| 1 |
/* |
| 2 |
URI manipulation routines. |
| 3 |
Copyright (C) 1999-2006, Joe Orton <[email protected]> |
| 4 |
|
| 5 |
This library is free software; you can redistribute it and/or |
| 6 |
modify it under the terms of the GNU Library General Public |
| 7 |
License as published by the Free Software Foundation; either |
| 8 |
version 2 of the License, or (at your option) any later version. |
| 9 |
|
| 10 |
This library is distributed in the hope that it will be useful, |
| 11 |
but WITHOUT ANY WARRANTY; without even the implied warranty of |
| 12 |
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
| 13 |
Library General Public License for more details. |
| 14 |
|
| 15 |
You should have received a copy of the GNU Library General Public |
| 16 |
License along with this library; if not, write to the Free |
| 17 |
Software Foundation, Inc., 59 Temple Place - Suite 330, Boston, |
| 18 |
MA 02111-1307, USA |
| 19 |
|
| 20 |
*/ |
| 21 |
|
| 22 |
#include "config.h" |
| 23 |
|
| 24 |
#ifdef HAVE_STRING_H |
| 25 |
#include <string.h> |
| 26 |
#endif |
| 27 |
#ifdef HAVE_STRINGS_H |
| 28 |
#include <strings.h> |
| 29 |
#endif |
| 30 |
#ifdef HAVE_UNISTD_H |
| 31 |
#include <unistd.h> |
| 32 |
#endif |
| 33 |
#ifdef HAVE_STDLIB_H |
| 34 |
#include <stdlib.h> |
| 35 |
#endif |
| 36 |
|
| 37 |
#include <stdio.h> |
| 38 |
|
| 39 |
#include <ctype.h> |
| 40 |
|
| 41 |
#include "ne_string.h" /* for ne_buffer */ |
| 42 |
#include "ne_alloc.h" |
| 43 |
#include "ne_uri.h" |
| 44 |
|
| 45 |
/* URI ABNF from RFC 3986: */ |
| 46 |
|
| 47 |
#define PS (0x0001) /* "+" */ |
| 48 |
#define PC (0x0002) /* "%" */ |
| 49 |
#define DS (0x0004) /* "-" */ |
| 50 |
#define DT (0x0008) /* "." */ |
| 51 |
#define US (0x0010) /* "_" */ |
| 52 |
#define TD (0x0020) /* "~" */ |
| 53 |
#define FS (0x0040) /* "/" */ |
| 54 |
#define CL (0x0080) /* ":" */ |
| 55 |
#define AT (0x0100) /* "@" */ |
| 56 |
#define QU (0x0200) /* "?" */ |
| 57 |
|
| 58 |
#define DG (0x0400) /* DIGIT */ |
| 59 |
#define AL (0x0800) /* ALPHA */ |
| 60 |
|
| 61 |
#define GD (0x1000) /* gen-delims = "#" / "[" / "]" |
| 62 |
* ... except ":", "/", "@", and "?" */ |
| 63 |
|
| 64 |
#define SD (0x2000) /* sub-delims = "!" / "$" / "&" / "'" / "(" / ")" |
| 65 |
* / "*" / "+" / "," / ";" / "=" |
| 66 |
* ... except "+" which is PS */ |
| 67 |
|
| 68 |
#define OT (0x4000) /* others */ |
| 69 |
|
| 70 |
#define URI_ALPHA (AL) |
| 71 |
#define URI_DIGIT (DG) |
| 72 |
|
| 73 |
/* unreserved = ALPHA / DIGIT / "-" / "." / "_" / "~" */ |
| 74 |
#define URI_UNRESERVED (AL | DG | DS | DT | US | TD) |
| 75 |
/* scheme = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." ) */ |
| 76 |
#define URI_SCHEME (AL | DG | PS | DS | DT) |
| 77 |
/* real sub-delims definition, including "+" */ |
| 78 |
#define URI_SUBDELIM (PS | SD) |
| 79 |
/* real gen-delims definition, including ":", "/", "@" and "?" */ |
| 80 |
#define URI_GENDELIM (GD | CL | FS | AT | QU) |
