/src/php-src/ext/uri/uriparser/src/UriCommon.c
Line | Count | Source |
1 | | /* |
2 | | * uriparser - RFC 3986 URI parsing library |
3 | | * |
4 | | * Copyright (C) 2007, Weijia Song <songweijia@gmail.com> |
5 | | * Copyright (C) 2007, Sebastian Pipping <sebastian@pipping.org> |
6 | | * All rights reserved. |
7 | | * |
8 | | * Redistribution and use in source and binary forms, with or without |
9 | | * modification, are permitted provided that the following conditions |
10 | | * are met: |
11 | | * |
12 | | * 1. Redistributions of source code must retain the above |
13 | | * copyright notice, this list of conditions and the following |
14 | | * disclaimer. |
15 | | * |
16 | | * 2. Redistributions in binary form must reproduce the above |
17 | | * copyright notice, this list of conditions and the following |
18 | | * disclaimer in the documentation and/or other materials |
19 | | * provided with the distribution. |
20 | | * |
21 | | * 3. Neither the name of the copyright holder nor the names of |
22 | | * its contributors may be used to endorse or promote products |
23 | | * derived from this software without specific prior written |
24 | | * permission. |
25 | | * |
26 | | * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS |
27 | | * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT |
28 | | * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS |
29 | | * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL |
30 | | * THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, |
31 | | * INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES |
32 | | * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR |
33 | | * SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) |
34 | | * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, |
35 | | * STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) |
36 | | * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED |
37 | | * OF THE POSSIBILITY OF SUCH DAMAGE. |
38 | | */ |
39 | | |
40 | | /* What encodings are enabled? */ |
41 | | #include <uriparser/UriDefsConfig.h> |
42 | | #if (!defined(URI_PASS_ANSI) && !defined(URI_PASS_UNICODE)) |
43 | | /* Include SELF twice */ |
44 | | # ifdef URI_ENABLE_ANSI |
45 | | # define URI_PASS_ANSI 1 |
46 | | # include "UriCommon.c" |
47 | | # undef URI_PASS_ANSI |
48 | | # endif |
49 | | # ifdef URI_ENABLE_UNICODE |
50 | | # define URI_PASS_UNICODE 1 |
51 | | # include "UriCommon.c" |
52 | | # undef URI_PASS_UNICODE |
53 | | # endif |
54 | | #else |
55 | | # ifdef URI_PASS_ANSI |
56 | | # include <uriparser/UriDefsAnsi.h> |
57 | | # else |
58 | | # include <uriparser/UriDefsUnicode.h> |
59 | | # include <wchar.h> |
60 | | # endif |
61 | | |
62 | | # ifndef URI_DOXYGEN |
63 | | # include <uriparser/Uri.h> |
64 | | # include "UriCommon.h" |
65 | | # include "UriSets.h" |
66 | | # endif |
67 | | |
68 | | # include <assert.h> |
69 | | |
70 | | /*extern*/ const URI_CHAR * const URI_FUNC(SafeToPointTo) = _UT("X"); |
71 | | /*extern*/ const URI_CHAR * const URI_FUNC(ConstPwd) = _UT("."); |
72 | | /*extern*/ const URI_CHAR * const URI_FUNC(ConstParent) = _UT(".."); |
73 | | |
74 | 0 | void URI_FUNC(ResetUri)(URI_TYPE(Uri) * uri) { |
75 | 0 | if (uri == NULL) { |
76 | 0 | return; |
77 | 0 | } |
78 | 0 | memset(uri, 0, sizeof(URI_TYPE(Uri))); |
79 | 0 | } Unexecuted instantiation: uriResetUriA Unexecuted instantiation: uriResetUriW |
80 | | |
81 | 0 | int URI_FUNC(FreeUriPath)(URI_TYPE(Uri) * uri, UriMemoryManager * memory) { |
82 | 0 | assert(uri != NULL); |
83 | 0 | assert(memory != NULL); |
84 | |
|
85 | 0 | if (uri->pathHead != NULL) { |
86 | 0 | URI_TYPE(PathSegment) * segWalk = uri->pathHead; |
87 | 0 | while (segWalk != NULL) { |
88 | 0 | URI_TYPE(PathSegment) * const next = segWalk->next; |
89 | 0 | if ((uri->owner == URI_TRUE) |
90 | 0 | && (segWalk->text.first != segWalk->text.afterLast)) { |
91 | 0 | memory->free(memory, (URI_CHAR *)segWalk->text.first); |
92 | 0 | } |
