved character was found. | |||
| 162 | */ | |||
| 163 | int url_reserved_p(char const *s) | |||
| 164 | { | |||
| 165 | if (s) | |||
| 166 | while (*s) { | |||
| 167 | unsigned char u = *s++; | |||
| 168 | ||||
| 169 | if (IS_EXCLUDED(u, RMASK1, RMASK2, RMASK3)(u <= ' ' || u >= '\177' || (u < 64 ? (0xbe19003f & (1 << (63 - u))) : (u < 96 ? (0x8000001e & (1 << (95 - u))) : (0x8000001d & (1 << (127 - u))))) != 0 )) | |||
| 170 | return 1; | |||
| 171 | } | |||
| 172 | ||||
| 173 | return 0; | |||
| 174 | } | |||
| 175 | ||||
| 176 | /** Calculate length of string when escaped with %-notation. | |||
| 177 | * | |||
| 178 | * Calculate the length of string @a s when the excluded or reserved | |||
| 179 | * characters in it have been escaped. | |||
| 180 | * | |||
| 181 | * @param s String with reserved URL characters. [IN | |||
| 182 | * @param reserved Optional array of reserved characters [IN] | |||
| 183 | * | |||
| 184 | * @return | |||
| 185 | * The number of characters in corresponding but escaped string. | |||
| 186 | * | |||
| 187 | * You can handle a part of URL with reserved characters like this: | |||
| 188 | * @code | |||
| 189 | * if (url_reserved_p(s)) { | |||
| 190 | * n = malloc(url_esclen(s, NULL) + 1); | |||
| 191 | * if (n) url_escape(n, s); | |||
| 192 | * } else { | |||
| 193 | * n = malloc(strlen(s) + 1); | |||
| 194 | * if (n) strcpy(n, s); | |||
| 195 | * } | |||
| 196 | * @endcode | |||
| 197 | */ | |||
| 198 | isize_t url_esclen(char const *s, char const reserved[]) | |||
| 199 | { | |||
| 200 | size_t n; | |||
| 201 | unsigned mask32, mask64, mask96; | |||
| 202 | ||||
| 203 | MASKS_WITH_RESERVED(reserved, mask32, mask64, mask96)if (reserved == ((void*)0)) { mask32 = 0xbe19003f, mask64 = 0x8000001e , mask96 = 0x8000001d; } else do { mask32 = 0xb400000a, mask64 = 0x0000001e, mask96 = 0x8000001d; for (;reserved[0]; reserved ++) { unsigned r = reserved[0]; if (r < 32) ; else if (r < 64) mask32 |= 1U << (63 - r); else if (r < 96) mask64 |= 1U << (95 - r); else if (r < 128) mask96 |= 1U << (127 - r); } } while (0); | |||
| 204 | ||||
| 205 | for (n = 0; s && *s; n++) { | |||
| 206 | unsigned char u = *s++; | |||
| 207 | ||||
| 208 | if (IS_EXCLUDED(u, mask32, mask64, mask96)(u <= ' ' || u >= '\177' || (u < 64 ? (mask32 & ( 1 << (63 - u))) : (u < 96 ? (mask64 & (1 << (95 - u))) : (mask96 & (1 << (127 - u))))) != 0)) | |||
| 209 | n += 2; | |||
| 210 | } | |||
| 211 | ||||
| 212 | return (isize_t)n; | |||
| 213 | } | |||
| 214 | ||||
| 215 | /** Escape a string. | |||
| 216 | * | |||
| 217 | * The function url_escape() copies the string pointed by @a s to the array | |||
| 218 | * pointed by @a d, @b excluding the terminating \\0 character. All reserved | |||
| 219 | * characters in @a s are copied in hexadecimal format, for instance, @c | |||
| 220 | * "$%#" is copied as @c "%24%25%23". The destination array @a d must be | |||
| 221 | * large enough to receive the escaped copy. | |||
| 222 | * | |||
| 223 | * @param d Destination buffer [OUT] | |||
| 224 | * @param s String to be copied [IN] | |||
| 225 | * @param reserved Array of reserved characters [IN] | |||
| 226 | * | |||
| 227 | * @return Pointer to the destination array. | |||
| 228 | */ | |||
| 229 | char *url_escape(char *d, char const *s, char const reserved[]) | |||
| 230 | { | |||
| 231 | char *retval = d; | |||
| 232 | unsigned mask32, mask64, mask96; | |||
| 233 | ||||
| 234 | MASKS_WITH_RESERVED(reserved, mask32, mask64, mask96)if (reserved == ((void*)0)) { mask32 = 0xbe19003f, mask64 = 0x8000001e , mask96 = 0x8000001d; } else do { mask32 = 0xb400000a, mask64 = 0x0000001e, mask96 = 0x8000001d; for (;reserved[0]; reserved ++) { unsigned r = reserved[0]; if (r < 32) ; else if (r < 64) mask32 |= 1U << (63 - r); else if (r < 96) mask64 |= 1U << (95 - r); else if (r < 128) mask96 |= 1U << (127 - r); } } while (0); | |||
| 235 | ||||
| 236 | while (s && *s) { | |||
| 237 | unsigned char u = *s++; | |||
| 238 | ||||
| 239 | if (IS_EXCLUDED(u, mask32, mask64, mask96)(u <= ' ' || u >= '\177' || (u < 64 ? (mask32 & ( 1 << (63 - u))) : (u < 96 ? (mask64 & (1 << (95 - u))) : (mask96 & (1 << (127 - u))))) != 0)) { | |||
| 240 | # define URL_HEXIFY(u) ((u) + '0' + ((u) >= 10 ? 'A' - '0' - 10 : 0)) | |||
| 241 | ||||
| 242 | *d++ = '%'; | |||
| 243 | *d++ = URL_HEXIFY(u >> 4); | |||
| 244 | *d++ = URL_HEXIFY(u & 15); | |||
| 245 | ||||
| 246 | # undef URL_HEXIFY | |||
| 247 | } | |||
| 248 | else { | |||
| 249 | *d++ = u; | |||
| 250 | } | |||
| 251 | } | |||
| 252 | ||||
| 253 | *d = '\0'; | |||
| 254 | ||||
| 255 | return retval; | |||
| 256 | } | |||
| 257 | ||||
| 258 | ||||
| 259 | /**Unescape url-escaped string fragment. | |||
| 260 | * | |||
| 261 | * Unescape @a n characters from string @a s to the buffer @a d, including | |||
| 262 | * the terminating \\0 character. All %-escaped triplets in @a s are | |||
| 263 | * unescaped, for instance, @c "%40%25%23" is copied as @c "@%#". The | |||
| 264 | * destination array @a d must be large enough to receive the escaped copy | |||
| 265 | * (@a n bytes is always enough). | |||
| 266 | * | |||
| 267 | * @param d destination buffer | |||
| 268 | * @param s string to be unescaped | |||
| 269 | * @param n maximum number of characters to unescape | |||
| 270 | * | |||
| 271 | * @return Length of unescaped string | |||
| 272 | * | |||
| 273 | * @NEW_1_12_4. | |||
| 274 | */ | |||
| 275 | size_t url_unescape_to(char *d, char const *s, size_t n) | |||
| 276 | { | |||
| 277 | size_t i = 0, j = 0; | |||
| 278 | ||||
| 279 | if (s == NULL((void*)0)) | |||
| 280 | return 0; | |||
| 281 | ||||
| 282 | i = j = strncspn(s, n, "%"); | |||
| 283 | ||||
| 284 | if (d && d != s) | |||
| 285 | memmove(d, s, i); | |||
| 286 | ||||
| 287 | for (; i < n;) { | |||
| 288 | char c = s[i++]; | |||
| 289 | ||||
| 290 | if (c == '\0') | |||
| 291 | break; | |||
| 292 | ||||
| 293 | if (c == '%' && i + 1 < n && IS_HEX(s[i])(((s[i]) >= '0' && (s[i]) <= '9') || ((s[i]) >= 'A' && (s[i]) <= 'F') || ((s[i]) >= 'a' && (s[i]) <= 'f')) && IS_HEX(s[i + 1])(((s[i + 1]) >= '0' && (s[i + 1]) <= '9') || (( s[i + 1]) >= 'A' && (s[i + 1]) <= 'F') || ((s[i + 1]) >= 'a' && (s[i + 1]) <= 'f'))) { | |||
| 294 | #define UNHEX(a) (a - (a >= 'a' ? 'a' - 10 : (a >= 'A' ? 'A' - 10 : '0'))) | |||
| 295 | c = (UNHEX(s[i]) << 4) | UNHEX(s[i + 1]); | |||
| 296 | #undef UNHEX | |||
| 297 | i += 2; | |||
| 298 | } | |||
| 299 | ||||
| 300 | if (d) | |||
| 301 | d[j] = c; | |||
| 302 | j++; | |||
| 303 | } | |||
| 304 | ||||
| 305 | return j; | |||
| 306 | } | |||
| 307 | ||||
| 308 | /**Unescape url-escaped string. | |||
| 309 | * | |||
| 310 | * Unescape string @a s to the buffer @a d, including the terminating \\0 | |||
| 311 | * character. All %-escaped triplets in @a s are unescaped, for instance, @c | |||
| 312 | * "%40%25%23" is copied as @c "@%#". The destination array @a d must be | |||
| 313 | * large enough to receive the escaped copy. | |||
| 314 | * | |||
| 315 | * @param d destination buffer | |||
| 316 | * @param s string to be copied | |||
| 317 | * | |||
| 318 | * @return Pointer to the destination buffer. | |||
| 319 | */ | |||
| 320 | char *url_unescape(char *d, char const *s) | |||
| 321 | { | |||
| 322 | size_t n = url_unescape_to(d, s, SIZE_MAX(18446744073709551615UL)); | |||
| 323 | if (d) | |||
| 324 | d[n] = '\0'; | |||
| 325 | return d; | |||
| 326 | } | |||
| 327 | ||||
| 328 | /** Canonize a URL component */ | |||
| 329 | static | |||
| 330 | char *url_canonize(char *d, char const *s, size_t n, | |||
| 331 | unsigned syn33, | |||
| 332 | char const allowed[]) | |||
| 333 | { | |||
| 334 | unsigned mask32 = 0xbe19003f, mask64 = 0x8000001e, mask96 = 0x8000001d; | |||
| 335 | ||||
| 336 | MASKS_WITH_ALLOWED(allowed, mask32, mask64, mask96)do { if (allowed) { for (;allowed[0]; allowed++) { unsigned a = allowed[0]; if (a < 32) ; else if (a < 64) mask32 &= ~(1U << (63 - a)); else if (a < 96) mask64 &= ~ (1U << (95 - a)); else if (a < 128) mask96 &= ~( 1U << (127 - a)); } } } while (0); | |||
| 337 | ||||
| 338 | return url_canonize2(d, s, n, syn33, mask32, mask64, mask96); | |||
| 339 | } | |||
| 340 | ||||
| 341 | #define SYN33(c)(1U << (c - 33)) (1U << (c - 33)) | |||
| 342 | #define IS_SYN33(syn33, c)((syn33 & (1U << (c - 33))) != 0) ((syn33 & (1U << (c - 33))) != 0) | |||
| 343 | ||||
| 344 | /** Canonize a URL component (with precomputed mask) */ | |||
| 345 | static | |||
| 346 | char *url_canonize2(char *d, char const * const s, size_t n, | |||
| 347 | unsigned syn33, | |||
| 348 | unsigned m32, unsigned m64, unsigned m96) | |||
| 349 | { | |||
| 350 | size_t i = 0; | |||
| 351 | ||||
| 352 | if (d == s) | |||
| 353 | for (;s[i] && i < n; d++, i++) | |||
| 354 | if (s[i] == '%') | |||
| 355 | break; | |||
| 356 | ||||
| 357 | for (;s[i] && i < n; d++, i++) { | |||
| 358 | unsigned char c = s[i], h1, h2; | |||
| 359 | ||||
| 360 | if (c != '%') { | |||
| 361 | if (!IS_SYN33(syn33, c)((syn33 & (1U << (c - 33))) != 0) && IS_EXCLUDED(c, m32, m64, m96)(c <= ' ' || c >= '\177' || (c < 64 ? (m32 & (1 << (63 - c))) : (c < 96 ? (m64 & (1 << (95 - c))) : (m96 & (1 << (127 - c))))) != 0)) | |||
| 362 | return NULL((void*)0); | |||
| 363 | *d = c; | |||
| 364 | continue; | |||
| 365 | } | |||
| 366 | ||||
| 367 | h1 = s[i + 1], h2 = s[i + 2]; | |||
| 368 | ||||
| 369 | if (!IS_HEX(h1)(((h1) >= '0' && (h1) <= '9') || ((h1) >= 'A' && (h1) <= 'F') || ((h1) >= 'a' && (h1 ) <= 'f')) || !IS_HEX(h2)(((h2) >= '0' && (h2) <= '9') || ((h2) >= 'A' && (h2) <= 'F') || ((h2) >= 'a' && (h2 ) <= 'f'))) { | |||
| 370 | *d = '\0'; | |||
| 371 | return NULL((void*)0); | |||
| 372 | } | |||
| 373 | ||||
| 374 | #define UNHEX(a) (a - (a >= 'a' ? 'a' - 10 : (a >= 'A' ? 'A' - 10 : '0'))) | |||
| 375 | c = (UNHEX(h1) << 4) | UNHEX(h2); | |||
| 376 | ||||
| 377 | if (!IS_EXCLUDED(c, m32, m64, m96)(c <= ' ' || c >= '\177' || (c < 64 ? (m32 & (1 << (63 - c))) : (c < 96 ? (m64 & (1 << (95 - c))) : (m96 & (1 << (127 - c))))) != 0)) { | |||
| 378 | /* Convert hex to normal character */ | |||
| 379 | *d = c, i += 2; | |||
| 380 | continue; | |||
| 381 | } | |||
| 382 | ||||
| 383 | /* Convert hex to uppercase */ | |||
| 384 | if (h1 >= 'a' /* && h1 <= 'f' */) | |||
| 385 | h1 = h1 - 'a' + 'A'; | |||
| 386 | if (h2 >= 'a' /* && h2 <= 'f' */) | |||
| 387 | h2 = h2 - 'a' + 'A'; | |||
| 388 | ||||
| 389 | d[0] = '%', d[1] = h1, d[2] = h2; | |||
| 390 | ||||
| 391 | d +=2, i += 2; | |||
| 392 | #undef UNHEX | |||
| 393 | } | |||
| 394 | ||||
| 395 | *d = '\0'; | |||
| 396 | ||||
| 397 | return d; | |||
| 398 | } | |||
| 399 | ||||
| 400 | ||||
| 401 | /** Canonize a URL component (with precomputed mask). | |||
| 402 | * | |||
| 403 | * This version does not flag error if *s contains character that should | |||
| 404 | * be escaped. | |||
| 405 | */ | |||
| 406 | static | |||
| 407 | char *url_canonize3(char *d, char const * const s, size_t n, | |||
| 408 | unsigned m32, unsigned m64, unsigned m96) | |||
| 409 | { | |||
| 410 | size_t i = 0; | |||
| 411 | ||||
| 412 | if (d == s) | |||
| 413 | for (;s[i] && i < n; d++, i++) | |||
| 414 | if (s[i] == '%') | |||
| 415 | break; | |||
| 416 | ||||
| 417 | for (;s[i] && i < n; d++, i++) { | |||
| 418 | unsigned char c = s[i], h1, h2; | |||
| 419 | ||||
| 420 | if (c != '%') { | |||
| 421 | *d = c; | |||
| 422 | continue; | |||
| 423 | } | |||
| 424 | ||||
| 425 | h1 = s[i + 1], h2 = s[i + 2]; | |||
| 426 | ||||
| 427 | if (!IS_HEX(h1)(((h1) >= '0' && (h1) <= '9') || ((h1) >= 'A' && (h1) <= 'F') || ((h1) >= 'a' && (h1 ) <= 'f')) || !IS_HEX(h2)(((h2) >= '0' && (h2) <= '9') || ((h2) >= 'A' && (h2) <= 'F') || ((h2) >= 'a' && (h2 ) <= 'f'))) { | |||
| 428 | *d = '\0'; | |||
| 429 | return NULL((void*)0); | |||
| 430 | } | |||
| 431 | ||||
| 432 | #define UNHEX(a) (a - (a >= 'a' ? 'a' - 10 : (a >= 'A' ? 'A' - 10 : '0'))) | |||
| 433 | c = (UNHEX(h1) << 4) | UNHEX(h2); | |||
| 434 | ||||
| 435 | if (!IS_EXCLUDED(c, m32, m64, m96)(c <= ' ' || c >= '\177' || (c < 64 ? (m32 & (1 << (63 - c))) : (c < 96 ? (m64 & (1 << (95 - c))) : (m96 & (1 << (127 - c))))) != 0)) { | |||
