uri.cpp@ 57723

Last change on this file since 57723 was 57723, checked in by vboxsync, 10 years ago
RTUri: Reworked the API to a one time parsing and make sure we get UTF-8 clean strings from it.
Property svn:eol-style set to `native` Property svn:keywords set to `Author Date Id Revision`
File size: 25.6 KB

Line
1	/* $Id: uri.cpp 57723 2015-09-11 20:02:15Z vboxsync $ */
2	/** @file
3	* IPRT - Uniform Resource Identifier handling.
4	*/
5
6	/*
7	* Copyright (C) 2011-2015 Oracle Corporation
8	*
9	* This file is part of VirtualBox Open Source Edition (OSE), as
10	* available from http://www.215389.xyz. This file is free software;
11	* you can redistribute it and/or modify it under the terms of the GNU
12	* General Public License (GPL) as published by the Free Software
13	* Foundation, in version 2 as it comes in the "COPYING" file of the
14	* VirtualBox OSE distribution. VirtualBox OSE is distributed in the
15	* hope that it will be useful, but WITHOUT ANY WARRANTY of any kind.
16	*
17	* The contents of this file may alternatively be used under the terms
18	* of the Common Development and Distribution License Version 1.0
19	* (CDDL) only, as it comes in the "COPYING.CDDL" file of the
20	* VirtualBox OSE distribution, in which case the provisions of the
21	* CDDL are applicable instead of those of the GPL.
22	*
23	* You may elect to license modified versions of this file under the
24	* terms and conditions of either the GPL or the CDDL or both.
25	*/
26
27
28	/*********************************************************************************************************************************
29	* Header Files *
30	*********************************************************************************************************************************/
31	#include <iprt/uri.h>
32
33	#include <iprt/assert.h>
34	#include <iprt/ctype.h>
35	#include <iprt/path.h>
36	#include <iprt/string.h>
37
38
39	/*********************************************************************************************************************************
40	* Defined Constants And Macros *
41	*********************************************************************************************************************************/
42	/** Internal magic value we use to check if a RTURIPARSED structure has made it thru RTUriParse. */
43	#define RTURIPARSED_MAGIC UINT32_C(0x439e0745)
44
45
46	/* General URI format:
47
48	foo://example.com:8042/over/there?name=ferret#nose
49	\_/ \______________/\_________/ \_________/ \__/
50	\| \| \| \| \|
51	scheme authority path query fragment
52	\| _____________________\|__
53	/ \ / \
54	urn:example:animal:ferret:nose
55	*/
56
57
58	/**
59	* The following defines characters which have to be % escaped:
60	* control = 00-1F
61	* space = ' '
62	* delims = '<' , '>' , '#' , '%' , '"'
63	* unwise = '{' , '}' , '\|' , '\' , '^' , '[' , ']' , '`'
64	*/
65	#define URI_EXCLUDED(a) \
66	( ((a) >= 0x0 && (a) <= 0x20) \
67	\|\| ((a) >= 0x5B && (a) <= 0x5E) \
68	\|\| ((a) >= 0x7B && (a) <= 0x7D) \
69	\|\| (a) == '<' \|\| (a) == '>' \|\| (a) == '#' \
70	\|\| (a) == '%' \|\| (a) == '"' \|\| (a) == '`' )
71
72	static char rtUriPercentEncodeN(const char pszString, size_t cchMax)
73	{
74	if (!pszString)
75	return NULL;
76
77	int rc = VINF_SUCCESS;
78
79	size_t cbLen = RT_MIN(strlen(pszString), cchMax);
80	/* The new string can be max 3 times in size of the original string. */
81	char pszNew = RTStrAlloc(cbLen 3 + 1);
82	if (!pszNew)
83	return NULL;
84
85	char *pszRes = NULL;
86	size_t iIn = 0;
87	size_t iOut = 0;
88	while (iIn < cbLen)
89	{
90	if (URI_EXCLUDED(pszString[iIn]))
91	{
92	char szNum[3] = { 0, 0, 0 };
93	RTStrFormatU8(&szNum[0], 3, pszString[iIn++], 16, 2, 2, RTSTR_F_CAPITAL \| RTSTR_F_ZEROPAD);
94	pszNew[iOut++] = '%';
95	pszNew[iOut++] = szNum[0];
96	pszNew[iOut++] = szNum[1];
97	}
98	else
99	pszNew[iOut++] = pszString[iIn++];
100	}
101	if (RT_SUCCESS(rc))
102	{
103	pszNew[iOut] = '\0';
104	if (iOut != iIn)
105	{
106	/* If the source and target strings have different size, recreate
107	* the target string with the correct size. */
108	pszRes = RTStrDupN(pszNew, iOut);
109	RTStrFree(pszNew);
110	}
111	else
112	pszRes = pszNew;
113	}
114	else
115	RTStrFree(pszNew);
116
117	return pszRes;
118	}
119
120
121	static char rtUriPercentDecodeN(const char pszString, size_t cchString)
122	{
123	AssertPtr(pszString);
124	Assert(strlen(pszString) >= cchString);
125
126	/*
127	* The new string can only get smaller, so use the input length as a
128	* staring buffer size.