93 | 0 | segWalk->text.first = NULL; |
94 | 0 | segWalk->text.afterLast = NULL; |
95 | 0 | segWalk->next = NULL; |
96 | 0 | memory->free(memory, segWalk); |
97 | 0 | segWalk = next; |
98 | 0 | } |
99 | 0 | uri->pathHead = NULL; |
100 | 0 | uri->pathTail = NULL; |
101 | 0 | } |
102 | |
|
103 | 0 | return URI_SUCCESS; |
104 | 0 | } Unexecuted instantiation: uriFreeUriPathA Unexecuted instantiation: uriFreeUriPathW |
105 | | |
106 | | /* Compares two text ranges for equal text content */ |
107 | 0 | int URI_FUNC(CompareRange)(const URI_TYPE(TextRange) * a, const URI_TYPE(TextRange) * b) { |
108 | 0 | int diff; |
109 | | |
110 | | /* NOTE: Both NULL means equal! */ |
111 | 0 | if ((a == NULL) || (b == NULL)) { |
112 | 0 | return ((a == NULL) ? 0 : 1) - ((b == NULL) ? 0 : 1); |
113 | 0 | } |
114 | | |
115 | | /* NOTE: Both NULL means equal! */ |
116 | 0 | if ((a->first == NULL) || (b->first == NULL)) { |
117 | 0 | return ((a->first == NULL) ? 0 : 1) - ((b->first == NULL) ? 0 : 1); |
118 | 0 | } |
119 | | |
120 | 0 | diff = ((int)(a->afterLast - a->first) - (int)(b->afterLast - b->first)); |
121 | 0 | if (diff > 0) { |
122 | 0 | return 1; |
123 | 0 | } else if (diff < 0) { |
124 | 0 | return -1; |
125 | 0 | } |
126 | | |
127 | 0 | diff = URI_STRNCMP(a->first, b->first, (a->afterLast - a->first)); |
128 | |
|
129 | 0 | if (diff > 0) { |
130 | 0 | return 1; |
131 | 0 | } else if (diff < 0) { |
132 | 0 | return -1; |
133 | 0 | } |
134 | | |
135 | 0 | return diff; |
136 | 0 | } Unexecuted instantiation: uriCompareRangeA Unexecuted instantiation: uriCompareRangeW |
137 | | |
138 | | UriBool URI_FUNC(CopyRange)(URI_TYPE(TextRange) * destRange, |
139 | | const URI_TYPE(TextRange) * sourceRange, |
140 | 0 | UriMemoryManager * memory) { |
141 | 0 | const int lenInChars = (int)(sourceRange->afterLast - sourceRange->first); |
142 | 0 | const int lenInBytes = lenInChars * sizeof(URI_CHAR); |
143 | 0 | URI_CHAR * dup = memory->malloc(memory, lenInBytes); |
144 | 0 | if (dup == NULL) { |
145 | 0 | return URI_FALSE; |
146 | 0 | } |
147 | 0 | memcpy(dup, sourceRange->first, lenInBytes); |
148 | 0 | destRange->first = dup; |
149 | 0 | destRange->afterLast = dup + lenInChars; |
150 | |
|
151 | 0 | return URI_TRUE; |
152 | 0 | } Unexecuted instantiation: uriCopyRangeA Unexecuted instantiation: uriCopyRangeW |
153 | | |
154 | | UriBool URI_FUNC(CopyRangeAsNeeded)(URI_TYPE(TextRange) * destRange, |
155 | | const URI_TYPE(TextRange) * sourceRange, |
156 | 0 | UriMemoryManager * memory) { |
157 | 0 | if (sourceRange->first == NULL) { |
158 | 0 | destRange->first = NULL; |
159 | 0 | destRange->afterLast = NULL; |
160 | 0 | } else if (sourceRange->first == sourceRange->afterLast) { |
161 | 0 | destRange->first = URI_FUNC(SafeToPointTo); |
162 | 0 | destRange->afterLast = URI_FUNC(SafeToPointTo); |
163 | 0 | } else { |
164 | 0 | return URI_FUNC(CopyRange)(destRange, sourceRange, memory); |
165 | 0 | } |
166 | | |
167 | 0 | return URI_TRUE; |
168 | 0 | } Unexecuted instantiation: uriCopyRangeAsNeededA Unexecuted instantiation: uriCopyRangeAsNeededW |
169 | | |
170 | | UriBool URI_FUNC(RemoveDotSegmentsEx)(URI_TYPE(Uri) * uri, UriBool relative, |
171 | 0 | UriBool pathOwned, UriMemoryManager * memory) { |
172 | 0 | URI_TYPE(PathSegment) * walker; |
173 | 0 | if ((uri == NULL) || (uri->pathHead == NULL)) { |
174 | 0 | return URI_TRUE; |
175 | 0 | } |
176 | | |
177 | 0 | walker = uri->pathHead; |
178 | 0 | walker->reserved = NULL; /* Prev pointer */ |
179 | 0 | do { |
180 | 0 | UriBool removeSegment = URI_FALSE; |
181 | 0 | int len = (int)(walker->text.afterLast - walker->text.first); |
182 | 0 | switch (len) { |
183 | 0 | case 1: |
184 | 0 | if ((walker->text.first)[0] == _UT('.')) { |
185 | | /* "." segment -> remove if not essential */ |
186 | 0 | URI_TYPE(PathSegment) * const prev = walker->reserved; |
187 | 0 | URI_TYPE(PathSegment) * const nextBackup = walker->next; |
188 | | |
189 | | /* |
190 | | * Is this dot segment essential, |