| 436 | *d = c, i += 2; | |||
| 437 | continue; | |||
| 438 | } | |||
| 439 | ||||
| 440 | /* Convert hex to uppercase */ | |||
| 441 | if (h1 >= 'a' /* && h1 <= 'f' */) | |||
| 442 | h1 = h1 - 'a' + 'A'; | |||
| 443 | if (h2 >= 'a' /* && h2 <= 'f' */) | |||
| 444 | h2 = h2 - 'a' + 'A'; | |||
| 445 | ||||
| 446 | d[0] = '%', d[1] = h1, d[2] = h2; | |||
| 447 | ||||
| 448 | d +=2, i += 2; | |||
| 449 | #undef UNHEX | |||
| 450 | } | |||
| 451 | ||||
| 452 | *d = '\0'; | |||
| 453 | ||||
| 454 | return d; | |||
| 455 | } | |||
| 456 | ||||
| 457 | ||||
| 458 | /** Get URL scheme. */ | |||
| 459 | char const* url_scheme(enum url_type_e url_type) | |||
| 460 | { | |||
| 461 | switch (url_type) { | |||
| 462 | case url_any: return "*"; | |||
| 463 | case url_sip: return "sip"; | |||
| 464 | case url_sips: return "sips"; | |||
| 465 | case url_tel: return "tel"; | |||
| 466 | case url_fax: return "fax"; | |||
| 467 | case url_modem: return "modem"; | |||
| 468 | case url_http: return "http"; | |||
| 469 | case url_https: return "https"; | |||
| 470 | case url_ftp: return "ftp"; | |||
| 471 | case url_file: return "file"; | |||
| 472 | case url_rtsp: return "rtsp"; | |||
| 473 | case url_rtspu: return "rtspu"; | |||
| 474 | case url_mailto: return "mailto"; | |||
| 475 | case url_im: return "im"; | |||
| 476 | case url_pres: return "pres"; | |||
| 477 | case url_cid: return "cid"; | |||
| 478 | case url_msrp: return "msrp"; | |||
| 479 | case url_msrps: return "msrps"; | |||
| 480 | case url_urn: return "urn"; | |||
| 481 | case url_wv: return "wv"; | |||
| 482 | default: | |||
| 483 | assert(url_type == url_unknown)((void) sizeof ((url_type == url_unknown) ? 1 : 0), __extension__ ({ if (url_type == url_unknown) ; else __assert_fail ("url_type == url_unknown" , "url.c", 483, __extension__ __PRETTY_FUNCTION__); })); | |||
| 484 | return NULL((void*)0); | |||
| 485 | } | |||
| 486 | } | |||
| 487 | ||||
| 488 | su_inlinestatic inline | |||
| 489 | int url_type_is_opaque(enum url_type_e url_type) | |||
| 490 | { | |||
| 491 | return | |||
| 492 | url_type == url_invalid || | |||
| 493 | url_type == url_tel || | |||
| 494 | url_type == url_modem || | |||
| 495 | url_type == url_fax || | |||
| 496 | url_type == url_cid; | |||
| 497 | } | |||
| 498 | ||||
| 499 | /** Init an url as given type */ | |||
| 500 | void url_init(url_t *url, enum url_type_e type) | |||
| 501 | { | |||
| 502 | memset(url, 0, sizeof(*url)); | |||
| 503 | url->url_type = type; | |||
| 504 | if (type > url_unknown) { | |||
| 505 | char const *scheme = url_scheme((enum url_type_e)url->url_type); | |||
| 506 | if (scheme) | |||
| 507 | url->url_scheme = scheme; | |||
| 508 | } | |||
| 509 | } | |||
| 510 | ||||
| 511 | /** Get url type */ | |||
| 512 | su_inlinestatic inline | |||
| 513 | enum url_type_e url_get_type(char const *scheme, size_t len) | |||
| 514 | { | |||
| 515 | #define test_scheme(s) \ | |||
| 516 | if (len == strlen(#s) && !strncasecmp(scheme, #s, len)) return url_##s | |||
| 517 | ||||
| 518 | switch (scheme[0]) { | |||
| 519 | case '*': if (strcmp(scheme, "*") == 0) return url_any; | |||
| 520 | case 'c': case 'C': | |||
| 521 | test_scheme(cid); break; | |||
| 522 | case 'f': case 'F': | |||
| 523 | test_scheme(ftp); test_scheme(file); test_scheme(fax); break; | |||
| 524 | case 'h': case 'H': | |||
| 525 | test_scheme(http); test_scheme(https); break; | |||
| 526 | case 'i': case 'I': | |||
| 527 | test_scheme(im); break; | |||
| 528 | case 'm': case 'M': | |||
| 529 | test_scheme(mailto); test_scheme(modem); | |||
| 530 | test_scheme(msrp); test_scheme(msrps); break; | |||
| 531 | case 'p': case 'P': | |||
| 532 | test_scheme(pres); break; | |||
| 533 | case 'r': case 'R': | |||
| 534 | test_scheme(rtsp); test_scheme(rtspu); break; | |||
| 535 | case 's': case 'S': | |||
| 536 | test_scheme(sip); test_scheme(sips); break; | |||
| 537 | case 't': case 'T': | |||
| 538 | test_scheme(tel); break; | |||
| 539 | case 'u': case 'U': | |||
| 540 | test_scheme(urn); break; | |||
| 541 | case 'w': case 'W': | |||
| 542 | test_scheme(wv); break; | |||
| 543 | ||||
| 544 | ||||
| 545 | default: break; | |||
| 546 | } | |||
| 547 | ||||
| 548 | #undef test_scheme | |||
| 549 | ||||
| 550 | if (len != span_unreserved(scheme)) | |||
| 551 | return url_invalid; | |||
| 552 | else | |||
| 553 | return url_unknown; | |||
| 554 | } | |||
| 555 | ||||
| 556 | /** | |||
| 557 | * Decode a URL. | |||
| 558 | * | |||
| 559 | * This function decodes a (SIP) URL string to a url_t structure. | |||
| 560 | * | |||
| 561 | * @param url structure to store the parsing result | |||
| 562 | * @param s NUL-terminated string to be parsed | |||
| 563 | * | |||
| 564 | * @note The parsed string @a s will be modified when parsing it. | |||
| 565 | * | |||
| 566 | * @retval 0 if successful, | |||
| 567 | * @retval -1 otherwise. | |||
| 568 | */ | |||
| 569 | static | |||
| 570 | int _url_d(url_t *url, char *s) | |||
| 571 | { | |||
| 572 | size_t n, p; | |||
| 573 | char rest_c, *host, *user; | |||
| 574 | int have_authority = 1; | |||
| 575 | ||||
| 576 | memset(url, 0, sizeof(*url)); | |||
| 577 | ||||
| 578 | if (strcmp(s, "*") == 0) { | |||
| 579 | url->url_type = url_any; | |||
| 580 | url->url_scheme = "*"; | |||
| 581 | return 0; | |||
| 582 | } | |||
| 583 | ||||
| 584 | n = strcspn(s, ":/?#"); | |||
| 585 | ||||
| 586 | if (n && s[n] == ':') { | |||
| 587 | char *scheme; | |||
| 588 | url->url_scheme = scheme = s; s[n] = '\0'; s = s + n + 1; | |||
| 589 | ||||
| 590 | if (!(scheme = url_canonize(scheme, scheme, SIZE_MAX(18446744073709551615UL), 0, "+"))) | |||
| 591 | return -1; | |||
| 592 | ||||
| 593 | n = scheme - url->url_scheme; | |||
| 594 | ||||
| 595 | url->url_type = url_get_type(url->url_scheme, n); | |||
| 596 | ||||
| 597 | have_authority = !url_type_is_opaque((enum url_type_e)url->url_type); | |||
| 598 | } | |||
| 599 | else { | |||
| 600 | url->url_type = url_unknown; | |||
| 601 | } | |||
| 602 | ||||
| 603 | user = NULL((void*)0), host = s; | |||
| 604 | ||||
| 605 | if (url->url_type == url_sip || url->url_type == url_sips) { | |||
| 606 | /* SIP URL may have /;? in user part but no path */ | |||
| 607 | /* user-unreserved = "&" / "=" / "+" / "$" / "," / ";" / "?" / "/" */ | |||
| 608 | /* Some #*@#* phones include unescaped # there, too */ | |||
| 609 | n = strcspn(s, "@/;?#"); | |||
| 610 | p = strcspn(s + n, "@"); | |||
| 611 | if (s[n + p] == '@') { | |||
| 612 | n += p; | |||
| 613 | user = s; | |||
| 614 | host = s + n + 1; | |||
| 615 | } | |||
| 616 | ||||
| 617 | n += strcspn(s + n, "/;?#"); | |||
| 618 | } | |||
| 619 | else if (have_authority) { | |||
| 620 | if (url->url_type == url_wv) { | |||
| 621 | /* WV URL may have / in user part */ | |||
| 622 | n = strcspn(s, "@#?;"); | |||
| 623 | if (s[n] == '@') { | |||
| 624 | user = s; | |||
| 625 | host = s + n + 1; | |||
| 626 | n += strcspn(s + n, ";?#"); | |||
| 627 | } | |||
| 628 | } | |||
| 629 | else if (host[0] == '/' && host[1] != '/') { | |||
| 630 | /* foo:/bar or /bar - no authority, just path */ | |||
| 631 | url->url_root = '/'; /* Absolute path */ | |||
| 632 | host = NULL((void*)0), n = 0; | |||
| 633 | } | |||
| 634 | else { | |||
| 635 | if (host[0] == '/' && host[1] == '/') { | |||
| 636 | /* We have authority, / / foo or foo */ | |||
| 637 | host += 2; s += 2, url->url_root = '/'; | |||
| 638 | n = strcspn(s, "/?#@[]"); | |||
| 639 | } | |||
| 640 | else | |||
| 641 | n = strcspn(s, "@;/?#"); | |||
| 642 | ||||
| 643 | if (s[n] == '@') | |||
| 644 | user = host, host = user + n + 1; | |||
| 645 | ||||
| 646 | n += strcspn(s + n, ";/?#"); /* Find path, query and/or fragment */ | |||
| 647 | } | |||
| 648 | } | |||
| 649 | else /* !have_authority */ { | |||
| 650 | user = host, host = NULL((void*)0); | |||
| 651 | if (url->url_type != url_invalid) | |||
| 652 | n = strcspn(s, "/;?#"); /* Find params, query and/or fragment */ | |||
| 653 | else | |||
| 654 | n = strcspn(s, "#"); | |||
| 655 | } | |||
| 656 | ||||
| 657 | rest_c = s[n]; s[n] = 0; s = rest_c ? s + n + 1 : NULL((void*)0); | |||
| 658 | ||||
| 659 | if (user) { | |||
| 660 | if (host) host[-1] = '\0'; | |||
| 661 | url->url_user = user; | |||
| 662 | if (url->url_type != url_unknown) { | |||
| 663 | n = strcspn(user, ":"); | |||
| 664 | if (user[n]) { | |||
| 665 | user[n] = '\0'; | |||
| 666 | url->url_password = user + n + 1; | |||
| 667 | } | |||
| 668 | } | |||
| 669 | } | |||
| 670 | ||||
| 671 | if (host) { | |||
| 672 | url->url_host = host; | |||
| 673 | /* IPv6 (and in some cases, IPv4) addresses are quoted with [] */ | |||
| 674 | if (host[0] == '[') { | |||
| 675 | n = strcspn(host, "]"); | |||
| 676 | if (host[n] && (host[n + 1] == '\0' || host[n + 1] == ':')) | |||
| 677 | n++; | |||
| 678 | else | |||
| 679 | n = 0; | |||
| 680 | } | |||
| 681 | else { | |||
| 682 | n = strcspn(host, ":"); | |||
| 683 | } | |||
| 684 | ||||
| 685 | /* We allow empty host by default */ | |||
| 686 | if (n == 0) switch (url->url_type) { | |||
| 687 | case url_sip: | |||
| 688 | case url_sips: | |||
| 689 | case url_im: | |||
| 690 | case url_pres: | |||
| 691 | return -1; | |||
| 692 | default: | |||
| 693 | break; | |||
| 694 | } | |||
| 695 | ||||
| 696 | if (host[n] == ':') { | |||
| 697 | char *port = host + n + 1; | |||
| 698 | url->url_port = port; | |||
| 699 | switch (url->url_type) { | |||
| 700 | case url_any: | |||
| 701 | case url_sip: | |||
| 702 | case url_sips: | |||
| 703 | case url_http: | |||
| 704 | case url_https: | |||
| 705 | case url_ftp: | |||
| 706 | case url_file: | |||
| 707 | case url_rtsp: | |||
| 708 | case url_rtspu: | |||
| 709 | if (!url_canonize2(port, port, SIZE_MAX(18446744073709551615UL), 0, RESERVED_MASK0xbe19003f, 0x8000001e, 0x8000001d)) | |||
| 710 | return -1; | |||
| 711 | ||||
| 712 | /* Check that port is really numeric or wildcard */ | |||
| 713 | /* Port can be *digit, empty string or "*" */ | |||
| 714 | while (*port >= '0' && *port <= '9') | |||
| 715 | port++; | |||
| 716 | ||||
| 717 | if (port != url->url_port) { | |||
| 718 | if (port[0] != '\0') | |||
| 719 | return -1; | |||
| 720 | } | |||
| 721 | else if (port[0] == '\0') | |||
| 722 | /* empty string */; | |||
| 723 | else if (port[0] == '*' && port[1] == '\0') | |||
| 724 | /* wildcard */; | |||
| 725 | else | |||
| 726 | return -1; | |||
| 727 | } | |||
| 728 | host[n] = 0; | |||
| 729 | } | |||
| 730 | } | |||
| 731 | ||||
| 732 | if (rest_c == '/') { | |||
| 733 | url->url_path = s; n = strcspn(s, "?#"); | |||
| 734 | rest_c = s[n]; s[n] = 0; s = rest_c ? s + n + 1 : NULL((void*)0); | |||
| 735 | } | |||
| 736 | if (rest_c == ';') { | |||
| 737 | url->url_params = s; n = strcspn(s, "?#"); | |||
| 738 | rest_c = s[n]; s[n] = 0; s = rest_c ? s + n + 1 : NULL((void*)0); | |||
| 739 | } | |||
| 740 | if (rest_c == '?') { | |||
| 741 | url->url_headers = s; n = strcspn(s, "#"); | |||
| 742 | rest_c = s[n]; s[n] = 0; s = rest_c ? s + n + 1 : NULL((void*)0); | |||
| 743 | } | |||
| 744 | if (rest_c == '#') { | |||
| 745 | url->url_fragment = s; | |||
| 746 | rest_c = '\0'; | |||
| 747 | } | |||
| 748 | if (rest_c) | |||
| 749 | return -1; | |||
| 750 | ||||
| 751 | return 0; | |||
| 752 | } | |||
| 753 | ||||
| 754 | /* Unreserved things */ | |||
| 755 | ||||
| 756 | /** | |||
| 757 | * Decode a URL. | |||
| 758 | * | |||
| 759 | * This function decodes a URL string to a url_t structure. | |||
| 760 | * | |||
| 761 | * @param url structure to store the parsing result | |||
| 762 | * @param s NUL-terminated string to be parsed | |||
| 763 | * | |||
| 764 | * @note The parsed string @a s will be modified when parsing it. | |||
| 765 | * | |||
| 766 | * @retval 0 if successful, | |||
| 767 | * @retval -1 otherwise. | |||
| 768 | */ | |||
| 769 | int url_d(url_t *url, char *s) | |||
| 770 | { | |||
| 771 | if (url == NULL((void*)0) || _url_d(url, s) < 0) | |||
| 772 | return -1; | |||
| 773 | ||||
| 774 | /* Canonize URL */ | |||
| 775 | /* scheme is canonized by _url_d() */ | |||