| 81 |
/* userinfo = *( unreserved / pct-encoded / sub-delims / ":" ) */ |
| 82 |
#define URI_USERINFO (URI_UNRESERVED | PC | URI_SUBDELIM | CL) |
| 83 |
/* pchar = unreserved / pct-encoded / sub-delims / ":" / "@" */ |
| 84 |
#define URI_PCHAR (URI_UNRESERVED | PC | URI_SUBDELIM | CL | AT) |
| 85 |
/* invented: segchar = pchar / "/" */ |
| 86 |
#define URI_SEGCHAR (URI_PCHAR | FS) |
| 87 |
/* query = *( pchar / "/" / "?" ) */ |
| 88 |
#define URI_QUERY (URI_PCHAR | FS | QU) |
| 89 |
/* fragment == query */ |
| 90 |
#define URI_FRAGMENT URI_QUERY |
| 91 |
|
| 92 |
/* any characters which should be path-escaped: */ |
| 93 |
#define URI_ESCAPE ((URI_GENDELIM & ~(FS)) | URI_SUBDELIM | OT | PC) |
| 94 |
|
| 95 |
static const unsigned int uri_chars[256] = { |
| 96 |
/* 0xXX x0 x2 x4 x6 x8 xA xC xE */ |
| 97 |
/* 0x */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, |
| 98 |
/* 1x */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, |
| 99 |
/* 2x */ OT, SD, OT, GD, SD, PC, SD, SD, SD, SD, SD, PS, SD, DS, DT, FS, |
| 100 |
/* 3x */ DG, DG, DG, DG, DG, DG, DG, DG, DG, DG, CL, SD, OT, SD, OT, QU, |
| 101 |
/* 4x */ AT, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, |
| 102 |
/* 5x */ AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, GD, OT, GD, OT, US, |
| 103 |
/* 6x */ OT, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, |
| 104 |
/* 7x */ AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, AL, OT, OT, OT, TD, OT, |
| 105 |
/* 8x */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, |
| 106 |
/* 9x */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, |
| 107 |
/* Ax */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, |
| 108 |
/* Bx */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, |
| 109 |
/* Cx */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, |
| 110 |
/* Dx */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, |
| 111 |
/* Ex */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, |
| 112 |
/* Fx */ OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT, OT |
| 113 |
}; |
| 114 |
|
| 115 |
#define uri_lookup(ch) (uri_chars[(unsigned char)ch]) |
| 116 |
|
| 117 |
char *ne_path_parent(const char *uri) |
| 118 |
{ |
| 119 |
size_t len = strlen(uri); |
| 120 |
const char *pnt = uri + len - 1; |
| 121 |
/* skip trailing slash (parent of "/foo/" is "/") */ |
| 122 |
if (pnt >= uri && *pnt == '/') |
| 123 |
pnt--; |
| 124 |
/* find previous slash */ |
| 125 |
while (pnt > uri && *pnt != '/') |
| 126 |
pnt--; |
| 127 |
if (pnt < uri || (pnt == uri && *pnt != '/')) |
| 128 |
return NULL; |
| 129 |
return ne_strndup(uri, pnt - uri + 1); |
| 130 |
} |
| 131 |
|
| 132 |
int ne_path_has_trailing_slash(const char *uri) |
| 133 |
{ |
| 134 |
size_t len = strlen(uri); |
| 135 |
return ((len > 0) && |
| 136 |
(uri[len-1] == '/')); |
| 137 |
} |
| 138 |
|
| 139 |
unsigned int ne_uri_defaultport(const char *scheme) |