191 | | * i.e. is there a chance of changing semantics by dropping this dot |
192 | | * segment? |
193 | | * |
194 | | * For example, changing "./http://foo" into "http://foo" would change |
195 | | * semantics and hence the dot segment is essential to that case and |
196 | | * cannot be removed. |
197 | | * |
198 | | * Other examples that would change semantics are: |
199 | | * - cutting "/.//" down to "//" |
200 | | * - cutting "scheme:/.//" down to "scheme://". |
201 | | */ |
202 | 0 | removeSegment = URI_TRUE; |
203 | 0 | if ((walker == uri->pathHead) && (walker->next != NULL)) { |
204 | | /* Detect case "/.//" (with or without scheme) */ |
205 | 0 | if ((walker->next->text.first == walker->next->text.afterLast) |
206 | 0 | && (URI_FUNC(HasHost)(uri) == URI_FALSE)) { |
207 | 0 | removeSegment = URI_FALSE; |
208 | | /* Detect case "./withcolon:" */ |
209 | 0 | } else if (relative) { |
210 | 0 | const URI_CHAR * ch = walker->next->text.first; |
211 | 0 | for (; ch < walker->next->text.afterLast; ch++) { |
212 | 0 | if (*ch == _UT(':')) { |
213 | 0 | removeSegment = URI_FALSE; |
214 | 0 | break; |
215 | 0 | } |
216 | 0 | } |
217 | 0 | } |
218 | 0 | } |
219 | |
|
220 | 0 | if (removeSegment) { |
221 | | /* .. then let's go remove that segment. */ |
222 | | /* Last segment? */ |
223 | 0 | if (walker->next != NULL) { |
224 | | /* Not last segment, i.e. first or middle segment |
225 | | * OLD: (prev|NULL) <- walker <- next |
226 | | * NEW: (prev|NULL) <----------- next */ |
227 | 0 | walker->next->reserved = prev; |
228 | |
|
229 | 0 | if (prev == NULL) { |
230 | | /* First but not last segment |
231 | | * OLD: head -> walker -> next |
232 | | * NEW: head -----------> next */ |
233 | 0 | uri->pathHead = walker->next; |
234 | 0 | } else { |
235 | | /* Middle segment |
236 | | * OLD: prev -> walker -> next |
237 | | * NEW: prev -----------> next */ |
238 | 0 | prev->next = walker->next; |
239 | 0 | } |
240 | |
|
241 | 0 | if (pathOwned && (walker->text.first != walker->text.afterLast)) { |
242 | 0 | memory->free(memory, (URI_CHAR *)walker->text.first); |
243 | 0 | } |
244 | 0 | memory->free(memory, walker); |
245 | 0 | } else { |
246 | | /* Last segment */ |
247 | 0 | if (pathOwned && (walker->text.first != walker->text.afterLast)) { |
248 | 0 | memory->free(memory, (URI_CHAR *)walker->text.first); |
249 | 0 | } |
250 | |
|
251 | 0 | if (prev == NULL) { |
252 | | /* Last and first */ |
253 | 0 | if (URI_FUNC(HasHost)(uri)) { |
254 | | /* Replace "." with empty segment to represent trailing |
255 | | * slash */ |
256 | 0 | walker->text.first = URI_FUNC(SafeToPointTo); |
257 | 0 | walker->text.afterLast = URI_FUNC(SafeToPointTo); |
258 | 0 | } else { |
259 | 0 | memory->free(memory, walker); |
260 | |
|
261 | 0 | uri->pathHead = NULL; |
262 | 0 | uri->pathTail = NULL; |
263 | 0 | } |
264 | 0 | } else { |
265 | | /* Last but not first, replace "." with empty segment to |
266 | | * represent trailing slash */ |
267 | 0 | walker->text.first = URI_FUNC(SafeToPointTo); |
268 | 0 | walker->text.afterLast = URI_FUNC(SafeToPointTo); |
269 | 0 | } |
270 | 0 | } |
271 | |
|
272 | 0 | walker = nextBackup; |
273 | 0 | } |
274 | 0 | } |
275 | 0 | break; |
276 | | |
277 | 0 | case 2: |
278 | 0 | if (((walker->text.first)[0] == _UT('.')) |
279 | 0 | && ((walker->text.first)[1] == _UT('.'))) { |
280 | | /* Path ".." -> remove this and the previous segment */ |
281 | 0 | URI_TYPE(PathSegment) * const prev = walker->reserved; |
282 | 0 | URI_TYPE(PathSegment) * prevPrev; |
283 | 0 | URI_TYPE(PathSegment) * const nextBackup = walker->next; |
284 | |
|
285 | 0 | removeSegment = URI_TRUE; |
286 | 0 | if (relative) { |
287 | 0 | if (prev == NULL) { |
288 | | /* We cannot remove traversal beyond because the |
289 | | * URI is relative and may be resolved later. |
290 | | * So we can simplify "a/../b/d" to "b/d" but |
291 | | * we cannot simplify "../b/d" (outside of reference resolution). |
292 | | */ |