| 776 | if (url->url_type == url_sip || url->url_type == url_sips) { | |||
| 777 | ||||
| 778 | # define SIP_USER_UNRESERVED"&=+$,;?/" "&=+$,;?/" | |||
| 779 | s = (char *)url->url_user; | |||
| 780 | if (s && !url_canonize(s, s, SIZE_MAX(18446744073709551615UL), 0, SIP_USER_UNRESERVED"&=+$,;?/")) | |||
| 781 | return -1; | |||
| 782 | ||||
| 783 | /* Having different charset in user and password does not make sense */ | |||
| 784 | /* but that is how it is defined in RFC 3261 */ | |||
| 785 | # define SIP_PASS_UNRESERVED"&=+$," "&=+$," | |||
| 786 | s = (char *)url->url_password; | |||
| 787 | if (s && !url_canonize(s, s, SIZE_MAX(18446744073709551615UL), 0, SIP_PASS_UNRESERVED"&=+$,")) | |||
| 788 | return -1; | |||
| 789 | ||||
| 790 | } | |||
| 791 | else { | |||
| 792 | ||||
| 793 | # define USER_UNRESERVED"&=+$,;" "&=+$,;" | |||
| 794 | s = (char *)url->url_user; | |||
| 795 | if (s && !url_canonize(s, s, SIZE_MAX(18446744073709551615UL), 0, USER_UNRESERVED"&=+$,;")) | |||
| 796 | return -1; | |||
| 797 | ||||
| 798 | # define PASS_UNRESERVED"&=+$,;:" "&=+$,;:" | |||
| 799 | s = (char *)url->url_password; | |||
| 800 | if (s && !url_canonize(s, s, SIZE_MAX(18446744073709551615UL), 0, PASS_UNRESERVED"&=+$,;:")) | |||
| 801 | return -1; | |||
| 802 | } | |||
| 803 | ||||
| 804 | s = (char *)url->url_host; | |||
| 805 | if (s && !url_canonize2(s, s, SIZE_MAX(18446744073709551615UL), 0, RESERVED_MASK0xbe19003f, 0x8000001e, 0x8000001d)) | |||
| 806 | return -1; | |||
| 807 | ||||
| 808 | /* port is canonized by _url_d() */ | |||
| 809 | s = (char *)url->url_path; | |||
| 810 | if (s && !url_canonize(s, s, SIZE_MAX(18446744073709551615UL), | |||
| 811 | /* Allow all URI characters but ? */ | |||
| 812 | /* Allow unescaped /;?@, - but do not convert */ | |||
| 813 | SYN33('/')(1U << ('/' - 33)) | SYN33(';')(1U << (';' - 33)) | SYN33('=')(1U << ('=' - 33)) | SYN33('@')(1U << ('@' - 33)) | | |||
| 814 | SYN33(',')(1U << (',' - 33)), | |||
| 815 | /* Convert escaped :&+$ to unescaped */ | |||
| 816 | ":&+$")) | |||
| 817 | return -1; | |||
| 818 | ||||
| 819 | s = (char *)url->url_params; | |||
| 820 | if (s && !url_canonize(s, s, SIZE_MAX(18446744073709551615UL), | |||
| 821 | /* Allow all URI characters but ? */ | |||
| 822 | /* Allow unescaped ;=@, - but do not convert */ | |||
| 823 | SYN33(';')(1U << (';' - 33)) | SYN33('=')(1U << ('=' - 33)) | SYN33('@')(1U << ('@' - 33)) | SYN33(',')(1U << (',' - 33)), | |||
| 824 | /* Convert escaped /:&+$ to unescaped */ | |||
| 825 | "/:&+$")) | |||
| 826 | return -1; | |||
| 827 | ||||
| 828 | /* Unhex alphanumeric and unreserved URI characters */ | |||
| 829 | s = (char *)url->url_headers; | |||
| 830 | if (s && !url_canonize3(s, s, SIZE_MAX(18446744073709551615UL), RESERVED_MASK0xbe19003f, 0x8000001e, 0x8000001d)) | |||
| 831 | return -1; | |||
| 832 | ||||
| 833 | /* Allow all URI characters (including reserved ones) */ | |||
| 834 | s = (char *)url->url_fragment; | |||
| 835 | if (s && !url_canonize2(s, s, SIZE_MAX(18446744073709551615UL), 0, URIC_MASK0xb400000a, 0x0000001e, 0x8000001d)) | |||
| 836 | return -1; | |||
| 837 | ||||
| 838 | return 0; | |||
| 839 | } | |||
| 840 | ||||
| 841 | /** Encode an URL. | |||
| 842 | * | |||
| 843 | * The function url_e() combines a URL from substrings in url_t structure | |||
| 844 | * according the @ref url_syntax "URL syntax" presented above. The encoded | |||
| 845 | * @a url is stored in a @a buffer of @a n bytes. | |||
| 846 | * | |||
| 847 | * @param buffer memory area to store the encoded @a url. | |||
| 848 | * @param n size of @a buffer. | |||
| 849 | * @param url URL to be encoded. | |||
| 850 | * | |||
| 851 | * @return | |||
| 852 | * Return the number of bytes in the encoding. | |||
| 853 | * | |||
| 854 | * @note The function follows the convention set by C99 snprintf(). Even if | |||
| 855 | * the result does not fit into the @a buffer and it is truncated, the | |||
| 856 | * function returns the number of bytes in an untruncated encoding. | |||
| 857 | */ | |||
| 858 | issize_t url_e(char buffer[], isize_t n, url_t const *url) | |||
| 859 | { | |||
| 860 | size_t i; | |||
| 861 | char *b = buffer; | |||
| 862 | size_t m = n; | |||
| 863 | int do_copy = n > 0; | |||
| 864 | ||||
| 865 | if (url == NULL((void*)0)) | |||
| 866 | return -1; | |||
| 867 | ||||
| 868 | if (URL_STRING_P(url)((url) && *((url_string_t*)(url))->us_str != 0)) { | |||
| 869 | char const *u = (char *)url; | |||
| 870 | i = strlen(u); | |||
| 871 | if (!buffer) | |||
| 872 | return i; | |||
| 873 | ||||
| 874 | if (i >= n) { | |||
| 875 | memcpy(buffer, u, n - 2); | |||
| 876 | buffer[n - 1] = '\0'; | |||
| 877 | } else { | |||
| 878 | memcpy(buffer, u, i + 1); | |||
| 879 | } | |||
| 880 | ||||
| 881 | return i; | |||
| 882 | } | |||
| 883 | ||||
| 884 | ||||
| 885 | if (url->url_type == url_any) { | |||
| 886 | if (b && m > 0) { | |||
| 887 | if (m > 1) strcpy(b, "*"); else b[0] = '\0'; | |||
| 888 | } | |||
| 889 | return 1; | |||
| 890 | } | |||
| 891 | ||||
| 892 | if (url->url_scheme && url->url_scheme[0]) { | |||
| 893 | i = strlen(url->url_scheme) + 1; | |||
| 894 | if (do_copy && (do_copy = i <= n)) { | |||
| 895 | memcpy(b, url->url_scheme, i - 1); | |||
| 896 | b[i - 1] = ':'; | |||
| 897 | } | |||
| 898 | b += i; n -= i; | |||
| 899 | } | |||
| 900 | ||||
| 901 | if (url->url_root && (url->url_host || url->url_user)) { | |||
| 902 | if (do_copy && (do_copy = 2 <= n)) | |||
| 903 | memcpy(b, "//", 2); | |||
| 904 | b += 2; n -= 2; | |||
| 905 | } | |||
| 906 | ||||
| 907 | if (url->url_user) { | |||
| 908 | i = strlen(url->url_user); | |||
| 909 | if (do_copy && (do_copy = i <= n)) | |||
| 910 | memcpy(b, url->url_user, i); | |||
| 911 | b += i; n -= i; | |||
| 912 | ||||
| 913 | if (url->url_password) { | |||
| 914 | if (do_copy && (do_copy = 1 <= n)) | |||
| 915 | *b = ':'; | |||
| 916 | b++; n--; | |||
| 917 | i = strlen(url->url_password); | |||
| 918 | if (do_copy && (do_copy = i <= n)) | |||
| 919 | memcpy(b, url->url_password, i); | |||
| 920 | b += i; n -= i; | |||
| 921 | } | |||
| 922 | ||||
| 923 | if (url->url_host) { | |||
| 924 | if (do_copy && (do_copy = 1 <= n)) | |||
| 925 | *b = '@'; | |||
| 926 | b++; n--; | |||
| 927 | } | |||
| 928 | } | |||
| 929 | ||||
| 930 | if (url->url_host) { | |||
| 931 | i = strlen(url->url_host); | |||
| 932 | if (do_copy && (do_copy = i <= n)) | |||
| 933 | memcpy(b, url->url_host, i); | |||
| 934 | b += i; n -= i; | |||
| 935 | ||||
| 936 | if (url->url_port) { | |||
| 937 | i = strlen(url->url_port) + 1; | |||
| 938 | if (do_copy && (do_copy = i <= n)) { | |||
| 939 | b[0] = ':'; | |||
| 940 | memcpy(b + 1, url->url_port, i - 1); | |||
| 941 | } | |||
| 942 | b += i; n -= i; | |||
| 943 | } | |||
| 944 | } | |||
| 945 | ||||
| 946 | if (url->url_path) { | |||
| 947 | if (url->url_root) { | |||
| 948 | if (do_copy && (do_copy = 1 <= n)) | |||
| 949 | b[0] = '/'; | |||
| 950 | b++, n--; | |||
| 951 | } | |||
| 952 | i = strlen(url->url_path); | |||
| 953 | if (do_copy && (do_copy = i < n)) | |||
| 954 | memcpy(b, url->url_path, i); | |||
| 955 | b += i; n -= i; | |||
| 956 | } | |||
| 957 | ||||
| 958 | { | |||
| 959 | static char const sep[] = ";?#"; | |||
| 960 | char const *pp[3]; | |||
| 961 | size_t j; | |||
| 962 | ||||
| 963 | pp[0] = url->url_params; | |||
| 964 | pp[1] = url->url_headers; | |||
| 965 | pp[2] = url->url_fragment; | |||
| 966 | ||||
| 967 | for (j = 0; j < 3; j++) { | |||
| 968 | char const *p = pp[j]; | |||
| 969 | if (!p) continue; | |||
| 970 | i = strlen(p) + 1; | |||
| 971 | if (do_copy && (do_copy = i <= n)) { | |||
| 972 | *b = sep[j]; | |||
| 973 | memcpy(b + 1, p, i - 1); | |||
| 974 | } | |||
| 975 | b += i; n -= i; | |||
| 976 | } | |||
| 977 | } | |||
| 978 | ||||
| 979 | if (do_copy && (do_copy = 1 <= n)) | |||
| 980 | *b = '\0'; | |||
| 981 | else if (buffer && m > 0) | |||
| 982 | buffer[m - 1] = '\0'; | |||
| 983 | ||||
| 984 | assert((size_t)(b - buffer) == (size_t)(m - n))((void) sizeof (((size_t)(b - buffer) == (size_t)(m - n)) ? 1 : 0), __extension__ ({ if ((size_t)(b - buffer) == (size_t)( m - n)) ; else __assert_fail ("(size_t)(b - buffer) == (size_t)(m - n)" , "url.c", 984, __extension__ __PRETTY_FUNCTION__); })); | |||
| 985 | ||||
| 986 | /* This follows the snprintf(C99) return value, | |||
| 987 | * Number of characters written (excluding NUL) | |||
| 988 | */ | |||
| 989 | return b - buffer; | |||
| 990 | } | |||
| 991 | ||||
| 992 | ||||
| 993 | /** Calculate the length of URL when encoded. | |||
| 994 | * | |||
| 995 | */ | |||
| 996 | isize_t url_len(url_t const * url) | |||
| 997 | { | |||
| 998 | size_t rv = 0; | |||
| 999 | ||||
| 1000 | if (url->url_scheme) rv += strlen(url->url_scheme) + 1; /* plus ':' */ | |||
| 1001 | if (url->url_user) { | |||
| 1002 | rv += strlen(url->url_user); | |||
| 1003 | if (url->url_password) | |||
| 1004 | rv += strlen(url->url_password) + 1; /* plus ':' */ | |||
| 1005 | rv += url->url_host != NULL((void*)0); /* plus '@' */ | |||
| 1006 | } | |||
| 1007 | if (url->url_host) rv += strlen(url->url_host); | |||
| 1008 | if (url->url_port) rv += strlen(url->url_port) + 1; /* plus ':' */ | |||
| 1009 | if (url->url_path) rv += strlen(url->url_path) + 1; /* plus initial / */ | |||
| 1010 | if (url->url_params) rv += strlen(url->url_params) + 1; /* plus initial ; */ | |||
| 1011 | if (url->url_headers) rv += strlen(url->url_headers) + 1; /* plus '?' */ | |||
| 1012 | if (url->url_fragment) rv += strlen(url->url_fragment) + 1; /* plus '#' */ | |||
| 1013 | ||||
| 1014 | return rv; | |||
| 1015 | } | |||
| 1016 | ||||
| 1017 | /**@def URL_E(buf, end, url) | |||
| 1018 | * Encode an URL: use @a buf up to @a end. | |||
| 1019 | * @hideinitializer | |||
| 1020 | */ | |||
| 1021 | ||||
| 1022 | /** | |||
| 1023 | * Calculate the size of strings associated with a #url_t sructure. | |||
| 1024 | * | |||
| 1025 | * @param url pointer to a #url_t structure or string | |||
| 1026 | * @return Number of bytes for URL | |||
| 1027 | */ | |||
| 1028 | isize_t url_xtra(url_t const *url) | |||
| 1029 | { | |||
| 1030 | size_t xtra; | |||
| 1031 | ||||
| 1032 | if (URL_STRING_P(url)((url) && *((url_string_t*)(url))->us_str != 0)) { | |||
| 1033 | xtra = strlen((char const *)url) + 1; | |||
| 1034 | } | |||
| 1035 | else { | |||
| 1036 | size_t len_scheme, len_user, len_password, | |||
| 1037 | len_host, len_port, len_path, len_params, | |||
| 1038 | len_headers, len_fragment; | |||
| 1039 | ||||
| 1040 | len_scheme = (url->url_type <= url_unknown && url->url_scheme) ? | |||
| 1041 | strlen(url->url_scheme) + 1 : 0; | |||
| 1042 | len_user = url->url_user ? strlen(url->url_user) + 1 : 0; | |||
| 1043 | len_password = url->url_password ? strlen(url->url_password) + 1 : 0; | |||
| 1044 | len_host = url->url_host ? strlen(url->url_host) + 1 : 0; | |||
| 1045 | len_port = url->url_port ? strlen(url->url_port) + 1 : 0; | |||
| 1046 | len_path = url->url_path ? strlen(url->url_path) + 1 : 0; | |||
| 1047 | len_params = url->url_params ? strlen(url->url_params) + 1 : 0; | |||
| 1048 | len_headers = url->url_headers ? strlen(url->url_headers) + 1 : 0; | |||
| 1049 | len_fragment = url->url_fragment ? strlen(url->url_fragment) + 1 : 0; | |||
| 1050 | ||||
| 1051 | xtra = | |||
| 1052 | len_scheme + len_user + len_password + len_host + len_port + | |||
| 1053 | len_path + len_params + len_headers + len_fragment; | |||
| 1054 | } | |||
| 1055 | ||||
| 1056 | return xtra; | |||
| 1057 | } | |||
| 1058 | ||||
| 1059 | su_inlinestatic inline | |||
| 1060 | char *copy(char *buf, char *end, char const *src) | |||
| 1061 | { | |||
| 1062 | #if HAVE_MEMCCPY1 | |||
| 1063 | char *b = memccpy(buf, src, '\0', end - buf); | |||
| 1064 | if (b) | |||
| 1065 | return b; | |||
| 1066 | else | |||
| 1067 | return end + strlen(src + (end - buf)) + 1; | |||
| 1068 | #else | |||
| 1069 | for (; buf < end && (*buf = *src); buf++, src++) | |||
| 1070 | ; | |||
| 1071 | ||||
| 1072 | if (buf >= end) | |||
| 1073 | while (*src++) | |||