| 140 |
{ |
| 141 |
/* RFC2616/3.2.3 says use case-insensitive comparisons here. */ |
| 142 |
if (ne_strcasecmp(scheme, "http") == 0) |
| 143 |
return 80; |
| 144 |
else if (ne_strcasecmp(scheme, "https") == 0) |
| 145 |
return 443; |
| 146 |
else |
| 147 |
return 0; |
| 148 |
} |
| 149 |
|
| 150 |
int ne_uri_parse(const char *uri, ne_uri *parsed) |
| 151 |
{ |
| 152 |
const char *p, *s; |
| 153 |
|
| 154 |
memset(parsed, 0, sizeof *parsed); |
| 155 |
|
| 156 |
p = s = uri; |
| 157 |
|
| 158 |
/* => s = p = URI-reference */ |
| 159 |
|
| 160 |
if (uri_lookup(*p) & URI_ALPHA) { |
| 161 |
while (uri_lookup(*p) & URI_SCHEME) |
| 162 |
p++; |
| 163 |
|
| 164 |
if (*p == ':') { |
| 165 |
parsed->scheme = ne_strndup(uri, p - s); |
| 166 |
s = p + 1; |
| 167 |
} |
| 168 |
} |
| 169 |
|
| 170 |
/* => s = heir-part, or s = relative-part */ |
| 171 |
|
| 172 |
if (s[0] == '/' && s[1] == '/') { |
| 173 |
const char *pa; |
| 174 |
|
| 175 |
/* => s = "//" authority path-abempty (from expansion of |
| 176 |
* either heir-part of relative-part) */ |
| 177 |
|
| 178 |
/* authority = [ userinfo "@" ] host [ ":" port ] */ |
| 179 |
|
| 180 |
s = pa = s + 2; /* => s = authority */ |
| 181 |
|
| 182 |
while (*pa != '/' && *pa != '\0') |
| 183 |
pa++; |
| 184 |
/* => pa = path-abempty */ |
| 185 |
|
| 186 |
p = s; |
| 187 |
while (p < pa && uri_lookup(*p) & URI_USERINFO) |
| 188 |
p++; |
| 189 |
|
| 190 |
if (*p == '@') { |
| 191 |
parsed->userinfo = ne_strndup(s, p - s); |
| 192 |
s = p + 1; |
| 193 |
} |
| 194 |
/* => s = host */ |
| 195 |
|
| 196 |
if (s[0] == '[') { |
| 197 |
p = s + 1; |
| 198 |
|
| 199 |
while (*p != ']' && p < pa) |
| 200 |
p++; |
| 201 |
|
| 202 |
if (p == pa || (p + 1 != pa && p[1] != ':')) { |
| 203 |
/* Ill-formed IP-literal. */ |
| 204 |
return -1; |
| 205 |
} |
| 206 |
|
| 207 |
p++; /* => p = colon */ |
| 208 |
} else { |
| 209 |
/* Find the colon. */ |
| 210 |
p = pa; |
| 211 |
while (*p != ':' && p > s) |
| 212 |
p--; |
| 213 |
} |
| 214 |
|
| 215 |
if (p == s) { |
| 216 |
p = pa; |
| 217 |
/* No colon; => p = path-abempty */ |
| 218 |
} else if (p + 1 != pa) { |
| 219 |
/* => p = colon */ |
| 220 |
parsed->port = atoi(p + 1); |
| 221 |
if(parsed->port == 0) |
| 222 |
parsed->port = UINT_MAX; |
| 223 |
} |
| 224 |
parsed->host = ne_strndup(s, p - s); |
| 225 |
|
| 226 |
s = pa; |
| 227 |
|
| 228 |
if (*s == '\0') { |
| 229 |
s = "/"; /* FIXME: scheme-specific. */ |
| 230 |
} |
| 231 |
} |
| 232 |
|
| 233 |
/* => s = path-abempty / path-absolute / path-rootless |
| 234 |
* / path-empty / path-noscheme */ |
| 235 |
|
| 236 |
p = s; |
| 237 |
|
| 238 |
while (uri_lookup(*p) & URI_SEGCHAR) |
| 239 |
p++; |
| 240 |
|
| 241 |
/* => p = [ "?" query ] [ "#" fragment ] */ |
| 242 |
|
| 243 |
parsed->path = ne_strndup(s, p - s); |
| 244 |
|
| 245 |
if (*p != '\0') { |
| 246 |
s = p++; |
| 247 |
|
| 248 |
while (uri_lookup(*p) & URI_QUERY) |