293 | 0 | removeSegment = URI_FALSE; |
294 | 0 | } else if ((prev != NULL) |
295 | 0 | && ((prev->text.afterLast - prev->text.first) == 2) |
296 | 0 | && ((prev->text.first)[0] == _UT('.')) |
297 | 0 | && ((prev->text.first)[1] == _UT('.'))) { |
298 | | /* We need to protect against mis-simplifying "a/../../b" to |
299 | | * "a/b". */ |
300 | 0 | removeSegment = URI_FALSE; |
301 | 0 | } |
302 | 0 | } |
303 | |
|
304 | 0 | if (removeSegment) { |
305 | 0 | if (prev != NULL) { |
306 | | /* Not first segment */ |
307 | 0 | prevPrev = prev->reserved; |
308 | 0 | if (prevPrev != NULL) { |
309 | | /* Not even prev is the first one |
310 | | * OLD: prevPrev -> prev -> walker -> (next|NULL) |
311 | | * NEW: prevPrev -------------------> (next|NULL) */ |
312 | 0 | prevPrev->next = walker->next; |
313 | 0 | if (walker->next != NULL) { |
314 | | /* Update parent relationship as well |
315 | | * OLD: prevPrev <- prev <- walker <- next |
316 | | * NEW: prevPrev <------------------- next */ |
317 | 0 | walker->next->reserved = prevPrev; |
318 | 0 | } else { |
319 | | /* Last segment -> insert "" segment to represent trailing |
320 | | * slash, update tail */ |
321 | 0 | URI_TYPE(PathSegment) * const segment = memory->calloc( |
322 | 0 | memory, 1, sizeof(URI_TYPE(PathSegment))); |
323 | 0 | if (segment == NULL) { |
324 | 0 | if (pathOwned |
325 | 0 | && (walker->text.first |
326 | 0 | != walker->text.afterLast)) { |
327 | 0 | memory->free(memory, |
328 | 0 | (URI_CHAR *)walker->text.first); |
329 | 0 | } |
330 | 0 | memory->free(memory, walker); |
331 | |
|
332 | 0 | if (pathOwned |
333 | 0 | && (prev->text.first != prev->text.afterLast)) { |
334 | 0 | memory->free(memory, |
335 | 0 | (URI_CHAR *)prev->text.first); |
336 | 0 | } |
337 | 0 | memory->free(memory, prev); |
338 | |
|
339 | 0 | return URI_FALSE; /* Raises malloc error */ |
340 | 0 | } |
341 | 0 | segment->text.first = URI_FUNC(SafeToPointTo); |
342 | 0 | segment->text.afterLast = URI_FUNC(SafeToPointTo); |
343 | 0 | prevPrev->next = segment; |
344 | 0 | uri->pathTail = segment; |
345 | 0 | } |
346 | | |
347 | 0 | if (pathOwned |
348 | 0 | && (walker->text.first != walker->text.afterLast)) { |
349 | 0 | memory->free(memory, (URI_CHAR *)walker->text.first); |
350 | 0 | } |
351 | 0 | memory->free(memory, walker); |
352 | |
|
353 | 0 | if (pathOwned && (prev->text.first != prev->text.afterLast)) { |
354 | 0 | memory->free(memory, (URI_CHAR *)prev->text.first); |
355 | 0 | } |
356 | 0 | memory->free(memory, prev); |
357 | |
|
358 | 0 | walker = nextBackup; |
359 | 0 | } else { |
360 | | /* Prev is the first segment */ |
361 | 0 | if (walker->next != NULL) { |
362 | 0 | uri->pathHead = walker->next; |
363 | 0 | walker->next->reserved = NULL; |
364 | |
|
365 | 0 | if (pathOwned |
366 | 0 | && (walker->text.first != walker->text.afterLast)) { |
367 | 0 | memory->free(memory, (URI_CHAR *)walker->text.first); |
368 | 0 | } |
369 | 0 | memory->free(memory, walker); |
370 | 0 | } else { |
371 | | /* Reuse segment for "" path segment to represent trailing |
372 | | * slash, update tail */ |
373 | 0 | URI_TYPE(PathSegment) * const segment = walker; |
374 | 0 | if (pathOwned |
375 | 0 | && (segment->text.first != segment->text.afterLast)) { |
376 | 0 | memory->free(memory, (URI_CHAR *)segment->text.first); |
377 | 0 | } |
378 | 0 | segment->text.first = URI_FUNC(SafeToPointTo); |
379 | 0 | segment->text.afterLast = URI_FUNC(SafeToPointTo); |
380 | 0 | uri->pathHead = segment; |
381 | 0 | uri->pathTail = segment; |
382 | 0 | } |
383 | |
|
384 | 0 | if (pathOwned && (prev->text.first != prev->text.afterLast)) { |
385 | 0 | memory->free(memory, (URI_CHAR *)prev->text.first); |
386 | 0 | } |
387 | 0 | memory->free(memory, prev); |
388 | |
|
389 | 0 | walker = nextBackup; |
390 | 0 | } |
391 | 0 | } else { |
392 | 0 | URI_TYPE(PathSegment) * const anotherNextBackup = walker->next; |
393 | 0 | int freeWalker = URI_TRUE; |
394 | | |
395 | | /* First segment */ |