| 1074 | buf++; | |||
| 1075 | ||||
| 1076 | return buf + 1; | |||
| 1077 | #endif | |||
| 1078 | } | |||
| 1079 | ||||
| 1080 | /** | |||
| 1081 | * Duplicate the url. | |||
| 1082 | * | |||
| 1083 | * The function url_dup() copies the url structure @a src and the strings | |||
| 1084 | * attached to it to @a url. The non-constant strings in @a src are copied | |||
| 1085 | * to @a buf. If the size of duplicated strings exceed @a bufsize, the | |||
| 1086 | * corresponding string fields in @a url are set to NULL. | |||
| 1087 | * | |||
| 1088 | * The calling function can calculate the size of buffer required by calling | |||
| 1089 | * url_dup() with zero as @a bufsize and NULL as @a dst. | |||
| 1090 | ||||
| 1091 | * @param buf Buffer for non-constant strings copied from @a src. | |||
| 1092 | * @param bufsize Size of @a buf. | |||
| 1093 | * @param dst Destination URL structure. | |||
| 1094 | * @param src Source URL structure. | |||
| 1095 | * | |||
| 1096 | * @return Number of characters required for | |||
| 1097 | * duplicating the strings in @a str, or -1 if an error | |||
| 1098 | * occurred. | |||
| 1099 | */ | |||
| 1100 | issize_t url_dup(char *buf, isize_t bufsize, url_t *dst, url_t const *src) | |||
| 1101 | { | |||
| 1102 | if (!src && !dst) | |||
| 1103 | return -1; | |||
| 1104 | else if (URL_STRING_P(src)((src) && *((url_string_t*)(src))->us_str != 0)) { | |||
| 1105 | size_t n = strlen((char *)src) + 1; | |||
| 1106 | if (n > bufsize || dst == NULL((void*)0)) | |||
| 1107 | return n; | |||
| 1108 | ||||
| 1109 | strcpy(buf, (char *)src); | |||
| 1110 | memset(dst, 0, sizeof(*dst)); | |||
| 1111 | if (url_d(dst, buf) < 0) | |||
| 1112 | return -1; | |||
| 1113 | ||||
| 1114 | return n; | |||
| 1115 | } | |||
| 1116 | else { | |||
| 1117 | char *b = buf; | |||
| 1118 | char *end = b + bufsize; | |||
| 1119 | char const **dstp; | |||
| 1120 | char const * const *srcp; | |||
| 1121 | url_t dst0[1]; | |||
| 1122 | ||||
| 1123 | if (dst == NULL((void*)0)) | |||
| 1124 | dst = dst0; | |||
| 1125 | ||||
| 1126 | memset(dst, 0, sizeof(*dst)); | |||
| 1127 | ||||
| 1128 | if (!src) | |||
| 1129 | return 0; | |||
| 1130 | ||||
| 1131 | memset(dst->url_pad, 0, sizeof dst->url_pad); | |||
| 1132 | dst->url_type = src->url_type; | |||
| 1133 | dst->url_root = src->url_root; | |||
| 1134 | ||||
| 1135 | dstp = &dst->url_scheme; | |||
| 1136 | srcp = &src->url_scheme; | |||
| 1137 | ||||
| 1138 | if (dst->url_type > url_unknown) | |||
| 1139 | *dstp = url_scheme((enum url_type_e)dst->url_type); | |||
| 1140 | ||||
| 1141 | if (*dstp != NULL((void*)0)) | |||
| 1142 | dstp++, srcp++; /* Skip scheme if it is constant */ | |||
| 1143 | ||||
| 1144 | if (dst != dst0 && buf != NULL((void*)0) && bufsize != 0) | |||
| 1145 | for (; srcp <= &src->url_fragment; srcp++, dstp++) | |||
| 1146 | if (*srcp) { | |||
| 1147 | char *next = copy(b, end, *srcp); | |||
| 1148 | ||||
| 1149 | if (next > end) | |||
| 1150 | break; | |||
| 1151 | ||||
| 1152 | *dstp = b, b = next; | |||
| 1153 | } | |||
| 1154 | ||||
| 1155 | for (; srcp <= &src->url_fragment; srcp++) | |||
| 1156 | if (*srcp) { | |||
| 1157 | b += strlen(*srcp) + 1; | |||
| 1158 | } | |||
| 1159 | ||||
| 1160 | return b - buf; | |||
| 1161 | } | |||
| 1162 | } | |||
| 1163 | ||||
| 1164 | /**@def URL_DUP(buf, end, dst, src) | |||
| 1165 | * Duplicate the url: use @a buf up to @a end. @HI | |||
| 1166 | * | |||
| 1167 | * The macro URL_DUP() duplicates the url. The non-constant strings in @a | |||
| 1168 | * src are copied to @a buf. However, no strings are copied past @a end. | |||
| 1169 | * In other words, the size of buffer is @a end - @a buf. | |||
| 1170 | * | |||
| 1171 | * The macro updates the buffer pointer @a buf, so that it points to the | |||
| 1172 | * first unused byte in the buffer. The buffer pointer @a buf is updated, | |||
| 1173 | * even if the buffer is too small for the duplicated strings. | |||
| 1174 | * | |||
| 1175 | * @param buf Buffer for non-constant strings copied from @a src. | |||
| 1176 | * @param end End of @a buf. | |||
| 1177 | * @param dst Destination URL structure. | |||
| 1178 | * @param src Source URL structure. | |||
| 1179 | * | |||
| 1180 | * @return | |||
| 1181 | * The macro URL_DUP() returns pointer to first unused byte in the | |||
| 1182 | * buffer @a buf. | |||
| 1183 | */ | |||
| 1184 | ||||
| 1185 | /** Duplicate the url to memory allocated via home. | |||
| 1186 | * | |||
| 1187 | * The function url_hdup() duplicates (deep copies) an #url_t structure. | |||
| 1188 | * Alternatively, it can be passed a string; string is then copied and | |||
| 1189 | * parsed to the #url_t structure. | |||
| 1190 | * | |||
| 1191 | * The function url_hdup() allocates the destination structure from @a home | |||
| 1192 | * as a single memory block. It is possible to free the copied url structure | |||
| 1193 | * and all the associated strings using a single call to su_free(). | |||
| 1194 | * | |||
| 1195 | * @param home memory home used to allocate new url object | |||
| 1196 | * @param src pointer to URL (or string) | |||
| 1197 | * | |||
| 1198 | * @return | |||
| 1199 | * The function url_hdup() returns a pointer to the newly allocated #url_t | |||
| 1200 | * structure, or NULL upon an error. | |||
| 1201 | */ | |||
| 1202 | url_t *url_hdup(su_home_t *home, url_t const *src) | |||
| 1203 | { | |||
| 1204 | if (src) { | |||
| 1205 | size_t len = sizeof(*src) + url_xtra(src); | |||
| 1206 | url_t *dst = su_alloc(home, len); | |||
| 1207 | if (dst) { | |||
| 1208 | ssize_t actual; | |||
| 1209 | actual = url_dup((char *)(dst + 1), len - sizeof(*src), dst, src); | |||
| 1210 | if (actual < 0) | |||
| 1211 | su_free(home, dst), dst = NULL((void*)0); | |||
| 1212 | else | |||
| 1213 | assert(len == sizeof(*src) + actual)((void) sizeof ((len == sizeof(*src) + actual) ? 1 : 0), __extension__ ({ if (len == sizeof(*src) + actual) ; else __assert_fail ("len == sizeof(*src) + actual" , "url.c", 1213, __extension__ __PRETTY_FUNCTION__); })); | |||
| 1214 | } | |||
| 1215 | return dst; | |||
| 1216 | } | |||
| 1217 | else | |||
| 1218 | return NULL((void*)0); | |||
| 1219 | } | |||
| 1220 | ||||
| 1221 | ||||
| 1222 | /** Convert an string to an url */ | |||
| 1223 | url_t *url_make(su_home_t *h, char const *str) | |||
| 1224 | { | |||
| 1225 | return url_hdup(h, URL_STRING_MAKE(str)((url_string_t *)((str) && *((char *)(str)) ? (str) : ((void*)0)))->us_url); | |||
| 1226 | } | |||
| 1227 | ||||
| 1228 | /** Print an URL */ | |||
| 1229 | url_t *url_format(su_home_t *h, char const *fmt, ...) | |||
| 1230 | { | |||
| 1231 | url_t *url; | |||
| 1232 | char *us; | |||
| 1233 | va_list ap; | |||
| 1234 | ||||
| 1235 | va_start(ap, fmt)__builtin_va_start(ap, fmt); | |||
| 1236 | ||||
| 1237 | us = su_vsprintf(h, fmt, ap); | |||
| 1238 | ||||
| 1239 | va_end(ap)__builtin_va_end(ap); | |||
| 1240 | ||||
| 1241 | if (us == NULL((void*)0)) | |||
| 1242 | return NULL((void*)0); | |||
| 1243 | ||||
| 1244 | url = url_hdup(h, URL_STRING_MAKE(us)((url_string_t *)((us) && *((char *)(us)) ? (us) : (( void*)0)))->us_url); | |||
| 1245 | ||||
| 1246 | su_free(h, us); | |||
| 1247 | ||||
| 1248 | return url; | |||
| 1249 | } | |||
| 1250 | ||||
| 1251 | ||||
| 1252 | /** Convert @a url to a string allocated from @a home. | |||
| 1253 | * | |||
| 1254 | * @param home memory home to allocate the new string | |||
| 1255 | * @param url url to convert to string | |||
| 1256 | * | |||
| 1257 | * The @a url can be a string, too. | |||
| 1258 | * | |||
| 1259 | * @return Newly allocated conversion result, or NULL upon an error. | |||
| 1260 | */ | |||
| 1261 | char *url_as_string(su_home_t *home, url_t const *url) | |||
| 1262 | { | |||
| 1263 | if (url) { | |||
| 1264 | int len = url_e(NULL((void*)0), 0, url); | |||
| 1265 | char *b = su_alloc(home, len + 1); | |||
| 1266 | url_e(b, len + 1, url); | |||
| 1267 | return b; | |||
| 1268 | } else { | |||
| 1269 | return NULL((void*)0); | |||
| 1270 | } | |||
| 1271 | } | |||
| 1272 | ||||
| 1273 | ||||
| 1274 | /** Test if param @a tag matches to parameter string @a p. | |||
| 1275 | */ | |||
| 1276 | #define URL_PARAM_MATCH(p, tag)(strncasecmp(p, tag, strlen(tag)) == 0 && (p[strlen(tag )] == '\0' || p[strlen(tag)] == ';' || p[strlen(tag)] == '=') ) \ | |||
| 1277 | (strncasecmp(p, tag, strlen(tag)) == 0 && \ | |||
| 1278 | (p[strlen(tag)] == '\0' || p[strlen(tag)] == ';' || p[strlen(tag)] == '=')) | |||
| 1279 | ||||
| 1280 | /** | |||
| 1281 | * Search for a parameter. | |||
| 1282 | * | |||
| 1283 | * This function searches for a parameter from a parameter list. | |||
| 1284 | * | |||
| 1285 | * If you want to test if there is parameter @b user=phone, | |||
| 1286 | * call this function like | |||
| 1287 | * @code if (url_param(url->url_param, "user=phone", NULL, 0)) | |||
| 1288 | * @endcode | |||
| 1289 | * | |||
| 1290 | * @param params URL parameter string (excluding first semicolon) | |||
| 1291 | * @param tag parameter name | |||
| 1292 | * @param value string to which the parameter value is copied | |||
| 1293 | * @param vlen length of string reserved for value | |||
| 1294 | * | |||
| 1295 | * @retval positive length of parameter value (including final NUL) if found | |||
| 1296 | * @retval zero if not found. | |||
| 1297 | */ | |||
| 1298 | isize_t url_param(char const *params, | |||
| 1299 | char const *tag, | |||
| 1300 | char value[], isize_t vlen) | |||
| 1301 | { | |||
| 1302 | size_t n, tlen, flen; | |||
| 1303 | char *p; | |||
| 1304 | ||||
| 1305 | if (!params) | |||
| 1306 | return 0; | |||
| 1307 | ||||
| 1308 | tlen = strlen(tag); | |||
| 1309 | if (tlen && tag[tlen - 1] == '=') | |||
| 1310 | tlen--; | |||
| 1311 | ||||
| 1312 | for (p = (char *)params; *p; p += n + 1) { | |||
| 1313 | n = strcspn(p, ";"); | |||
| 1314 | if (n < tlen) { | |||
| 1315 | if (p[n]) continue; else break; | |||
| 1316 | } | |||
| 1317 | if (strncasecmp(p, tag, tlen) == 0) { | |||
| 1318 | if (n == tlen) { | |||
| 1319 | if (vlen > 0) | |||
| 1320 | value[0] = '\0'; | |||
| 1321 | return 1; | |||
| 1322 | } | |||
| 1323 | if (p[tlen] != '=') | |||
| 1324 | continue; | |||
| 1325 | flen = n - tlen - 1; | |||
| 1326 | if (flen >= (size_t)vlen) | |||
| 1327 | return flen + 1; | |||
| 1328 | memcpy(value, p + tlen + 1, flen); | |||
| 1329 | value[flen] = '\0'; | |||
| 1330 | return flen + 1; | |||
| 1331 | } | |||
| 1332 | if (!p[n]) | |||
| 1333 | break; | |||
| 1334 | } | |||
| 1335 | ||||
| 1336 | return 0; | |||
| 1337 | } | |||
| 1338 | ||||
| 1339 | /** Check for a parameter. | |||
| 1340 | * | |||
| 1341 | * @deprecated | |||
| 1342 | * Bad grammar. Use url_has_param(). | |||
| 1343 | */ | |||
| 1344 | isize_t url_have_param(char const *params, char const *tag) | |||
| 1345 | { | |||
| 1346 | return url_param(params, tag, NULL((void*)0), 0); | |||
| 1347 | } | |||
| 1348 | ||||
| 1349 | /** Check for a parameter. */ | |||
| 1350 | int url_has_param(url_t const *url, char const *tag) | |||
| 1351 | { | |||
| 1352 | return url && url->url_params && url_param(url->url_params, tag, NULL((void*)0), 0); | |||
| 1353 | } | |||
| 1354 | ||||
| 1355 | /** Add an parameter. */ | |||
| 1356 | int url_param_add(su_home_t *h, url_t *url, char const *param) | |||
| 1357 | { | |||
| 1358 | /* XXX - should remove existing parameters with same name? */ | |||
| 1359 | size_t n = url->url_params ? strlen(url->url_params) + 1: 0; | |||
| 1360 | size_t nn = strlen(param) + 1; | |||
| 1361 | char *s = su_alloc(h, n + nn); | |||
| 1362 | ||||
| 1363 | if (!s) | |||
| 1364 | return -1; | |||
| 1365 | ||||
| 1366 | if (url->url_params) | |||
| 1367 | strcpy(s, url->url_params)[n - 1] = ';'; | |||
| 1368 | strcpy(s + n, param); | |||
| 1369 | url->url_params = s; | |||
| 1370 | ||||
| 1371 | return 0; | |||
| 1372 | } | |||
| 1373 | ||||
| 1374 | /** Remove a named parameter from url_param string. | |||
| 1375 | * | |||
| 1376 | * Remove a named parameter and its possible value from the URL parameter | |||
| 1377 | * string (url_s##url_param). | |||