| 249 |
p++; |
| 250 |
|
| 251 |
/* => p = [ "#" fragment ] */ |
| 252 |
/* => s = [ "?" query ] [ "#" fragment ] */ |
| 253 |
|
| 254 |
if (*s == '?') { |
| 255 |
parsed->query = ne_strndup(s + 1, p - s - 1); |
| 256 |
|
| 257 |
if (*p != '\0') { |
| 258 |
s = p++; |
| 259 |
|
| 260 |
while (uri_lookup(*p) & URI_FRAGMENT) |
| 261 |
p++; |
| 262 |
} |
| 263 |
} |
| 264 |
|
| 265 |
/* => p now points to the next character after the |
| 266 |
* URI-reference; which should be the NUL byte. */ |
| 267 |
|
| 268 |
if (*s == '#') { |
| 269 |
parsed->fragment = ne_strndup(s + 1, p - s - 1); |
| 270 |
} |
| 271 |
else if (*p || *s != '?') { |
| 272 |
return -1; |
| 273 |
} |
| 274 |
} |
| 275 |
|
| 276 |
return 0; |
| 277 |
} |
| 278 |
|
| 279 |
/* This function directly implements the "Merge Paths" algorithm |
| 280 |
* described in RFC 3986 section 5.2.3. */ |
| 281 |
static char *merge_paths(const ne_uri *base, const char *path) |
| 282 |
{ |
| 283 |
const char *p; |
| 284 |
|
| 285 |
if (base->host && base->path[0] == '\0') { |
| 286 |
return ne_concat("/", path, NULL); |
| 287 |
} |
| 288 |
|
| 289 |
p = strrchr(base->path, '/'); |
| 290 |
if (p == NULL) { |
| 291 |
return ne_strdup(path); |
| 292 |
} else { |
| 293 |
size_t len = p - base->path + 1; |
| 294 |
char *ret = ne_malloc(strlen(path) + len + 1); |
| 295 |
|
| 296 |
memcpy(ret, base->path, len); |
| 297 |
memcpy(ret + len, path, strlen(path) + 1); |
| 298 |
return ret; |
| 299 |
} |
| 300 |
} |
| 301 |
|
| 302 |
/* This function directly implements the "Remove Dot Segments" |
| 303 |
* algorithm described in RFC 3986 section 5.2.4. */ |
| 304 |
static char *remove_dot_segments(const char *path) |
| 305 |
{ |
| 306 |
char *in, *inc, *out; |
| 307 |
|
| 308 |
inc = in = ne_strdup(path); |
| 309 |
out = ne_malloc(strlen(path) + 1); |
| 310 |
out[0] = '\0'; |
| 311 |
|
| 312 |
while (in[0]) { |
| 313 |
/* case 2.A: */ |
| 314 |
if (strncmp(in, "./", 2) == 0) { |
| 315 |
in += 2; |
| 316 |
} |
| 317 |
else if (strncmp(in, "../", 3) == 0) { |
| 318 |
in += 3; |
| 319 |
} |
| 320 |
|
| 321 |
/* case 2.B: */ |
| 322 |
else if (strncmp(in, "/./", 3) == 0) { |
| 323 |
in += 2; |
| 324 |
} |
| 325 |
else if (strcmp(in, "/.") == 0) { |
| 326 |
in[1] = '\0'; |
| 327 |
} |
| 328 |
|
| 329 |
/* case 2.C: */ |
| 330 |
else if (strncmp(in, "/../", 4) == 0 || strcmp(in, "/..") == 0) { |
| 331 |
char *p; |
| 332 |
|
| 333 |
/* Make the next character in the input buffer a "/": */ |
| 334 |
if (in[3] == '\0') { |
| 335 |
/* terminating "/.." case */ |
| 336 |
in += 2; |
| 337 |
in[0] = '/'; |
| 338 |
} else { |
| 339 |
/* "/../" prefix case */ |
| 340 |
in += 3; |
| 341 |
} |
| 342 |
|
| 343 |
/* Trim the last component from the output buffer, or |
| 344 |
* empty it. */ |
| 345 |
p = strrchr(out, '/'); |
| 346 |
if (p) { |
| 347 |
*p = '\0'; |
| 348 |
} else { |
| 349 |
out[0] = '\0'; |
| 350 |
} |
| 351 |
} |