396 | 0 | if (walker->next != NULL) { |
397 | | /* First segment of multiple -> update head |
398 | | * OLD: head -> walker -> next |
399 | | * NEW: head -----------> next */ |
400 | 0 | uri->pathHead = walker->next; |
401 | | |
402 | | /* Update parent link as well |
403 | | * OLD: head <- walker <- next |
404 | | * NEW: head <----------- next */ |
405 | 0 | walker->next->reserved = NULL; |
406 | 0 | } else { |
407 | 0 | if (uri->absolutePath) { |
408 | | /* First and only segment -> update head |
409 | | * OLD: head -> walker -> NULL |
410 | | * NEW: head -----------> NULL */ |
411 | 0 | uri->pathHead = NULL; |
412 | | |
413 | | /* Last segment -> update tail |
414 | | * OLD: tail -> walker |
415 | | * NEW: tail -> NULL */ |
416 | 0 | uri->pathTail = NULL; |
417 | 0 | } else { |
418 | | /* Reuse segment for "" path segment to represent trailing |
419 | | * slash, then update head and tail */ |
420 | 0 | if (pathOwned |
421 | 0 | && (walker->text.first != walker->text.afterLast)) { |
422 | 0 | memory->free(memory, (URI_CHAR *)walker->text.first); |
423 | 0 | } |
424 | 0 | walker->text.first = URI_FUNC(SafeToPointTo); |
425 | 0 | walker->text.afterLast = URI_FUNC(SafeToPointTo); |
426 | 0 | freeWalker = URI_FALSE; |
427 | 0 | } |
428 | 0 | } |
429 | |
|
430 | 0 | if (freeWalker) { |
431 | 0 | if (pathOwned |
432 | 0 | && (walker->text.first != walker->text.afterLast)) { |
433 | 0 | memory->free(memory, (URI_CHAR *)walker->text.first); |
434 | 0 | } |
435 | 0 | memory->free(memory, walker); |
436 | 0 | } |
437 | |
|
438 | 0 | walker = anotherNextBackup; |
439 | 0 | } |
440 | 0 | } |
441 | 0 | } |
442 | 0 | break; |
443 | 0 | } /* end of switch */ |
444 | | |
445 | 0 | if (!removeSegment) { |
446 | | /* .. then let's move to the next element, and start again. */ |
447 | 0 | if (walker->next != NULL) { |
448 | 0 | walker->next->reserved = walker; |
449 | 0 | } else { |
450 | | /* Last segment -> update tail */ |
451 | 0 | uri->pathTail = walker; |
452 | 0 | } |
453 | 0 | walker = walker->next; |
454 | 0 | } |
455 | 0 | } while (walker != NULL); |
456 | | |
457 | 0 | return URI_TRUE; |
458 | 0 | } Unexecuted instantiation: uriRemoveDotSegmentsExA Unexecuted instantiation: uriRemoveDotSegmentsExW |
459 | | |
460 | | /* Properly removes "." and ".." path segments */ |
461 | | UriBool URI_FUNC(RemoveDotSegmentsAbsolute)(URI_TYPE(Uri) * uri, |
462 | 0 | UriMemoryManager * memory) { |
463 | 0 | const UriBool ABSOLUTE = URI_FALSE; |
464 | 0 | if (uri == NULL) { |
465 | 0 | return URI_TRUE; |
466 | 0 | } |
467 | 0 | return URI_FUNC(RemoveDotSegmentsEx)(uri, ABSOLUTE, uri->owner, memory); |
468 | 0 | } Unexecuted instantiation: uriRemoveDotSegmentsAbsoluteA Unexecuted instantiation: uriRemoveDotSegmentsAbsoluteW |
469 | | |
470 | 0 | unsigned char URI_FUNC(HexdigToInt)(URI_CHAR hexdig) { |
471 | 0 | switch (hexdig) { |
472 | 0 | case URI_SET_DIGIT(_UT): |
473 | 0 | return (unsigned char)(9 + hexdig - _UT('9')); |
474 | 0 | case URI_SET_HEX_LETTER_LOWER(_UT): |
475 | 0 | return (unsigned char)(15 + hexdig - _UT('f')); |
476 | 0 | case URI_SET_HEX_LETTER_UPPER(_UT): |
477 | 0 | return (unsigned char)(15 + hexdig - _UT('F')); |
478 | | |
479 | 0 | default: |
480 | 0 | return 0; |
481 | 0 | } |
482 | 0 | } Unexecuted instantiation: uriHexdigToIntA Unexecuted instantiation: uriHexdigToIntW |
483 | | |
484 | 0 | URI_CHAR URI_FUNC(HexToLetterEx)(unsigned int value, UriBool uppercase) { |
485 | 0 | switch (value) { |
486 | 0 | case 0: |
487 | 0 | return _UT('0'); |
488 | 0 | case 1: |
489 | 0 | return _UT('1'); |
490 | 0 | case 2: |
491 | 0 | return _UT('2'); |
492 | 0 | case 3: |
493 | 0 | return _UT('3'); |
494 | 0 | case 4: |
495 | 0 | return _UT('4'); |
496 | 0 | case 5: |
497 | 0 | return _UT('5'); |
498 | 0 | case 6: |
499 | 0 | return _UT('6'); |
500 | 0 | case 7: |
501 | 0 | return _UT('7'); |
502 | 0 | case 8: |
503 | 0 | return _UT('8'); |
504 | 0 | case 9: |
505 | 0 | return _UT('9'); |