| 1378 | * | |||
| 1379 | * @return Pointer to modified string, or NULL if nothing is left in there. | |||
| 1380 | */ | |||
| 1381 | char *url_strip_param_string(char *params, char const *name) | |||
| 1382 | { | |||
| 1383 | if (params && name) { | |||
| 1384 | size_t i, n = strlen(name), remove, rest; | |||
| 1385 | ||||
| 1386 | for (i = 0; params[i];) { | |||
| 1387 | if (strncasecmp(params + i, name, n) || | |||
| 1388 | (params[i + n] != '=' && params[i + n] != ';' && params[i + n])) { | |||
| 1389 | i = i + strcspn(params + i, ";"); | |||
| 1390 | if (!params[i++]) | |||
| 1391 | break; | |||
| 1392 | continue; | |||
| 1393 | } | |||
| 1394 | remove = n + strcspn(params + i + n, ";"); | |||
| 1395 | if (params[i + remove] == ';') | |||
| 1396 | remove++; | |||
| 1397 | ||||
| 1398 | if (i == 0) { | |||
| 1399 | params += remove; | |||
| 1400 | continue; | |||
| 1401 | } | |||
| 1402 | ||||
| 1403 | rest = strlen(params + i + remove); | |||
| 1404 | if (!rest) { | |||
| 1405 | if (i == 0) | |||
| 1406 | return NULL((void*)0); /* removed everything */ | |||
| 1407 | params[i - 1] = '\0'; | |||
| 1408 | break; | |||
| 1409 | } | |||
| 1410 | memmove(params + i, params + i + remove, rest + 1); | |||
| 1411 | } | |||
| 1412 | ||||
| 1413 | if (!params[0]) | |||
| 1414 | return NULL((void*)0); | |||
| 1415 | } | |||
| 1416 | ||||
| 1417 | return params; | |||
| 1418 | } | |||
| 1419 | ||||
| 1420 | int url_string_p(url_string_t const *url) | |||
| 1421 | { | |||
| 1422 | return URL_STRING_P(url)((url) && *((url_string_t*)(url))->us_str != 0); | |||
| 1423 | } | |||
| 1424 | ||||
| 1425 | int url_is_string(url_string_t const *url) | |||
| 1426 | { | |||
| 1427 | return URL_IS_STRING(url)((url) && *((url_string_t*)(url))->us_str != 0); | |||
| 1428 | } | |||
| 1429 | ||||
| 1430 | /** Strip transport-specific stuff. */ | |||
| 1431 | static | |||
| 1432 | int url_strip_transport2(url_t *url, int modify) | |||
| 1433 | { | |||
| 1434 | char *p, *d; | |||
| 1435 | size_t n; | |||
| 1436 | int semi; | |||
| 1437 | ||||
| 1438 | if (url->url_type != url_sip && url->url_type != url_sips) | |||
| 1439 | return 0; | |||
| 1440 | ||||
| 1441 | if (url->url_port != NULL((void*)0)) { | |||
| 1442 | if (!modify) | |||
| 1443 | return 1; | |||
| 1444 | url->url_port = NULL((void*)0); | |||
| 1445 | } | |||
| 1446 | ||||
| 1447 | if (!url->url_params) | |||
| 1448 | return 0; | |||
| 1449 | ||||
| 1450 | for (d = p = (char *)url->url_params; *p; p += n + semi) { | |||
| 1451 | n = strcspn(p, ";"); | |||
| 1452 | semi = (p[n] != '\0'); | |||
| 1453 | ||||
| 1454 | if (modify && n == 0) | |||
| 1455 | continue; | |||
| 1456 | if (URL_PARAM_MATCH(p, "method")(strncasecmp(p, "method", strlen("method")) == 0 && ( p[strlen("method")] == '\0' || p[strlen("method")] == ';' || p [strlen("method")] == '='))) | |||
| 1457 | continue; | |||
| 1458 | if (URL_PARAM_MATCH(p, "maddr")(strncasecmp(p, "maddr", strlen("maddr")) == 0 && (p[ strlen("maddr")] == '\0' || p[strlen("maddr")] == ';' || p[strlen ("maddr")] == '='))) | |||
| 1459 | continue; | |||
| 1460 | if (URL_PARAM_MATCH(p, "ttl")(strncasecmp(p, "ttl", strlen("ttl")) == 0 && (p[strlen ("ttl")] == '\0' || p[strlen("ttl")] == ';' || p[strlen("ttl" )] == '='))) | |||
| 1461 | continue; | |||
| 1462 | if (URL_PARAM_MATCH(p, "transport")(strncasecmp(p, "transport", strlen("transport")) == 0 && (p[strlen("transport")] == '\0' || p[strlen("transport")] == ';' || p[strlen("transport")] == '='))) | |||
| 1463 | continue; | |||
| 1464 | ||||
| 1465 | if (p != d) { | |||
| 1466 | if (d != url->url_params) | |||
| 1467 | d++; | |||
| 1468 | if (p != d) { | |||
| 1469 | if (!modify) | |||
| 1470 | return 1; | |||
| 1471 | memmove(d, p, n + 1); | |||
| 1472 | } | |||
| 1473 | } | |||
| 1474 | d += n; | |||
| 1475 | } | |||
| 1476 | ||||
| 1477 | if (d == p) | |||
| 1478 | return 0; | |||
| 1479 | else if (d + 1 == p) /* empty param */ | |||
| 1480 | return 0; | |||
| 1481 | else if (!modify) | |||
| 1482 | return 1; | |||
| 1483 | ||||
| 1484 | if (d != url->url_params) | |||
| 1485 | *d = '\0'; | |||
| 1486 | else | |||
| 1487 | url->url_params = NULL((void*)0); | |||
| 1488 | ||||
| 1489 | return 1; | |||
| 1490 | } | |||
| 1491 | ||||
| 1492 | /** Strip transport-specific stuff. | |||
| 1493 | * | |||
| 1494 | * The function url_strip_transport() removes transport-specific parameters | |||
| 1495 | * from a SIP or SIPS URI. These parameters include: | |||
| 1496 | * - the port number | |||
| 1497 | * - "maddr=" parameter | |||
| 1498 | * - "transport=" parameter | |||
| 1499 | * - "ttl=" parameter | |||
| 1500 | * - "method=" parameter | |||
| 1501 | * | |||
| 1502 | * @note | |||
| 1503 | * The @a url must be a pointer to a URL structure. It is stripped in-place. | |||
| 1504 | * | |||
| 1505 | * @note | |||
| 1506 | * If the parameter string contains empty parameters, they are stripped, too. | |||
| 1507 | * | |||
| 1508 | * @return | |||
| 1509 | * The function url_strip_transport() returns @e true, if the URL was | |||
| 1510 | * modified, @e false otherwise. | |||
| 1511 | */ | |||
| 1512 | int url_strip_transport(url_t *url) | |||
| 1513 | { | |||
| 1514 | return url_strip_transport2(url, 1); | |||
| 1515 | } | |||
| 1516 | ||||
| 1517 | /** Check for transport-specific stuff. | |||
| 1518 | * | |||
| 1519 | * The function url_have_transport() tests if there are transport-specific | |||
| 1520 | * parameters in a SIP or SIPS URI. These parameters include: | |||
| 1521 | * - the port number | |||
| 1522 | * - "maddr=" parameters | |||
| 1523 | * - "transport=" parameters | |||
| 1524 | * | |||
| 1525 | * @note | |||
| 1526 | * The @a url must be a pointer to a URL structure. | |||
| 1527 | * | |||
| 1528 | * @return The function url_have_transport() returns @e true, if the URL | |||
| 1529 | * contains transport parameters, @e false otherwise. | |||
| 1530 | */ | |||
| 1531 | int url_have_transport(url_t const *url) | |||
| 1532 | { | |||
| 1533 | return url_strip_transport2((url_t *)url, 0); | |||
| 1534 | } | |||
| 1535 | ||||
| 1536 | /**Lazily compare two URLs. | |||
| 1537 | * | |||
| 1538 | * Compare essential parts of URLs: schema, host, port, and username. | |||
| 1539 | * | |||
| 1540 | * any_url compares 0 with any other URL. | |||
| 1541 | * | |||
| 1542 | * pres: and im: URIs compares 0 with SIP URIs. | |||
| 1543 | * | |||
| 1544 | * @note | |||
| 1545 | * The @a a and @a b must be pointers to URL structures. | |||
| 1546 | * | |||
| 1547 | * @note Currently, the url parameters are not compared. This is because the | |||
| 1548 | * url_cmp() is used to sort URLs: taking parameters into account makes that | |||
| 1549 | * impossible. | |||
| 1550 | */ | |||
| 1551 | int url_cmp(url_t const *a, url_t const *b) | |||
| 1552 | { | |||
| 1553 | int rv; | |||
| 1554 | int url_type; | |||
| 1555 | ||||
| 1556 | if ((a && a->url_type == url_any) || (b && b->url_type == url_any)) | |||
| 1557 | return 0; | |||
| 1558 | ||||
| 1559 | if (!a || !b) | |||
| 1560 | return (a != NULL((void*)0)) - (b != NULL((void*)0)); | |||
| 1561 | ||||
| 1562 | if ((rv = a->url_type - b->url_type)) { | |||
| 1563 | #if 0 | |||
| 1564 | /* presence and instant messaging URLs match magically with SIP */ | |||
| 1565 | enum url_type_e a_type = a->url_type; | |||
| 1566 | enum url_type_e b_type = b->url_type; | |||
| 1567 | ||||
| 1568 | if (a_type == url_im || a_type == url_pres) | |||
| 1569 | a_type = url_sip; | |||
| 1570 | ||||
| 1571 | if (b_type == url_im || b_type == url_pres) | |||
| 1572 | b_type = url_sip; | |||
| 1573 | ||||
| 1574 | if (a_type != b_type) | |||
| 1575 | #endif | |||
| 1576 | return rv; | |||
| 1577 | } | |||
| 1578 | ||||
| 1579 | url_type = a->url_type; /* Or b->url_type, they are equal! */ | |||
| 1580 | ||||
| 1581 | if (url_type <= url_unknown && | |||
| 1582 | ((rv = !a->url_scheme - !b->url_scheme) || | |||
| 1583 | (a->url_scheme && b->url_scheme && | |||
| 1584 | (rv = strcasecmp(a->url_scheme, b->url_scheme))))) | |||
| 1585 | return rv; | |||
| 1586 | ||||
| 1587 | if ((rv = host_cmp(a->url_host, b->url_host))) | |||
| 1588 | return rv; | |||
| 1589 | ||||
| 1590 | if (a->url_port != b->url_port) { | |||
| 1591 | char const *a_port; | |||
| 1592 | char const *b_port; | |||
| 1593 | ||||
| 1594 | if (url_type != url_sip && url_type != url_sips) | |||
| 1595 | a_port = b_port = url_port_default((enum url_type_e)url_type); | |||
| 1596 | else if (host_is_ip_address(a->url_host)) | |||
| 1597 | a_port = b_port = url_port_default((enum url_type_e)url_type); | |||
| 1598 | else | |||
| 1599 | a_port = b_port = ""; | |||
| 1600 | ||||
| 1601 | if (a->url_port) a_port = a->url_port; | |||
| 1602 | if (b->url_port) b_port = b->url_port; | |||
| 1603 | ||||
| 1604 | if ((rv = strcmp(a_port, b_port))) | |||
| 1605 | return rv; | |||
| 1606 | } | |||
| 1607 | ||||
| 1608 | if (a->url_user != b->url_user) { | |||
| 1609 | if (a->url_user == NULL((void*)0)) return -1; | |||
| 1610 | if (b->url_user == NULL((void*)0)) return +1; | |||
| 1611 | switch (url_type) { | |||
| 1612 | case url_tel: case url_modem: case url_fax: | |||
| 1613 | rv = url_tel_cmp_numbers(a->url_user, b->url_user); | |||
| 1614 | break; | |||
| 1615 | default: | |||
| 1616 | rv = strcmp(a->url_user, b->url_user); | |||
| 1617 | break; | |||
| 1618 | } | |||
| 1619 | if (rv) | |||
| 1620 | return rv; | |||
| 1621 | } | |||
| 1622 | ||||
| 1623 | #if 0 | |||
| 1624 | if (a->url_path != b->url_path) { | |||
| 1625 | if (a->url_path == NULL((void*)0)) return -1; | |||
| 1626 | if (b->url_path == NULL((void*)0)) return +1; | |||
| 1627 | if ((rv = strcmp(a->url_path, b->url_path))) | |||
| 1628 | return rv; | |||
| 1629 | } | |||
| 1630 | #endif | |||
| 1631 | ||||
| 1632 | return 0; | |||
| 1633 | } | |||
| 1634 | ||||
| 1635 | static | |||
| 1636 | int url_tel_cmp_numbers(char const *A, char const *B) | |||
| 1637 | { | |||
| 1638 | short a, b; | |||
| 1639 | int rv; | |||
| 1640 | ||||
| 1641 | while (*A && *B) { | |||
| 1642 | #define UNHEX(a) (a - (a >= 'a' ? 'a' - 10 : (a >= 'A' ? 'A' - 10 : '0'))) | |||
| 1643 | /* Skip visual-separators */ | |||
| 1644 | do { | |||
| 1645 | a = *A++; | |||
| 1646 | if (a == '%' && IS_HEX(A[0])(((A[0]) >= '0' && (A[0]) <= '9') || ((A[0]) >= 'A' && (A[0]) <= 'F') || ((A[0]) >= 'a' && (A[0]) <= 'f')) && IS_HEX(A[1])(((A[1]) >= '0' && (A[1]) <= '9') || ((A[1]) >= 'A' && (A[1]) <= 'F') || ((A[1]) >= 'a' && (A[1]) <= 'f'))) | |||
| 1647 | a = (UNHEX(A[0]) << 4) | UNHEX(A[1]), A +=2; | |||
| 1648 | } while (a == ' ' || a == '-' || a == '.' || a == '(' || a == ')'); | |||
| 1649 | ||||
| 1650 | if (isupper(a)((*__ctype_b_loc ())[(int) ((a))] & (unsigned short int) _ISupper )) | |||
| 1651 | a = tolower(a); | |||
| 1652 | ||||
| 1653 | do { | |||
| 1654 | b = *B++; | |||
| 1655 | if (b == '%' && IS_HEX(B[0])(((B[0]) >= '0' && (B[0]) <= '9') || ((B[0]) >= 'A' && (B[0]) <= 'F') || ((B[0]) >= 'a' && (B[0]) <= 'f')) && IS_HEX(B[1])(((B[1]) >= '0' && (B[1]) <= '9') || ((B[1]) >= 'A' && (B[1]) <= 'F') || ((B[1]) >= 'a' && (B[1]) <= 'f'))) | |||
| 1656 | b = (UNHEX(B[0]) << 4) | UNHEX(B[1]), B +=2; | |||
| 1657 | } while (b == ' ' || b == '-' || b == '.' || b == '(' || b == ')'); | |||
| 1658 | ||||
| 1659 | if (isupper(b)((*__ctype_b_loc ())[(int) ((b))] & (unsigned short int) _ISupper )) | |||
| 1660 | b = tolower(b); | |||
| 1661 | ||||
| 1662 | if ((rv = a - b)) | |||
| 1663 | return rv; | |||
| 1664 | } | |||
| 1665 | ||||
| 1666 | return (int)*A - (int)*B; | |||
| 1667 | } | |||
| 1668 | ||||
| 1669 | /**Conservative comparison of urls. | |||
| 1670 | * | |||
| 1671 | * Compare all parts of URLs. | |||
| 1672 | * | |||
| 1673 | * @note | |||
| 1674 | * The @a a and @a b must be pointers to URL structures. | |||
| 1675 | * | |||
| 1676 | */ | |||
| 1677 | int url_cmp_all(url_t const *a, url_t const *b) | |||
| 1678 | { | |||
| 1679 | int rv, url_type; | |||
| 1680 | ||||
| 1681 | if (!a || !b) | |||
| 1682 | return (a != NULL((void*)0)) - (b != NULL((void*)0)); | |||