129	*/
130	char *pszDecoded = RTStrAlloc(cchString + 1);
131	if (pszDecoded)
132	{
133	/*
134	* Knowing that the pszString itself is valid UTF-8, we only have to
135	* validate the escape sequences.
136	*/
137	size_t cchLeft = cchString;
138	char const *pchSrc = pszString;
139	char *pchDst = pszDecoded;
140	while (cchLeft > 0)
141	{
142	const char pchPct = (const char )memchr(pchSrc, '%', cchLeft);
143	if (pchPct)
144	{
145	size_t cchBefore = pchPct - pchSrc;
146	if (cchBefore)
147	{
148	memcpy(pchDst, pchSrc, cchBefore);
149	pchDst += cchBefore;
150	pchSrc += cchBefore;
151	cchLeft -= cchBefore;
152	}
153
154	char chHigh, chLow;
155	if ( cchLeft >= 3
156	&& RT_C_IS_XDIGIT(chHigh = pchSrc[1])
157	&& RT_C_IS_XDIGIT(chLow = pchSrc[2]))
158	{
159	uint8_t b = RT_C_IS_DIGIT(chHigh) ? chHigh - '0' : (chHigh & ~0x20) - 'A' + 10;
160	b <<= 4;
161	b \|= RT_C_IS_DIGIT(chLow) ? chLow - '0' : (chLow & ~0x20) - 'A' + 10;
162	*pchDst++ = (char)b;
163	pchSrc += 3;
164	cchLeft -= 3;
165	}
166	else
167	{
168	AssertFailed();
169	pchDst++ = pchSrc++;
170	cchLeft--;
171	}
172	}
173	else
174	{
175	memcpy(pchDst, pchSrc, cchLeft);
176	pchDst += cchLeft;
177	pchSrc += cchLeft;
178	cchLeft = 0;
179	break;
180	}
181	}
182
183	*pchDst = '\0';
184
185	/*
186	* If we've got lof space room in the result string, reallocate it.
187	*/
188	size_t cchDecoded = pchDst - pszDecoded;
189	Assert(cchDecoded <= cchString);
190	// if (cchString - cchDecoded > 64) - enable later!
191	RTStrRealloc(&pszDecoded, cchDecoded + 1);
192	}
193	return pszDecoded;
194	}
195
196
197	static int rtUriParse(const char *pszUri, PRTURIPARSED pParsed)
198	{
199	/*
200	* Validate the input and clear the output.
201	*/
202	AssertPtrReturn(pParsed, VERR_INVALID_POINTER);
203	RT_ZERO(*pParsed);
204	pParsed->uAuthorityPort = UINT32_MAX;
205
206	AssertPtrReturn(pszUri, VERR_INVALID_POINTER);
207
208	size_t const cchUri = strlen(pszUri);
209	if (RT_LIKELY(cchUri >= 3)) { /* likely */ }
210	else return cchUri ? VERR_URI_TOO_SHORT : VERR_URI_EMPTY;
211
212	/*
213	* Validating escaped text sequences is much simpler if we know that
214	* that the base URI string is valid. Also, we don't necessarily trust
215	* the developer calling us to remember to do this.