| 352 |
|
| 353 |
/* case 2.D: */ |
| 354 |
else if (strcmp(in, ".") == 0 || strcmp(in, "..") == 0) { |
| 355 |
in[0] = '\0'; |
| 356 |
} |
| 357 |
|
| 358 |
/* case 2.E */ |
| 359 |
else { |
| 360 |
char *p; |
| 361 |
|
| 362 |
/* Search for the *second* "/" if the leading character is |
| 363 |
* already "/": */ |
| 364 |
p = strchr(in + (in[0] == '/'), '/'); |
| 365 |
/* Otherwise, copy the whole string */ |
| 366 |
if (p == NULL) p = strchr(in, '\0'); |
| 367 |
|
| 368 |
strncat(out, in, p - in); |
| 369 |
in = p; |
| 370 |
} |
| 371 |
} |
| 372 |
|
| 373 |
ne_free(inc); |
| 374 |
|
| 375 |
return out; |
| 376 |
} |
| 377 |
|
| 378 |
/* Copy authority components from 'src' to 'dest' if defined. */ |
| 379 |
static void copy_authority(ne_uri *dest, const ne_uri *src) |
| 380 |
{ |
| 381 |
if (src->host) dest->host = ne_strdup(src->host); |
| 382 |
dest->port = src->port; |
| 383 |
if (src->userinfo) dest->userinfo = ne_strdup(src->userinfo); |
| 384 |
} |
| 385 |
|
| 386 |
/* This function directly implements the "Transform References" |
| 387 |
* algorithm described in RFC 3986 section 5.2.2. */ |
| 388 |
ne_uri *ne_uri_resolve(const ne_uri *base, const ne_uri *relative, |
| 389 |
ne_uri *target) |
| 390 |
{ |
| 391 |
memset(target, 0, sizeof *target); |
| 392 |
|
| 393 |
if (relative->scheme) { |
| 394 |
target->scheme = ne_strdup(relative->scheme); |
| 395 |
copy_authority(target, relative); |
| 396 |
target->path = remove_dot_segments(relative->path); |
| 397 |
if (relative->query) target->query = ne_strdup(relative->query); |
| 398 |
} else { |
| 399 |
if (relative->host) { |
| 400 |
copy_authority(target, relative); |
| 401 |
target->path = remove_dot_segments(relative->path); |
| 402 |
if (relative->query) target->query = ne_strdup(relative->query); |
| 403 |
} else { |
| 404 |
if (relative->path[0] == '\0') { |
| 405 |
target->path = ne_strdup(base->path); |
| 406 |
if (relative->query) { |
| 407 |
target->query = ne_strdup(relative->query); |
| 408 |
} else if (base->query) { |
| 409 |
target->query = ne_strdup(base->query); |
| 410 |
} |
| 411 |
} else { |
| 412 |
if (relative->path[0] == '/') { |
| 413 |
target->path = remove_dot_segments(relative->path); |
| 414 |
} else { |
| 415 |
char *merged = merge_paths(base, relative->path); |
| 416 |
target->path = remove_dot_segments(merged); |
| 417 |
ne_free(merged); |
| 418 |
} |
| 419 |
if (relative->query) target->query = ne_strdup(relative->query); |
| 420 |
} |
| 421 |
copy_authority(target, base); |
| 422 |
} |
| 423 |
if (base->scheme) target->scheme = ne_strdup(base->scheme); |
| 424 |
} |
| 425 |
|
| 426 |
if (relative->fragment) target->fragment = ne_strdup(relative->fragment); |
| 427 |
|
| 428 |
return target; |
| 429 |
} |
| 430 |
|
| 431 |
ne_uri *ne_uri_copy(ne_uri *dest, const ne_uri *src) |
| 432 |
{ |
| 433 |
memset(dest, 0, sizeof *dest); |
| 434 |
|
| 435 |