| 1683 | ||||
| 1684 | if ((rv = a->url_type - b->url_type)) | |||
| 1685 | return rv; | |||
| 1686 | ||||
| 1687 | url_type = a->url_type; /* Or b->url_type, they are equal! */ | |||
| 1688 | ||||
| 1689 | if (url_type <= url_unknown && | |||
| 1690 | ((rv = !a->url_scheme - !b->url_scheme) || | |||
| 1691 | (a->url_scheme && b->url_scheme && | |||
| 1692 | (rv = strcasecmp(a->url_scheme, b->url_scheme))))) | |||
| 1693 | return rv; | |||
| 1694 | ||||
| 1695 | if ((rv = a->url_root - b->url_root)) | |||
| 1696 | return rv; | |||
| 1697 | ||||
| 1698 | if ((rv = host_cmp(a->url_host, b->url_host))) | |||
| 1699 | return rv; | |||
| 1700 | ||||
| 1701 | if (a->url_port != b->url_port) { | |||
| 1702 | char const *a_port; | |||
| 1703 | char const *b_port; | |||
| 1704 | ||||
| 1705 | if (url_type != url_sip && url_type != url_sips) | |||
| 1706 | a_port = b_port = url_port_default((enum url_type_e)url_type); | |||
| 1707 | else if (host_is_ip_address(a->url_host)) | |||
| 1708 | a_port = b_port = url_port_default((enum url_type_e)url_type); | |||
| 1709 | else | |||
| 1710 | a_port = b_port = ""; | |||
| 1711 | ||||
| 1712 | if (a->url_port) a_port = a->url_port; | |||
| 1713 | if (b->url_port) b_port = b->url_port; | |||
| 1714 | ||||
| 1715 | if ((rv = strcmp(a_port, b_port))) | |||
| 1716 | return rv; | |||
| 1717 | } | |||
| 1718 | ||||
| 1719 | if (a->url_user != b->url_user) { | |||
| 1720 | if (a->url_user == NULL((void*)0)) return -1; | |||
| 1721 | if (b->url_user == NULL((void*)0)) return +1; | |||
| 1722 | ||||
| 1723 | switch (url_type) { | |||
| 1724 | case url_tel: case url_modem: case url_fax: | |||
| 1725 | rv = url_tel_cmp_numbers(a->url_user, b->url_user); | |||
| 1726 | break; | |||
| 1727 | default: | |||
| 1728 | rv = strcmp(a->url_user, b->url_user); | |||
| 1729 | break; | |||
| 1730 | } | |||
| 1731 | if (rv) | |||
| 1732 | return rv; | |||
| 1733 | } | |||
| 1734 | ||||
| 1735 | if (a->url_path != b->url_path) { | |||
| 1736 | if (a->url_path == NULL((void*)0)) return -1; | |||
| 1737 | if (b->url_path == NULL((void*)0)) return +1; | |||
| 1738 | if ((rv = strcmp(a->url_path, b->url_path))) | |||
| 1739 | return rv; | |||
| 1740 | } | |||
| 1741 | ||||
| 1742 | if (a->url_params != b->url_params) { | |||
| 1743 | if (a->url_params == NULL((void*)0)) return -1; | |||
| 1744 | if (b->url_params == NULL((void*)0)) return +1; | |||
| 1745 | if ((rv = strcmp(a->url_params, b->url_params))) | |||
| 1746 | return rv; | |||
| 1747 | } | |||
| 1748 | ||||
| 1749 | if (a->url_headers != b->url_headers) { | |||
| 1750 | if (a->url_headers == NULL((void*)0)) return -1; | |||
| 1751 | if (b->url_headers == NULL((void*)0)) return +1; | |||
| 1752 | if ((rv = strcmp(a->url_headers, b->url_headers))) | |||
| 1753 | return rv; | |||
| 1754 | } | |||
| 1755 | ||||
| 1756 | if (a->url_headers != b->url_headers) { | |||
| 1757 | if (a->url_headers == NULL((void*)0)) return -1; | |||
| 1758 | if (b->url_headers == NULL((void*)0)) return +1; | |||
| 1759 | if ((rv = strcmp(a->url_headers, b->url_headers))) | |||
| 1760 | return rv; | |||
| 1761 | } | |||
| 1762 | ||||
| 1763 | if (a->url_fragment != b->url_fragment) { | |||
| 1764 | if (a->url_fragment == NULL((void*)0)) return -1; | |||
| 1765 | if (b->url_fragment == NULL((void*)0)) return +1; | |||
| 1766 | if ((rv = strcmp(a->url_fragment, b->url_fragment))) | |||
| 1767 | return rv; | |||
| 1768 | } | |||
| 1769 | ||||
| 1770 | return 0; | |||
| 1771 | } | |||
| 1772 | ||||
| 1773 | /** Return default port number corresponding to the url type */ | |||
| 1774 | char const *url_port_default(enum url_type_e url_type) | |||
| 1775 | { | |||
| 1776 | switch (url_type) { | |||
| 1777 | case url_sip: /* "sip:" */ | |||
| 1778 | return "5060"; | |||
| 1779 | case url_sips: /* "sips:" */ | |||
| 1780 | return "5061"; | |||
| 1781 | case url_http: /* "http:" */ | |||
| 1782 | return "80"; | |||
| 1783 | case url_https: /* "https:" */ | |||
| 1784 | return "443"; | |||
| 1785 | case url_ftp: /* "ftp:" */ | |||
| 1786 | case url_file: /* "file:" */ | |||
| 1787 | return "21"; | |||
| 1788 | case url_rtsp: /* "rtsp:" */ | |||
| 1789 | case url_rtspu: /* "rtspu:" */ | |||
| 1790 | return "554"; | |||
| 1791 | case url_mailto: /* "mailto:" */ | |||
| 1792 | return "25"; | |||
| 1793 | ||||
| 1794 | case url_any: /* "*" */ | |||
| 1795 | return "*"; | |||
| 1796 | ||||
| 1797 | case url_msrp: | |||
| 1798 | case url_msrps: | |||
| 1799 | return "9999"; /* XXXX */ | |||
| 1800 | ||||
| 1801 | case url_tel: | |||
| 1802 | case url_urn: | |||
| 1803 | case url_fax: | |||
| 1804 | case url_modem: | |||
| 1805 | case url_im: | |||
| 1806 | case url_pres: | |||
| 1807 | case url_cid: | |||
| 1808 | case url_wv: | |||
| 1809 | ||||
| 1810 | default: /* Unknown scheme */ | |||
| 1811 | return ""; | |||
| 1812 | } | |||
| 1813 | } | |||
| 1814 | ||||
| 1815 | /** Return default transport name corresponding to the url type */ | |||
| 1816 | char const *url_tport_default(enum url_type_e url_type) | |||
| 1817 | { | |||
| 1818 | switch (url_type) { | |||
| 1819 | case url_sip: | |||
| 1820 | return "*"; | |||
| 1821 | case url_sips: | |||
| 1822 | return "tls"; | |||
| 1823 | case url_http: | |||
| 1824 | return "tcp"; | |||
| 1825 | case url_https: | |||
| 1826 | return "tls"; | |||
| 1827 | case url_ftp: | |||
| 1828 | case url_file: | |||
| 1829 | return "tcp"; | |||
| 1830 | case url_rtsp: | |||
| 1831 | return "tcp"; | |||
| 1832 | case url_rtspu: | |||
| 1833 | return "udp"; | |||
| 1834 | case url_mailto: | |||
| 1835 | return "tcp"; | |||
| 1836 | case url_msrp: | |||
| 1837 | return "tcp"; | |||
| 1838 | case url_msrps: | |||
| 1839 | return "tls"; | |||
| 1840 | ||||
| 1841 | case url_any: /* "*" */ | |||
| 1842 | case url_tel: | |||
| 1843 | case url_fax: | |||
| 1844 | case url_modem: | |||
| 1845 | case url_im: | |||
| 1846 | case url_pres: | |||
| 1847 | case url_cid: | |||
| 1848 | case url_urn: | |||
| 1849 | case url_wv: | |||
| 1850 | ||||
| 1851 | default: /* Unknown scheme */ | |||
| 1852 | return "*"; | |||
| 1853 | } | |||
| 1854 | } | |||
| 1855 | ||||
| 1856 | ||||
| 1857 | /** Return the URL port string */ | |||
| 1858 | char const *url_port(url_t const *u) | |||
| 1859 | { | |||
| 1860 | if (!u) | |||
| 1861 | return ""; | |||
| 1862 | else if (u->url_port && u->url_port[0]) | |||
| 1863 | return u->url_port; | |||
| 1864 | ||||
| 1865 | if (u->url_type == url_sips || u->url_type == url_sip) | |||
| 1866 | if (!host_is_ip_address(u->url_host)) | |||
| 1867 | return ""; | |||
| 1868 | ||||
| 1869 | return url_port_default((enum url_type_e)u->url_type); | |||
| 1870 | } | |||
| 1871 | ||||
| 1872 | /** Sanitize URL. | |||
| 1873 | * | |||
| 1874 | * The function url_sanitize() adds a scheme to an incomplete URL. It | |||
| 1875 | * modifies its parameter structure @a url. Currently, the function follows | |||
| 1876 | * simple heuristics: | |||
| 1877 | * | |||
| 1878 | * - URL with host name starting with @c ftp. is an FTP URL | |||
| 1879 | * - URL with host name starting with @c www. is an HTTP URL | |||
| 1880 | * - URL with host and path, e.g., @c host/foo;bar, is an HTTP URL | |||
| 1881 | * - URL with host name, no path is a SIP URL. | |||
| 1882 | * | |||
| 1883 | * @param url pointer to URL struct to be sanitized (IN/OUT) | |||
| 1884 | * | |||
| 1885 | * @return | |||
| 1886 | * The function url_sanitize() returns 0 if it considers URL to be | |||
| 1887 | * sane, and -1 otherwise. | |||
| 1888 | */ | |||
| 1889 | int url_sanitize(url_t *url) | |||
| 1890 | { | |||
| 1891 | if (!url) | |||
| 1892 | return -1; | |||
| 1893 | else if (url->url_scheme != NULL((void*)0)) | |||
| 1894 | /* xyzzy */; | |||
| 1895 | else if (url->url_host == NULL((void*)0)) | |||
| 1896 | return -1; | |||
| 1897 | else if (strncasecmp(url->url_host, "ftp.", strlen("ftp.")) == 0) | |||
| 1898 | url->url_type = url_ftp, url->url_scheme = "ftp", url->url_root = '/'; | |||
| 1899 | else if (strncasecmp(url->url_host, "www.", strlen("www.")) == 0 | |||
| 1900 | || url->url_path) | |||
| 1901 | url->url_type = url_http, url->url_scheme = "http", url->url_root = '/'; | |||
| 1902 | else | |||
| 1903 | url->url_type = url_sip, url->url_scheme = "sip"; | |||
| 1904 | ||||
| 1905 | return 0; | |||
| 1906 | } | |||
| 1907 | ||||
| 1908 | #include <sofia-sip/su_md5.h> | |||
| 1909 | ||||
| 1910 | static | |||
| 1911 | void canon_update(su_md5_t *md5, char const *s, size_t n, char const *allow) | |||
| 1912 | { | |||
| 1913 | size_t i, j; | |||
| 1914 | ||||
| 1915 | for (i = 0, j = 0; i < n && s[i]; i++) { | |||
| 1916 | char c; | |||
| 1917 | ||||
| 1918 | if (s[i] == '%' && i + 2 < n && IS_HEX(s[i+1])(((s[i+1]) >= '0' && (s[i+1]) <= '9') || ((s[i+ 1]) >= 'A' && (s[i+1]) <= 'F') || ((s[i+1]) >= 'a' && (s[i+1]) <= 'f')) && IS_HEX(s[i+2])(((s[i+2]) >= '0' && (s[i+2]) <= '9') || ((s[i+ 2]) >= 'A' && (s[i+2]) <= 'F') || ((s[i+2]) >= 'a' && (s[i+2]) <= 'f'))) { | |||
| 1919 | #define UNHEX(a) (a - (a >= 'a' ? 'a' - 10 : (a >= 'A' ? 'A' - 10 : '0'))) | |||
| 1920 | c = (UNHEX(s[i+1]) << 4) | UNHEX(s[i+2]); | |||
| 1921 | #undef UNHEX | |||
| 1922 | if (c != '%' && c > ' ' && c < '\177' && | |||
| 1923 | (!strchr(EXCLUDED";/?:@&=+$," "<>#%\"" "{}|\\^[]`", c) || strchr(allow, c))) { | |||
| 1924 | if (i != j) | |||
| 1925 | su_md5_iupdate(md5, s + j, i - j); | |||
| 1926 | su_md5_iupdate(md5, &c, 1); | |||
| 1927 | j = i + 3; | |||
| 1928 | } | |||
| 1929 | i += 2; | |||
| 1930 | } | |||
| 1931 | } | |||
| 1932 | ||||
| 1933 | if (i != j) | |||
| 1934 | su_md5_iupdate(md5, s + j, i - j); | |||
| 1935 | } | |||
| 1936 | ||||
| 1937 | /** Update MD5 sum with url-string contents */ | |||
| 1938 | static | |||
| 1939 | void url_string_update(su_md5_t *md5, char const *s) | |||
| 1940 | { | |||
| 1941 | size_t n, p; | |||
| 1942 | int have_authority = 1; | |||
| 1943 | enum url_type_e type = url_any; | |||
| 1944 | char const *at, *colon; | |||
| 1945 | char schema[48]; | |||
| 1946 | ||||
| 1947 | if (s == NULL((void*)0) || strlen(s) == 0 || strcmp(s, "*") == 0) { | |||
| 1948 | su_md5_update(md5, "*\0\0*", 4); | |||
| 1949 | return; | |||
| 1950 | } | |||
| 1951 | ||||
| 1952 | n = strcspn(s, ":/?#"); | |||
| 1953 | if (n >= sizeof schema) { | |||
| 1954 | su_md5_update(md5, ":", 1); | |||
| 1955 | } | |||
| 1956 | else if (n && s[n] == ':' ) { | |||
| 1957 | at = url_canonize(schema, s, n, 0, "+"); | |||
| 1958 | ||||
| 1959 | type = url_get_type(schema, at - schema); | |||
| 1960 | su_md5_iupdate(md5, schema, at - schema); | |||
| 1961 | ||||
| 1962 | have_authority = !url_type_is_opaque(type); | |||
| 1963 | s += n + 1; | |||
| 1964 | } | |||
| 1965 | else { | |||
| 1966 | su_md5_update(md5, "", 1); | |||
| 1967 | } | |||
| 1968 | ||||
| 1969 | if (type == url_sip || type == url_sips) { | |||
| 1970 | /* SIP URL may have /;? in user part but no path */ | |||
| 1971 | /* user-unreserved = "&" / "=" / "+" / "$" / "," / ";" / "?" / "/" */ | |||
| 1972 | /* Some #*@#* phones include unescaped # there, too */ | |||
| 1973 | n = strcspn(s, "@/;?#"); | |||
| 1974 | p = strcspn(s + n, "@"); | |||
| 1975 | if (s[n + p] == '@') { | |||
| 1976 | n += p; | |||
| 1977 | /* Ignore password in hash */ | |||
| 1978 | colon = memchr(s, ':', n); | |||
| 1979 | p = colon ? (size_t)(colon - s) : n; | |||
| 1980 | canon_update(md5, s, p, SIP_USER_UNRESERVED"&=+$,;?/"); | |||
| 1981 | s += n + 1; n = 0; | |||
| 1982 | } | |||
| 1983 | else | |||
| 1984 | su_md5_iupdate(md5, "", 1); /* user */ | |||
| 1985 | n += strcspn(s + n, "/;?#"); | |||
| 1986 | } | |||
| 1987 | else if (have_authority) { | |||
| 1988 | if (type == url_wv) { /* WV URL may have / in user part */ | |||
| 1989 | n = strcspn(s, "@;?#"); | |||
| 1990 | } | |||
| 1991 | else if (type != url_wv && s[0] == '/' && s[1] != '/') { | |||
| 1992 | /* foo:/bar */ | |||
| 1993 | su_md5_update(md5, "\0\0", 2); /* user, host */ | |||
| 1994 | su_md5_striupdate(md5, url_port_default(type)); | |||
| 1995 | return; | |||
| 1996 | } | |||