506 | | |
507 | 0 | case 10: |
508 | 0 | return (uppercase == URI_TRUE) ? _UT('A') : _UT('a'); |
509 | 0 | case 11: |
510 | 0 | return (uppercase == URI_TRUE) ? _UT('B') : _UT('b'); |
511 | 0 | case 12: |
512 | 0 | return (uppercase == URI_TRUE) ? _UT('C') : _UT('c'); |
513 | 0 | case 13: |
514 | 0 | return (uppercase == URI_TRUE) ? _UT('D') : _UT('d'); |
515 | 0 | case 14: |
516 | 0 | return (uppercase == URI_TRUE) ? _UT('E') : _UT('e'); |
517 | 0 | default: |
518 | 0 | return (uppercase == URI_TRUE) ? _UT('F') : _UT('f'); |
519 | 0 | } |
520 | 0 | } Unexecuted instantiation: uriHexToLetterExA Unexecuted instantiation: uriHexToLetterExW |
521 | | |
522 | | /* Checks if a URI has the host component set. */ |
523 | 0 | UriBool URI_FUNC(HasHost)(const URI_TYPE(Uri) * uri) { |
524 | | /* NOTE: .hostData.ipFuture.first is not being checked, * |
525 | | * because we do check .hostText.first and * |
526 | | * .hostData.ipFuture.first has to be identical to * |
527 | | * .hostText.first if set, and hence there is * |
528 | | * no more information to be gained. */ |
529 | 0 | return (uri != NULL) |
530 | 0 | && ((uri->hostText.first != NULL) || (uri->hostData.ip4 != NULL) |
531 | 0 | || (uri->hostData.ip6 != NULL)); |
532 | 0 | } Unexecuted instantiation: uriHasHostA Unexecuted instantiation: uriHasHostW |
533 | | |
534 | | /* Copies the path segment list from one URI to another. */ |
535 | | UriBool URI_FUNC(CopyPath)(URI_TYPE(Uri) * dest, const URI_TYPE(Uri) * source, |
536 | 0 | UriMemoryManager * memory) { |
537 | 0 | if (source->pathHead == NULL) { |
538 | | /* No path component */ |
539 | 0 | dest->pathHead = NULL; |
540 | 0 | dest->pathTail = NULL; |
541 | 0 | } else { |
542 | | /* Copy list but not the text contained */ |
543 | 0 | URI_TYPE(PathSegment) * sourceWalker = source->pathHead; |
544 | 0 | URI_TYPE(PathSegment) * destPrev = NULL; |
545 | 0 | do { |
546 | 0 | URI_TYPE(PathSegment) * cur = |
547 | 0 | memory->malloc(memory, sizeof(URI_TYPE(PathSegment))); |
548 | 0 | if (cur == NULL) { |
549 | | /* Fix broken list */ |
550 | 0 | if (destPrev != NULL) { |
551 | 0 | destPrev->next = NULL; |
552 | 0 | } |
553 | 0 | return URI_FALSE; /* Raises malloc error */ |
554 | 0 | } |
555 | | |
556 | | /* From this functions usage we know that * |
557 | | * the dest URI cannot be uri->owner */ |
558 | 0 | cur->text = sourceWalker->text; |
559 | 0 | if (destPrev == NULL) { |
560 | | /* First segment ever */ |
561 | 0 | dest->pathHead = cur; |
562 | 0 | } else { |
563 | 0 | destPrev->next = cur; |
564 | 0 | } |
565 | 0 | destPrev = cur; |
566 | 0 | sourceWalker = sourceWalker->next; |
567 | 0 | } while (sourceWalker != NULL); |
568 | 0 | dest->pathTail = destPrev; |
569 | 0 | dest->pathTail->next = NULL; |
570 | 0 | } |
571 | | |
572 | 0 | dest->absolutePath = source->absolutePath; |
573 | 0 | return URI_TRUE; |
574 | 0 | } Unexecuted instantiation: uriCopyPathA Unexecuted instantiation: uriCopyPathW |
575 | | |
576 | | /* Copies the authority part of an URI over to another. */ |
577 | | UriBool URI_FUNC(CopyAuthority)(URI_TYPE(Uri) * dest, const URI_TYPE(Uri) * source, |
578 | 0 | UriMemoryManager * memory) { |
579 | | /* From this functions usage we know that * |
580 | | * the dest URI cannot be uri->owner */ |
581 | | |
582 | | /* Copy userInfo */ |
583 | 0 | dest->userInfo = source->userInfo; |
584 | | |
585 | | /* Copy hostText */ |
586 | 0 | dest->hostText = source->hostText; |
587 | | |
588 | | /* Copy hostData */ |
589 | 0 | if (source->hostData.ip4 != NULL) { |
590 | 0 | dest->hostData.ip4 = memory->malloc(memory, sizeof(UriIp4)); |
591 | 0 | if (dest->hostData.ip4 == NULL) { |
592 | 0 | return URI_FALSE; /* Raises malloc error */ |
593 | 0 | } |
594 | 0 | *(dest->hostData.ip4) = *(source->hostData.ip4); |
595 | 0 | dest->hostData.ip6 = NULL; |
596 | 0 | dest->hostData.ipFuture.first = NULL; |
597 | 0 | dest->hostData.ipFuture.afterLast = NULL; |
598 | 0 | } else if (source->hostData.ip6 != NULL) { |