if (src->scheme) dest->scheme = ne_strdup(src->scheme); |
| 436 |
copy_authority(dest, src); |
| 437 |
if (src->path) dest->path = ne_strdup(src->path); |
| 438 |
if (src->query) dest->query = ne_strdup(src->query); |
| 439 |
if (src->fragment) dest->fragment = ne_strdup(src->fragment); |
| 440 |
|
| 441 |
return dest; |
| 442 |
} |
| 443 |
|
| 444 |
void ne_uri_free(ne_uri *u) |
| 445 |
{ |
| 446 |
if (u->host) ne_free(u->host); |
| 447 |
if (u->path) ne_free(u->path); |
| 448 |
if (u->scheme) ne_free(u->scheme); |
| 449 |
if (u->userinfo) ne_free(u->userinfo); |
| 450 |
if (u->fragment) ne_free(u->fragment); |
| 451 |
if (u->query) ne_free(u->query); |
| 452 |
memset(u, 0, sizeof *u); |
| 453 |
} |
| 454 |
|
| 455 |
char *ne_path_unescape(const char *uri) |
| 456 |
{ |
| 457 |
const char *pnt; |
| 458 |
char *ret, *retpos, buf[5] = { "0x00" }; |
| 459 |
retpos = ret = ne_malloc(strlen(uri) + 1); |
| 460 |
for (pnt = uri; *pnt != '\0'; pnt++) { |
| 461 |
if (*pnt == '%') { |
| 462 |
if (!isxdigit((unsigned char) pnt[1]) || |
| 463 |
!isxdigit((unsigned char) pnt[2])) { |
| 464 |
/* Invalid URI */ |
| 465 |
ne_free(ret); |
| 466 |
return NULL; |
| 467 |
} |
| 468 |
buf[2] = *++pnt; buf[3] = *++pnt; /* bit faster than memcpy */ |
| 469 |
*retpos++ = (char)strtol(buf, NULL, 16); |
| 470 |
} else { |
| 471 |
*retpos++ = *pnt; |
| 472 |
} |
| 473 |
} |
| 474 |
*retpos = '\0'; |
| 475 |
return ret; |
| 476 |
} |
| 477 |
|
| 478 |
/* CH must be an unsigned char; evaluates to 1 if CH should be |
| 479 |
* percent-encoded. */ |
| 480 |
#define path_escape_ch(ch) (uri_lookup(ch) & URI_ESCAPE) |
| 481 |
|
| 482 |
char *ne_path_escape(const char *path) |
| 483 |
{ |
| 484 |
const unsigned char *pnt; |
| 485 |
char *ret, *p; |
| 486 |
size_t count = 0; |
| 487 |
|
| 488 |
for (pnt = (const unsigned char *)path; *pnt != '\0'; pnt++) { |
| 489 |
count += path_escape_ch(*pnt); |
| 490 |
} |
| 491 |
|
| 492 |
if (count == 0) { |
| 493 |
return ne_strdup(path); |
| 494 |
} |
| 495 |
|
| 496 |
p = ret = ne_malloc(strlen(path) + 2 * count + 1); |
| 497 |
for (pnt = (const unsigned char *)path; *pnt != '\0'; pnt++) { |
| 498 |
if (path_escape_ch(*pnt)) { |
| 499 |
/* Escape it - %<hex><hex> */ |
| 500 |
sprintf(p, "%%%02x", (unsigned char) *pnt); |
| 501 |
p += 3; |
| 502 |
} else { |
| 503 |
*p++ = *pnt; |
| 504 |
} |
| 505 |
} |
| 506 |
*p = '\0'; |
| 507 |
return ret; |
| 508 |
} |
| 509 |
|
| 510 |
#undef path_escape_ch |
| 511 |
|
| 512 |
#define CMPWITH(field, func) \ |
| 513 |
do { \ |
| 514 |
if (u1->field) { \ |
| 515 |
if (!u2->field) return -1; \ |
| 516 |
n = func(u1->field, u2->field); \ |
| 517 |
if (n) return n; \ |
| 518 |
} else if (u2->field) { \ |
| 519 |
return 1; \ |
| 520 |
} \ |
| 521 |
} while (0) |
| 522 |
|
| 523 |
#define CMP(field) CMPWITH(field, strcmp) |
| 524 |
#define CASECMP(field) CMPWITH(field, ne_strcasecmp) |
| 525 |
|
| 526 |
/* As specified by RFC 2616, section 3.2.3. */ |