| 1997 | else if (s[0] == '/' && s[1] == '/') { | |||
| 1998 | /* We have authority, / / foo or foo */ | |||
| 1999 | s += 2; | |||
| 2000 | n = strcspn(s, "/?#@[]"); | |||
| 2001 | } | |||
| 2002 | else | |||
| 2003 | n = strcspn(s, "@;/?#"); | |||
| 2004 | ||||
| 2005 | if (s[n] == '@') { | |||
| 2006 | /* Ignore password in hash */ | |||
| 2007 | colon = type != url_unknown ? memchr(s, ':', n) : NULL((void*)0); | |||
| 2008 | p = colon ? (size_t)(colon - s) : n; | |||
| 2009 | canon_update(md5, s, p, SIP_USER_UNRESERVED"&=+$,;?/"); | |||
| 2010 | s += n + 1; | |||
| 2011 | n = strcspn(s, "/;?#"); /* Until path, query or fragment */ | |||
| 2012 | } | |||
| 2013 | else { | |||
| 2014 | su_md5_iupdate(md5, "", 1); /* user */ | |||
| 2015 | n += strcspn(s + n, "/;?#"); /* Until path, query or fragment */ | |||
| 2016 | } | |||
| 2017 | } | |||
| 2018 | else /* if (!have_authority) */ { | |||
| 2019 | n = strcspn(s, ":/;?#"); /* Until pass, path, query or fragment */ | |||
| 2020 | ||||
| 2021 | canon_update(md5, s, n, ""); /* user */ | |||
| 2022 | su_md5_update(md5, "\0", 1); /* host, no port */ | |||
| 2023 | su_md5_striupdate(md5, url_port_default(type)); | |||
| 2024 | return; | |||
| 2025 | } | |||
| 2026 | ||||
| 2027 | if (n > 0 && s[0] == '[') { /* IPv6reference */ | |||
| 2028 | colon = memchr(s, ']', n); | |||
| 2029 | if (colon == NULL((void*)0) || ++colon == s + n || *colon != ':') | |||
| 2030 | colon = NULL((void*)0); | |||
| 2031 | } | |||
| 2032 | else | |||
| 2033 | colon = memchr(s, ':', n); | |||
| 2034 | ||||
| 2035 | if (colon) { | |||
| 2036 | canon_update(md5, s, colon - s, ""); /* host */ | |||
| 2037 | canon_update(md5, colon + 1, (s + n) - (colon + 1), ""); | |||
| 2038 | } | |||
| 2039 | else { | |||
| 2040 | canon_update(md5, s, n, ""); /* host */ | |||
| 2041 | su_md5_strupdate(md5, url_port_default(type)); /* port */ | |||
| 2042 | } | |||
| 2043 | ||||
| 2044 | /* ignore parameters/path/headers.... */ | |||
| 2045 | } | |||
| 2046 | ||||
| 2047 | ||||
| 2048 | /** Update md5 digest with contents of URL. | |||
| 2049 | * | |||
| 2050 | */ | |||
| 2051 | void url_update(su_md5_t *md5, url_t const *url) | |||
| 2052 | { | |||
| 2053 | if (url_string_p((url_string_t *)url)) { | |||
| 2054 | url_string_update(md5, (char const *)url); | |||
| 2055 | } | |||
| 2056 | else { | |||
| 2057 | SU_MD5_STRI0UPDATE(md5, url->url_scheme)su_md5_iupdate(md5, (url->url_scheme) ? (url->url_scheme ) : "", (url->url_scheme) ? strlen(url->url_scheme) : 1 ); | |||
| 2058 | SU_MD5_STRI0UPDATE(md5, url->url_user)su_md5_iupdate(md5, (url->url_user) ? (url->url_user) : "", (url->url_user) ? strlen(url->url_user) : 1); | |||
| 2059 | SU_MD5_STRI0UPDATE(md5, url->url_host)su_md5_iupdate(md5, (url->url_host) ? (url->url_host) : "", (url->url_host) ? strlen(url->url_host) : 1); | |||
| 2060 | su_md5_striupdate(md5, URL_PORT(url)((url) && (url)->url_port ? (url)->url_port : url_port_default ((url) ? (enum url_type_e)(url)->url_type : url_any))); | |||
| 2061 | /* XXX - parameters/path.... */ | |||
| 2062 | /* SU_MD5_STRI0UPDATE(md5, url->url_path); */ | |||
| 2063 | } | |||
| 2064 | } | |||
| 2065 | ||||
| 2066 | /** Calculate a digest from URL contents. */ | |||
| 2067 | void url_digest(void *hash, int hsize, url_t const *url, char const *key) | |||
| 2068 | { | |||
| 2069 | su_md5_t md5[1]; | |||
| 2070 | uint8_t digest[SU_MD5_DIGEST_SIZE16]; | |||
| 2071 | ||||
| 2072 | su_md5_init(md5); | |||
| 2073 | if (key) su_md5_strupdate(md5, key); | |||
| ||||
| 2074 | url_update(md5, url); | |||
| 2075 | su_md5_digest(md5, digest); | |||
| 2076 | ||||
| 2077 | if (hsize > SU_MD5_DIGEST_SIZE16) { | |||
| 2078 | memset((char *)hash + SU_MD5_DIGEST_SIZE16, 0, hsize - SU_MD5_DIGEST_SIZE16); | |||
| 2079 | hsize = SU_MD5_DIGEST_SIZE16; | |||
| 2080 | } | |||
| 2081 | ||||
| 2082 | memcpy(hash, digest, hsize); | |||
| 2083 | } | |||
| 2084 | ||||
| 2085 | /** Convert a URL query to a header string. | |||
| 2086 | * | |||
| 2087 | * URL query is converted by replacing each "=" in header name "=" value | |||
| 2088 | * pair with semicolon (":"), and the "&" separating header-name-value pairs | |||
| 2089 | * with line feed ("\n"). The "body" pseudoheader is moved last in the | |||
| 2090 | * string. The %-escaping is removed. Note that if the @a query contains %00, | |||
| 2091 | * the resulting string will be truncated. | |||
| 2092 | * | |||
| 2093 | * @param home memory home used to alloate string (if NULL, malloc() it) | |||
| 2094 | * @param query query part from SIP URL | |||
| 2095 | * | |||
| 2096 | * The result string is allocated from @a home, and it can be used as | |||
| 2097 | * argument to msg_header_parse_str(), msg_header_add_str() or | |||
| 2098 | * SIPTAG_HEADER_STR(). | |||
| 2099 | * | |||
| 2100 | * @sa msg_header_add_str(), SIPTAG_HEADER_STR(), | |||
| 2101 | * sip_headers_as_url_query(), sip_url_query_as_taglist(), | |||
| 2102 | * @RFC3261 section 19.1.1 "Headers", #url_t, url_s#url_headers, | |||
| 2103 | * url_unescape(), url_unescape_to() | |||
| 2104 | * | |||
| 2105 | * @since New in @VERSION_1_12_4. | |||
| 2106 | */ | |||
| 2107 | char *url_query_as_header_string(su_home_t *home, | |||
| 2108 | char const *query) | |||
| 2109 | { | |||
| 2110 | size_t i, j, n, b_start = 0, b_len = 0; | |||
| 2111 | char *s = su_strdup(home, query); | |||
| 2112 | ||||
| 2113 | if (!s) | |||
| 2114 | return NULL((void*)0); | |||
| 2115 | ||||
| 2116 | for (i = 0, j = 0; s[i];) { | |||
| 2117 | n = strcspn(s + i, "="); | |||
| 2118 | if (!s[i + n]) | |||
| 2119 | break; | |||
| 2120 | if (n == 4 && strncasecmp(s + i, "body", 4) == 0) { | |||
| 2121 | if (b_start) | |||
| 2122 | break; | |||
| 2123 | b_start = i + n + 1, b_len = strcspn(s + b_start, "&"); | |||
| 2124 | i = b_start + b_len + 1; | |||
| 2125 | continue; | |||
| 2126 | } | |||
| 2127 | if (i != j) | |||
| 2128 | memmove(s + j, s + i, n); | |||
| 2129 | s[j + n] = ':'; | |||
| 2130 | i += n + 1, j += n + 1; | |||
| 2131 | n = strcspn(s + i, "&"); | |||
| 2132 | j += url_unescape_to(s + j, s + i, n); | |||
| 2133 | i += n; | |||
| 2134 | if (s[i]) { | |||
| 2135 | s[j++] = '\n', i++; | |||
| 2136 | } | |||
| 2137 | } | |||
| 2138 | ||||
| 2139 | if (s[i]) | |||
| 2140 | return (void)su_free(home, s), NULL((void*)0); | |||
| 2141 | ||||
| 2142 | if (b_start) { | |||
| 2143 | s[j++] = '\n', s[j++] = '\n'; | |||
| 2144 | j += url_unescape_to(s + j, query + b_start, b_len); | |||
| 2145 | } | |||
| 2146 | s[j] = '\0'; assert(j <= i)((void) sizeof ((j <= i) ? 1 : 0), __extension__ ({ if (j <= i) ; else __assert_fail ("j <= i", "url.c", 2146, __extension__ __PRETTY_FUNCTION__); })); | |||
| 2147 | ||||
| 2148 | return s; | |||
| 2149 | } |
| 1 | /* | |||||
| 2 | * This file is part of the Sofia-SIP package | |||||
| 3 | * | |||||
| 4 | * Copyright (C) 2005 Nokia Corporation. | |||||
| 5 | * | |||||
| 6 | * Contact: Pekka Pessi <pekka.pessi@nokia.com> | |||||
| 7 | * | |||||
| 8 | * This library is free software; you can redistribute it and/or | |||||
| 9 | * modify it under the terms of the GNU Lesser General Public License | |||||
| 10 | * as published by the Free Software Foundation; either version 2.1 of | |||||
| 11 | * the License, or (at your option) any later version. | |||||
| 12 | * | |||||
| 13 | * This library is distributed in the hope that it will be useful, but | |||||
| 14 | * WITHOUT ANY WARRANTY; without even the implied warranty of | |||||
| 15 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |||||
| 16 | * Lesser General Public License for more details. | |||||
| 17 | * | |||||
| 18 | * You should have received a copy of the GNU Lesser General Public | |||||
| 19 | * License along with this library; if not, write to the Free Software | |||||
| 20 | * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA | |||||
| 21 | * 02110-1301 USA | |||||
| 22 | * | |||||
| 23 | */ | |||||
| 24 | ||||||
| 25 | #ifndef BNF_H /** Defined when <sofia-sip/bnf.h> has been included. */ | |||||
| 26 | #define BNF_H | |||||
| 27 | ||||||
| 28 | /**@file sofia-sip/bnf.h | |||||
| 29 | * | |||||
| 30 | * Parsing macros and prototypes for HTTP-like protocols. | |||||
| 31 | * | |||||
| 32 | * @author Pekka Pessi <Pekka.Pessi@nokia.com> | |||||
| 33 | * | |||||
| 34 | * @date Created: Tue Jun 06 10:59:34 2000 ppessi | |||||
| 35 | * | |||||
| 36 | */ | |||||
| 37 | ||||||
| 38 | #include <sofia-sip/su_types.h> | |||||
| 39 | ||||||
| 40 | #include <string.h> | |||||
| 41 | ||||||
| 42 | SOFIA_BEGIN_DECLS | |||||
| 43 | ||||||
| 44 | /* Parsing tokens */ | |||||
| 45 | /** Control characters. */ | |||||
| 46 | #define CTL"\001\002\003\004\005\006\007" "\010\011\012\013\014\015\016\017" "\020\021\022\023\024\025\026\027" "\030\031\032\033\034\035\036\037" "\177" "\0" "\001\002\003\004\005\006\007" \ | |||||
| 47 | "\010\011\012\013\014\015\016\017" \ | |||||
| 48 | "\020\021\022\023\024\025\026\027" \ | |||||
| 49 | "\030\031\032\033\034\035\036\037" "\177" "\0" | |||||
| 50 | /** Space */ | |||||
| 51 | #define SP" " " " | |||||
| 52 | /** Horizontal tab */ | |||||
| 53 | #define HT"\t" "\t" | |||||
| 54 | /** Carriage return */ | |||||
| 55 | /* CR conflicts with Windows SDK 10, so it is now _CR */ | |||||
| 56 | #define _CR"\r" "\r" | |||||
| 57 | /** Line feed */ | |||||
| 58 | #define LF"\n" "\n" | |||||
| 59 | /** Line-ending characters */ | |||||
| 60 | #define CRLF"\r" "\n" _CR"\r" LF"\n" | |||||
| 61 | /** Whitespace */ | |||||
| 62 | #define WS" " "\t" SP" " HT"\t" | |||||
| 63 | /** Linear whitespace */ | |||||
| 64 | #define LWS" " "\t" "\r" "\n" SP" " HT"\t" _CR"\r" LF"\n" | |||||
| 65 | /** Lower-case alphabetic characters */ | |||||
| 66 | #define LOALPHA"abcdefghijklmnopqrstuvwxyz" "abcdefghijklmnopqrstuvwxyz" | |||||
| 67 | /** Upper-case alphabetic characters */ | |||||
| 68 | #define UPALPHA"ABCDEFGHIJKLMNOPQRSTUVWXYZ" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" | |||||
| 69 | /** Alphabetic characters */ | |||||
| 70 | #define ALPHA"abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" LOALPHA"abcdefghijklmnopqrstuvwxyz" UPALPHA"ABCDEFGHIJKLMNOPQRSTUVWXYZ" | |||||
| 71 | /** Digits */ | |||||
| 72 | #define DIGIT"0123456789" "0123456789" | |||||
| 73 | /** RTSP safe characters */ | |||||
| 74 | #define SAFE"$-_." "$-_." /* RTSP stuff */ | |||||
| 75 | #define ALPHANUM"0123456789" "abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" DIGIT"0123456789" ALPHA"abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" | |||||
| 76 | #define HEX"0123456789" "ABCDEF" "abcdef" DIGIT"0123456789" "ABCDEF" "abcdef" | |||||
| 77 | ||||||
| 78 | /** SIP token characters. | |||||
| 79 | * @note $|&^# were token chars in RFC 2543, but no more in RFC 3261. | |||||
| 80 | */ | |||||
| 81 | #define SIP_TOKEN"0123456789" "abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "-.!%*_+`'~" ALPHANUM"0123456789" "abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "-.!%*_+`'~" | |||||
| 82 | /** SIP separator characters */ | |||||
| 83 | #define SIP_SEPARATOR"()<>@,;:\\\"/[]?={}" " " "\t" "()<>@,;:\\\"/[]?={}" SP" " HT"\t" | |||||
| 84 | ||||||
| 85 | /** SIP Word characters (that are not token characters) */ | |||||
| 86 | #define SIP_WORD"()<>:\\\"/[]?{}" "()<>:\\\"/[]?{}" | |||||
| 87 | ||||||
| 88 | /** Skip whitespace (SP HT) */ | |||||
| 89 | #define skip_ws(ss)(*(ss) += strspn(*(ss), " " "\t")) (*(ss) += span_ws(*(ss))strspn(*(ss), " " "\t")) | |||||
| 90 | ||||||
| 91 | /** Skip linear whitespace (SP HT CR LF) */ | |||||
| 92 | #define skip_lws(ss)(*(ss) += span_lws(*(ss))) (*(ss) += span_lws(*(ss))) | |||||
| 93 | ||||||
| 94 | /** Skip [a-zA-Z] */ | |||||
| 95 | #define skip_alpha(ss)(*(ss) += span_alpha(*(ss))) (*(ss) += span_alpha(*(ss))) | |||||
| 96 | ||||||
| 97 | /** Skip digits */ | |||||
| 98 | #define skip_digit(ss)(*(ss) += span_digit(*(ss))) (*(ss) += span_digit(*(ss))) | |||||
| 99 | ||||||
| 100 | /** Skip characters belonging to an RTSP token. */ | |||||
| 101 | #define skip_alpha_digit_safe(ss)(*(ss) += span_alpha_digit_safe(*(ss))) (*(ss) += span_alpha_digit_safe(*(ss))) | |||||