599 | 0 | dest->hostData.ip4 = NULL; |
600 | 0 | dest->hostData.ip6 = memory->malloc(memory, sizeof(UriIp6)); |
601 | 0 | if (dest->hostData.ip6 == NULL) { |
602 | 0 | return URI_FALSE; /* Raises malloc error */ |
603 | 0 | } |
604 | 0 | *(dest->hostData.ip6) = *(source->hostData.ip6); |
605 | 0 | dest->hostData.ipFuture.first = NULL; |
606 | 0 | dest->hostData.ipFuture.afterLast = NULL; |
607 | 0 | } else { |
608 | 0 | dest->hostData.ip4 = NULL; |
609 | 0 | dest->hostData.ip6 = NULL; |
610 | 0 | dest->hostData.ipFuture = source->hostData.ipFuture; |
611 | 0 | } |
612 | | |
613 | | /* Copy portText */ |
614 | 0 | dest->portText = source->portText; |
615 | |
|
616 | 0 | return URI_TRUE; |
617 | 0 | } Unexecuted instantiation: uriCopyAuthorityA Unexecuted instantiation: uriCopyAuthorityW |
618 | | |
619 | 0 | UriBool URI_FUNC(FixAmbiguity)(URI_TYPE(Uri) * uri, UriMemoryManager * memory) { |
620 | 0 | URI_TYPE(PathSegment) * segment; |
621 | |
|
622 | 0 | if (/* Case 1: absolute path, empty first segment */ |
623 | 0 | (uri->absolutePath && (uri->pathHead != NULL) |
624 | 0 | && (uri->pathHead->text.afterLast == uri->pathHead->text.first)) |
625 | | |
626 | | /* Case 2: relative path, empty first and second segment */ |
627 | 0 | || (!uri->absolutePath && (uri->pathHead != NULL) && (uri->pathHead->next != NULL) |
628 | 0 | && (uri->pathHead->text.afterLast == uri->pathHead->text.first) |
629 | 0 | && (uri->pathHead->next->text.afterLast |
630 | 0 | == uri->pathHead->next->text.first))) { |
631 | | /* NOOP */ |
632 | 0 | } else { |
633 | 0 | return URI_TRUE; |
634 | 0 | } |
635 | | |
636 | 0 | segment = memory->malloc(memory, 1 * sizeof(URI_TYPE(PathSegment))); |
637 | 0 | if (segment == NULL) { |
638 | 0 | return URI_FALSE; /* Raises malloc error */ |
639 | 0 | } |
640 | | |
641 | | /* Insert "." segment in front */ |
642 | 0 | segment->next = uri->pathHead; |
643 | 0 | segment->text.first = URI_FUNC(ConstPwd); |
644 | 0 | segment->text.afterLast = URI_FUNC(ConstPwd) + 1; |
645 | 0 | uri->pathHead = segment; |
646 | 0 | return URI_TRUE; |
647 | 0 | } Unexecuted instantiation: uriFixAmbiguityA Unexecuted instantiation: uriFixAmbiguityW |
648 | | |
649 | | static UriBool URI_FUNC(PrependNewDotSegment)(URI_TYPE(Uri) * uri, |
650 | 0 | UriMemoryManager * memory) { |
651 | 0 | assert(uri != NULL); |
652 | 0 | assert(memory != NULL); |
653 | |
|
654 | 0 | URI_TYPE(PathSegment) * const segment = |
655 | 0 | memory->malloc(memory, 1 * sizeof(URI_TYPE(PathSegment))); |
656 | |
|
657 | 0 | if (segment == NULL) { |
658 | 0 | return URI_FALSE; /* i.e. raise malloc error */ |
659 | 0 | } |
660 | | |
661 | 0 | segment->next = uri->pathHead; |
662 | |
|
663 | 0 | URI_TYPE(TextRange) dotRange; |
664 | 0 | dotRange.first = URI_FUNC(ConstPwd); |
665 | 0 | dotRange.afterLast = URI_FUNC(ConstPwd) + 1; |
666 | |
|
667 | 0 | if (uri->owner == URI_TRUE) { |
668 | 0 | if (URI_FUNC(CopyRange)(&(segment->text), &dotRange, memory) == URI_FALSE) { |
669 | 0 | memory->free(memory, segment); |
670 | 0 | return URI_FALSE; /* i.e. raise malloc error */ |
671 | 0 | } |
672 | 0 | } else { |
673 | 0 | segment->text = dotRange; /* copies all members */ |
674 | 0 | } |
675 | | |
676 | 0 | uri->pathHead = segment; |
677 | |
|
678 | 0 | return URI_TRUE; |
679 | 0 | } Unexecuted instantiation: UriCommon.c:uriPrependNewDotSegmentA Unexecuted instantiation: UriCommon.c:uriPrependNewDotSegmentW |
680 | | |
681 | | /* When dropping a scheme from a URI without a host and with a colon (":") |
682 | | * in the first path segment, a consecutive reparse would rightfully |
683 | | * mis-classify the first path segment as a scheme due to the colon. |
684 | | * To protect against this case, we prepend an artificial "." segment |
685 | | * to the path in here; the function is called after the scheme has |
686 | | * just been dropped. |
687 | | * |
688 | | * 0. We start with parsed URI "scheme:path1:/path2/path3". |