| 527 |
int ne_uri_cmp(const ne_uri *u1, const ne_uri *u2) |
| 528 |
{ |
| 529 |
int n; |
| 530 |
|
| 531 |
CMP(path); |
| 532 |
CASECMP(host); |
| 533 |
CASECMP(scheme); |
| 534 |
CMP(query); |
| 535 |
CMP(fragment); |
| 536 |
CMP(userinfo); |
| 537 |
|
| 538 |
return u2->port - u1->port; |
| 539 |
} |
| 540 |
|
| 541 |
#undef CMP |
| 542 |
#undef CASECMP |
| 543 |
#undef CMPWITH |
| 544 |
|
| 545 |
/* TODO: implement properly */ |
| 546 |
int ne_path_compare(const char *a, const char *b) |
| 547 |
{ |
| 548 |
int ret = ne_strcasecmp(a, b); |
| 549 |
if (ret) { |
| 550 |
/* This logic says: "If the lengths of the two URIs differ by |
| 551 |
* exactly one, and the LONGER of the two URIs has a trailing |
| 552 |
* slash and the SHORTER one DOESN'T, then..." */ |
| 553 |
int traila = ne_path_has_trailing_slash(a), |
| 554 |
trailb = ne_path_has_trailing_slash(b), |
| 555 |
lena = strlen(a), lenb = strlen(b); |
| 556 |
if (traila != trailb && abs(lena - lenb) == 1 && |
| 557 |
((traila && lena > lenb) || (trailb && lenb > lena))) { |
| 558 |
/* Compare them, ignoring the trailing slash on the longer |
| 559 |
* URI */ |
| 560 |
if (strncasecmp(a, b, lena < lenb ? lena : lenb) == 0) |
| 561 |
ret = 0; |
| 562 |
} |
| 563 |
} |
| 564 |
return ret; |
| 565 |
} |
| 566 |
|
| 567 |
char *ne_uri_unparse(const ne_uri *uri) |
| 568 |
{ |
| 569 |
ne_buffer *buf = ne_buffer_create(); |
| 570 |
|
| 571 |
if (uri->scheme) { |
| 572 |
ne_buffer_concat(buf, uri->scheme, ":", NULL); |
| 573 |
} |
| 574 |
|
| 575 |
if (uri->host) { |
| 576 |
ne_buffer_czappend(buf, "//"); |
| 577 |
if (uri->userinfo) { |
| 578 |
ne_buffer_concat(buf, uri->userinfo, "@", NULL); |
| 579 |
} |
| 580 |
ne_buffer_zappend(buf, uri->host); |
| 581 |
|
| 582 |
if (uri->port > 0 |
| 583 |
&& (!uri->scheme |
| 584 |
|| ne_uri_defaultport(uri->scheme) != uri->port)) { |
| 585 |
char str[20]; |
| 586 |
ne_snprintf(str, 20, ":%d", uri->port); |
| 587 |
ne_buffer_zappend(buf, str); |
| 588 |
} |
| 589 |
} |
| 590 |
|
| 591 |
ne_buffer_zappend(buf, uri->path); |
| 592 |
|
| 593 |
if (uri->query) { |
| 594 |
ne_buffer_concat(buf, "?", uri->query, NULL); |
| 595 |
} |
| 596 |
|
| 597 |
if (uri->fragment) { |
| 598 |
ne_buffer_concat(buf, "#", uri->fragment, NULL); |
| 599 |
} |
| 600 |
|
| 601 |
return ne_buffer_finish(buf); |
| 602 |
} |
| 603 |
|
| 604 |
/* Give it a path segment, it returns non-zero if child is |
| 605 |
* a child of parent. */ |
| 606 |
int ne_path_childof(const char *parent, const char *child) |
| 607 |
{ |
| 608 |
char *root = ne_strdup(child); |
| 609 |
int ret; |
| 610 |
if (strlen(parent) >= strlen(child)) { |
| 611 |
ret = 0; |
| 612 |
} else { |
| 613 |
/* root is the first of child, equal to length of parent */ |
| 614 |
root[strlen(parent)] = '\0'; |
| 615 |
ret = (ne_path_compare(parent, root) == 0); |
| 616 |
} |
| 617 |
ne_free(root); |
| 618 |
return ret; |
| 619 |
} |