| 102 | ||||||
| 103 | /** Skip characters belonging to a SIP token. */ | |||||
| 104 | #define skip_token(ss)(*(ss) += span_token(*(ss))) (*(ss) += span_token(*(ss))) | |||||
| 105 | ||||||
| 106 | /** Skip characters belonging to a SIP parameter value. */ | |||||
| 107 | #define skip_param(ss)(*(ss) += span_param(*(ss))) (*(ss) += span_param(*(ss))) | |||||
| 108 | ||||||
| 109 | /** Skip characters belonging to a SIP word. */ | |||||
| 110 | #define skip_word(ss)(*(ss) += span_word(*(ss))) (*(ss) += span_word(*(ss))) | |||||
| 111 | ||||||
| 112 | /** Test if @c is CR or LF */ | |||||
| 113 | #define IS_CRLF(c)((c) == '\r' || (c) == '\n') ((c) == '\r' || (c) == '\n') | |||||
| 114 | /** Test if @c is linear whitespace */ | |||||
| 115 | #define IS_LWS(c)((c) == ' ' || (c) == '\t' || (c) == '\r' || (c) == '\n') ((c) == ' ' || (c) == '\t' || (c) == '\r' || (c) == '\n') | |||||
| 116 | /*#define IS_LWS(c) ((_bnf_table[(unsigned char)(c)] & bnf_lws))*/ | |||||
| 117 | /** Test if @c is normal whitespace */ | |||||
| 118 | #define IS_WS(c)((c) == ' ' || (c) == '\t') ((c) == ' ' || (c) == '\t') | |||||
| 119 | /** Test if @c is not whitespace (and not NUL). */ | |||||
| 120 | #define IS_NON_WS(c)(c && !((c) == ' ' || (c) == '\t')) (c && !IS_WS(c)((c) == ' ' || (c) == '\t')) | |||||
| 121 | /*#define IS_NON_WS(c) (c && !(_bnf_table[(unsigned char)c] & bnf_ws))*/ | |||||
| 122 | /** Test if @c is not linear whitespace (and not NUL). */ | |||||
| 123 | #define IS_NON_LWS(c)(c && !((c) == ' ' || (c) == '\t' || (c) == '\r' || ( c) == '\n')) (c && !IS_LWS(c)((c) == ' ' || (c) == '\t' || (c) == '\r' || (c) == '\n')) | |||||
| 124 | /*#define IS_NON_LWS(c) (c && !(_bnf_table[(unsigned char)c] & bnf_lws))*/ | |||||
| 125 | /** Test if @c is a digit. */ | |||||
| 126 | #define IS_DIGIT(c)((c) >= '0' && (c) <= '9') ((c) >= '0' && (c) <= '9') | |||||
| 127 | /** Test if @c is alphabetic. */ | |||||
| 128 | #define IS_ALPHA(c)(c && ((_bnf_table[(unsigned char)c] & bnf_alpha) )) (c && ((_bnf_table[(unsigned char)c] & bnf_alpha))) | |||||
| 129 | /** Test if @c is alphanumeric. */ | |||||
| 130 | #define IS_ALPHANUM(c)(c && (((c) >= '0' && (c) <= '9') || (c && ((_bnf_table[(unsigned char)c] & bnf_alpha))) )) (c && (IS_DIGIT(c)((c) >= '0' && (c) <= '9') || IS_ALPHA(c)(c && ((_bnf_table[(unsigned char)c] & bnf_alpha) )))) | |||||
| 131 | /** Test if @c is URL-unreserved. */ | |||||
| 132 | #define IS_UNRESERVED(c)((_bnf_table[(unsigned char)c] & bnf_unreserved)) ((_bnf_table[(unsigned char)c] & bnf_unreserved)) | |||||
| 133 | /** Test if @c is URL-reserved. */ | |||||
| 134 | #define IS_RESERVED(c)(c && !(_bnf_table[(unsigned char)c] & bnf_unreserved )) (c && !(_bnf_table[(unsigned char)c] & bnf_unreserved)) | |||||
| 135 | /** Test if @c is valid in tokens. */ | |||||
| 136 | #define IS_TOKEN(c)((_bnf_table[(unsigned char)c] & bnf_token)) ((_bnf_table[(unsigned char)c] & bnf_token)) | |||||
| 137 | /** Test if @c is valid for SIP parameter value. */ | |||||
| 138 | #define IS_PARAM(c)((_bnf_table[(unsigned char)c] & (bnf_token|bnf_param))) ((_bnf_table[(unsigned char)c] & (bnf_token|bnf_param))) | |||||
| 139 | /** Test if @c is a hex digit. */ | |||||
| 140 | #define IS_HEX(c)(((c) >= '0' && (c) <= '9') || ((c) >= 'A' && (c) <= 'F') || ((c) >= 'a' && (c) <= 'f')) (((c) >= '0' && (c) <= '9') || ((c) >= 'A' && (c) <= 'F') || ((c) >= 'a' && (c) <= 'f')) | |||||
| 141 | /** Test if @c is a linear whitespace or valid in tokens. */ | |||||
| 142 | #define IS_TOKENLWS(c)((_bnf_table[(unsigned char)c] & (bnf_token|bn_lws))) ((_bnf_table[(unsigned char)c] & (bnf_token|bn_lws))) | |||||
| 143 | ||||||
| 144 | enum { | |||||
| 145 | bnf_ws = 1, /**< Whitespace character */ | |||||
| 146 | bnf_crlf = 2, /**< Line end character */ | |||||
| 147 | bnf_lws = 3, /**< Linear whitespace */ | |||||
| 148 | bnf_alpha = 4, /**< Alphabetic */ | |||||
| 149 | bnf_safe = 8, /**< RTSP safe */ | |||||
| 150 | bnf_mark = 16, /**< URL mark */ | |||||
| 151 | bnf_unreserved = bnf_alpha | bnf_mark, /**< URL unreserved */ | |||||
| 152 | bnf_separator = 32, /**< SIP separator */ | |||||
| 153 | /** SIP token, not alphabetic (0123456789-.!%*_+`'~) */ | |||||
| 154 | bnf_token0 = 64 | bnf_safe, | |||||
| 155 | bnf_token = bnf_token0 | bnf_alpha, /**< SIP token */ | |||||
| 156 | bnf_param0 = 128, /**< SIP parameter, not token */ | |||||
| 157 | bnf_param = bnf_token | bnf_param0 /**< SIP/HTTP parameter */ | |||||
| 158 | }; | |||||
| 159 | ||||||
| 160 | /** Table for determining class of a character. */ | |||||
| 161 | SOFIAPUBVARextern unsigned char const _bnf_table[256]; | |||||
| 162 | ||||||
| 163 | /** Get number of characters before CRLF */ | |||||
| 164 | #define span_non_crlf(s)strcspn(s, "\r" "\n") strcspn(s, _CR"\r" LF"\n") | |||||
| 165 | ||||||
| 166 | /** Get number of characters before whitespace */ | |||||
| 167 | #define span_non_ws(s)strcspn(s, " " "\t") strcspn(s, WS" " "\t") | |||||
| 168 | ||||||
| 169 | /** Get number of whitespace characters */ | |||||
| 170 | #define span_ws(s)strspn(s, " " "\t") strspn(s, WS" " "\t") | |||||
| 171 | ||||||
| 172 | /** Get number of characters before linear whitespace */ | |||||
| 173 | #define span_non_lws(s)strcspn(s, " " "\t" "\r" "\n") strcspn(s, LWS" " "\t" "\r" "\n") | |||||
| 174 | ||||||
| 175 | /** Calculate span of a linear whitespace. | |||||
| 176 | * LWS = [*WSP CRLF] 1*WSP | |||||
| 177 | */ | |||||
| 178 | su_inlinestatic inline isize_t span_lws(char const *s) | |||||
| 179 | { | |||||
| 180 | char const *e = s; | |||||
| 181 | int i = 0; | |||||
| 182 | e += strspn(s, WS" " "\t"); | |||||
| 183 | if (e[i] == '\r') i++; | |||||
| 184 | if (e[i] == '\n') i++; | |||||
| 185 | if (IS_WS(e[i])((e[i]) == ' ' || (e[i]) == '\t')) | |||||
| 186 | e += i + strspn(e + i, WS" " "\t"); | |||||
| 187 | return e - s; | |||||
| 188 | } | |||||
| 189 | ||||||
| 190 | /** Calculate span of a token or linear whitespace characters. */ | |||||
| 191 | su_inlinestatic inline isize_t span_token_lws(char const *s) | |||||
| 192 | { | |||||
| 193 | char const *e = s; | |||||
| 194 | while (_bnf_table[(unsigned char)(*e)] & (bnf_token | bnf_lws)) | |||||
| 195 | e++; | |||||
| 196 | return e - s; | |||||
| 197 | } | |||||
| 198 | ||||||
| 199 | /** Calculate span of a token characters. */ | |||||
| 200 | su_inlinestatic inline isize_t span_token(char const *s) | |||||
| 201 | { | |||||
| 202 | char const *e = s; | |||||
| 203 | while (_bnf_table[(unsigned char)(*e)] & bnf_token) | |||||
| 204 | e++; | |||||
| 205 | return e - s; | |||||
| 206 | } | |||||
| 207 | ||||||
| 208 | /** Calculate span of a alphabetic characters. */ | |||||
| 209 | su_inlinestatic inline isize_t span_alpha(char const *s) | |||||
| 210 | { | |||||
| 211 | char const *e = s; | |||||
| 212 | while (_bnf_table[(unsigned char)(*e)] & bnf_alpha) | |||||
| 213 | e++; | |||||
| 214 | return e - s; | |||||
| 215 | } | |||||
| 216 | ||||||
| 217 | /** Calculate span of a digits. */ | |||||
| 218 | su_inlinestatic inline isize_t span_digit(char const *s) | |||||
| 219 | { | |||||
| 220 | char const *e = s; | |||||
| 221 | while (*e >= '0' && *e <= '9') | |||||
| 222 | e++; | |||||
| 223 | return e - s; | |||||
| 224 | } | |||||
| 225 | ||||||
| 226 | /** Calculate span of a hex. */ | |||||
| 227 | su_inlinestatic inline isize_t span_hexdigit(char const *s) | |||||
| 228 | { | |||||
| 229 | char const *e = s; | |||||
| 230 | while (IS_HEX(*e)(((*e) >= '0' && (*e) <= '9') || ((*e) >= 'A' && (*e) <= 'F') || ((*e) >= 'a' && (*e ) <= 'f'))) | |||||
| 231 | e++; | |||||
| 232 | return e - s; | |||||
| 233 | } | |||||
| 234 | ||||||
| 235 | /** Calculate span of characters belonging to an RTSP token */ | |||||
| 236 | su_inlinestatic inline isize_t span_alpha_digit_safe(char const *s) | |||||
| 237 | { | |||||
| 238 | char const *e = s; | |||||
| 239 | while (_bnf_table[(unsigned char)(*e)] & (bnf_alpha | bnf_safe)) | |||||
| 240 | e++; | |||||
| 241 | return e - s; | |||||
| 242 | } | |||||
| 243 | ||||||
| 244 | /** Calculate span of a characters valid in parameters. */ | |||||
| 245 | su_inlinestatic inline isize_t span_param(char const *s) | |||||
| 246 | { | |||||
| 247 | char const *e = s; | |||||
| 248 | while (IS_PARAM(*e)((_bnf_table[(unsigned char)*e] & (bnf_token|bnf_param)))) | |||||
| 249 | e++; | |||||
| 250 | return e - s; | |||||
| 251 | } | |||||
| 252 | ||||||
| 253 | /** Calculate span of a SIP word. */ | |||||
| 254 | su_inlinestatic inline isize_t span_word(char const *s) | |||||
| 255 | { | |||||
| 256 | char const *e = s; | |||||
| 257 | while (*e && (IS_TOKEN(*e)((_bnf_table[(unsigned char)*e] & bnf_token)) || strchr(SIP_WORD"()<>:\\\"/[]?{}", *e))) | |||||
| 258 | e++; | |||||
| 259 | return e - s; | |||||
| 260 | } | |||||
| 261 | ||||||
| 262 | /** Calculate span of a unreserved characters. */ | |||||
| 263 | su_inlinestatic inline isize_t span_unreserved(char const *s) | |||||
| 264 | { | |||||
| 265 | char const *e = s; | |||||
| 266 | while (IS_UNRESERVED(*e)((_bnf_table[(unsigned char)*e] & bnf_unreserved))) | |||||
| ||||||
| 267 | e++; | |||||
| 268 | return e - s; | |||||
| 269 | } | |||||
| 270 | ||||||
| 271 | /** Calculate span of a double quoted string (with escaped chars inside) */ | |||||
| 272 | su_inlinestatic inline isize_t span_quoted(char const *s) | |||||
| 273 | { | |||||
| 274 | char const *b = s; | |||||
| 275 | ||||||
| 276 | if (*s++ != '"') | |||||
| 277 | return 0; | |||||
| 278 | ||||||
| 279 | for (;;) { | |||||
| 280 | s += strcspn(s, "\\\""); | |||||
| 281 | if (!*s) | |||||
| 282 | return 0; | |||||
| 283 | if (*s++ == '"') | |||||
| 284 | return s - b; | |||||
| 285 | if (!*s++) | |||||
| 286 | return 0; | |||||
| 287 | } | |||||
| 288 | } | |||||
| 289 | ||||||
| 290 | /* RFC 2396 defines URL chars */ | |||||
| 291 | /** Reserved in URLs */ | |||||
| 292 | #define URL_RESERVED";/?:=+$," ";/?:=+$," | |||||
| 293 | ||||||
| 294 | /** Non-alphanumeric characters without syntactical meaning. */ | |||||
| 295 | #define URL_MARK"-_.!~*'()" "-_.!~*'()" | |||||
| 296 | ||||||
| 297 | /** Unreserved characters. */ | |||||
| 298 | #define URL_UNRESERVED"0123456789" "abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "-_.!~*'()" ALPHANUM"0123456789" "abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" URL_MARK"-_.!~*'()" | |||||
| 299 | ||||||
| 300 | /** URL hex escape. */ | |||||
| 301 | #define URL_ESCAPED"%" "%" | |||||
| 302 | #define URL_DELIMS"<>#%\"" "<>#%\"" | |||||
| 303 | #define URL_UNWISE"{}|\\^[]`" "{}|\\^[]`" | |||||
| 304 | #define URL_SCHEME"0123456789" "abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "+-." ALPHANUM"0123456789" "abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "+-." | |||||
| 305 | ||||||
| 306 | /** Get number of characters belonging to url scheme */ | |||||
| 307 | #define span_url_scheme(s)strspn(s, "0123456789" "abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "+-.") strspn(s, URL_SCHEME"0123456789" "abcdefghijklmnopqrstuvwxyz" "ABCDEFGHIJKLMNOPQRSTUVWXYZ" "+-.") | |||||
| 308 | ||||||
| 309 | SOFIAPUBFUN int span_ip4_address(char const *host); | |||||
| 310 | SOFIAPUBFUN int span_ip6_address(char const *host); | |||||
| 311 | SOFIAPUBFUN int span_ip6_reference(char const *host); | |||||
| 312 | SOFIAPUBFUN int span_ip_address(char const *host); | |||||
| 313 | SOFIAPUBFUN isize_t span_domain(char const *host); | |||||
| 314 | SOFIAPUBFUN isize_t span_host(char const *host); | |||||
| 315 | ||||||
| 316 | SOFIAPUBFUN int scan_ip4_address(char **inout_host); | |||||
| 317 | SOFIAPUBFUN int scan_ip6_address(char **inout_host); | |||||
| 318 | SOFIAPUBFUN int scan_ip6_reference(char **inout_host); | |||||
| 319 | SOFIAPUBFUN int scan_ip_address(char **inout_host); | |||||
| 320 | SOFIAPUBFUN issize_t scan_domain(char **inout_host); | |||||
| 321 | SOFIAPUBFUN issize_t scan_host(char **inout_host); | |||||
| 322 | ||||||
| 323 | SOFIA_END_DECLS | |||||
| 324 | ||||||
| 325 | #endif /* !defined BNF_H */ |