689 | | * 1. We drop the scheme naively and yield "path1:/path2/path3". |
690 | | * 2. We prepend "." and yield unambiguous "./path1:/path2/path3". |
691 | | * |
692 | | * From the view of the RFC 3986 grammar, this is replacing rule path-rootless |
693 | | * by path-noscheme content. |
694 | | * |
695 | | * Returns URI_TRUE for (a) nothing to do or (b) successful changes. |
696 | | * Returns URI_FALSE to signal out-of-memory. |
697 | | */ |
698 | 0 | UriBool URI_FUNC(FixPathNoScheme)(URI_TYPE(Uri) * uri, UriMemoryManager * memory) { |
699 | 0 | assert(uri != NULL); |
700 | 0 | assert(memory != NULL); |
701 | |
|
702 | 0 | if ((uri->absolutePath == URI_TRUE) || (uri->pathHead == NULL) |
703 | 0 | || (uri->scheme.first != NULL) || URI_FUNC(HasHost)(uri)) { |
704 | 0 | return URI_TRUE; /* i.e. nothing to do */ |
705 | 0 | } |
706 | | |
707 | | /* Check for troublesome first path segment containing a colon */ |
708 | 0 | UriBool colonFound = URI_FALSE; |
709 | 0 | const URI_CHAR * walker = uri->pathHead->text.first; |
710 | |
|
711 | 0 | while (walker < uri->pathHead->text.afterLast) { |
712 | 0 | if (walker[0] == _UT(':')) { |
713 | 0 | colonFound = URI_TRUE; |
714 | 0 | break; |
715 | 0 | } |
716 | 0 | walker++; |
717 | 0 | } |
718 | |
|
719 | 0 | assert((walker == uri->pathHead->text.afterLast) || (colonFound == URI_TRUE)); |
720 | |
|
721 | 0 | if (colonFound == URI_FALSE) { |
722 | 0 | return URI_TRUE; /* i.e. nothing to do */ |
723 | 0 | } |
724 | | |
725 | | /* Insert "." segment in front */ |
726 | 0 | return URI_FUNC(PrependNewDotSegment)(uri, memory); |
727 | 0 | } Unexecuted instantiation: uriFixPathNoSchemeA Unexecuted instantiation: uriFixPathNoSchemeW |
728 | | |
729 | | /* When dropping a host from a URI without a scheme, an absolute path |
730 | | * and and empty first path segment, a consecutive reparse would rightfully |
731 | | * mis-classify the first path segment as a host marker due to the "//". |
732 | | * To protect against this case, we prepend an artificial "." segment |
733 | | * to the path in here; the function is called after the host has |
734 | | * just been dropped. |
735 | | * |
736 | | * 0. We start with parsed URI "//host//path1/path2". |
737 | | * 1. We drop the host naively and yield "//path1/path2". |
738 | | * 2. We insert "./" and yield unambiguous "/.//path1/path2". |
739 | | * |
740 | | * Returns URI_TRUE for (a) nothing to do or (b) successful changes. |
741 | | * Returns URI_FALSE to signal out-of-memory. |
742 | | */ |
743 | | UriBool URI_FUNC(EnsureThatPathIsNotMistakenForHost)(URI_TYPE(Uri) * uri, |
744 | 0 | UriMemoryManager * memory) { |
745 | 0 | assert(uri != NULL); |
746 | 0 | assert(memory != NULL); |
747 | |
|
748 | 0 | if ((URI_FUNC(HasHost)(uri) == URI_TRUE) || (uri->absolutePath == URI_FALSE) |
749 | 0 | || (uri->pathHead == NULL) |
750 | 0 | || (uri->pathHead == uri->pathTail) /* i.e. no second slash */ |
751 | 0 | || (uri->pathHead->text.first != uri->pathHead->text.afterLast)) { |
752 | 0 | return URI_TRUE; /* i.e. nothing to do */ |
753 | 0 | } |
754 | | |
755 | | /* Insert "." segment in front */ |
756 | 0 | return URI_FUNC(PrependNewDotSegment)(uri, memory); |
757 | 0 | } Unexecuted instantiation: uriEnsureThatPathIsNotMistakenForHostA Unexecuted instantiation: uriEnsureThatPathIsNotMistakenForHostW |
758 | | |
759 | 0 | void URI_FUNC(FixEmptyTrailSegment)(URI_TYPE(Uri) * uri, UriMemoryManager * memory) { |
760 | | /* Fix path if only one empty segment */ |
761 | 0 | if (!uri->absolutePath && !URI_FUNC(HasHost)(uri) && (uri->pathHead != NULL) |
762 | 0 | && (uri->pathHead->next == NULL) |
763 | 0 | && (uri->pathHead->text.first == uri->pathHead->text.afterLast)) { |
764 | 0 | memory->free(memory, uri->pathHead); |
765 | 0 | uri->pathHead = NULL; |
766 | | uri->pathTail = NULL; |
767 | 0 | } |
768 | 0 | } Unexecuted instantiation: uriFixEmptyTrailSegmentA Unexecuted instantiation: uriFixEmptyTrailSegmentW |
769 | | |
770 | | #endif |