216	*/
217	int rc = RTStrValidateEncoding(pszUri);
218	AssertRCReturn(rc, rc);
219
220	/*
221	* RFC-3986, section 3.1:
222	* scheme = ALPHA *( ALPHA / DIGIT / "+" / "-" / "." )
223	*
224	* The scheme ends with a ':', which we also skip here.
225	*/
226	size_t off = 0;
227	char ch = pszUri[off++];
228	if (RT_LIKELY(RT_C_IS_ALPHA(ch))) { /* likely */ }
229	else return VERR_URI_INVALID_SCHEME;
230	for (;;)
231	{
232	ch = pszUri[off];
233	if (ch == ':')
234	break;
235	if (RT_LIKELY(RT_C_IS_ALNUM(ch) \|\| ch == '.' \|\| ch == '-' \|\| ch == '+')) { /* likely */ }
236	else return VERR_URI_INVALID_SCHEME;
237	off++;
238	}
239	pParsed->cchScheme = off;
240
241	/* Require the scheme length to be at least two chars so we won't confuse
242	it with a path starting with a DOS drive letter specification. */
243	if (RT_LIKELY(off >= 2)) { /* likely */ }
244	else return VERR_URI_INVALID_SCHEME;
245
246	off++; /* (skip colon) */
247
248	/*
249	* Find the end of the path, we'll need this several times.
250	* Also, while we're potentially scanning the whole thing, check for '%'.
251	*/
252	size_t const offHash = RTStrOffCharOrTerm(&pszUri[off], '#') + off;
253	size_t const offQuestionMark = RTStrOffCharOrTerm(&pszUri[off], '?') + off;
254
255	if (memchr(pszUri, '%', cchUri) != NULL)
256	pParsed->fFlags \|= RTURIPARSED_F_CONTAINS_ESCAPED_CHARS;
257
258	/*
259	* RFC-3986, section 3.2:
260	* The authority component is preceeded by a double slash ("//")...
261	*/
262	if ( pszUri[off] == '/'
263	&& pszUri[off + 1] == '/')
264	{
265	off += 2;
266	pParsed->offAuthority = pParsed->offAuthorityUsername = pParsed->offAuthorityPassword = pParsed->offAuthorityHost = off;
267	pParsed->fFlags \|= RTURIPARSED_F_HAVE_AUTHORITY;
268
269	/*
270	* RFC-3986, section 3.2:
271	* ...and is terminated by the next slash ("/"), question mark ("?"),
272	* or number sign ("#") character, or by the end of the URI.
273	*/
274	const char *pszAuthority = &pszUri[off];
275	size_t cchAuthority = RTStrOffCharOrTerm(pszAuthority, '/');
276	cchAuthority = RT_MIN(cchAuthority, offHash - off);
277	cchAuthority = RT_MIN(cchAuthority, offQuestionMark - off);
278	pParsed->cchAuthority = cchAuthority;
279
280	/* The Authority can be empty, like for: file:///usr/bin/grep */
281	if (cchAuthority > 0)
282	{
283	pParsed->cchAuthorityHost = cchAuthority;
284
285	/*
286	* If there is a userinfo part, it is ended by a '@'.
287	*/
288	const char pszAt = (const char )memchr(pszAuthority, '@', cchAuthority);
289	if (pszAt)
290	{
291	size_t cchTmp = pszAt - pszAuthority;
292	pParsed->offAuthorityHost += cchTmp + 1;
293	pParsed->cchAuthorityHost -= cchTmp + 1;
294
295	/* If there is a password part, it's separated from the username with a colon. */
296	const char pszColon = (const char )memchr(pszAuthority, ':', cchTmp);
297	if (pszColon)
298	{
299	pParsed->cchAuthorityUsername = pszColon - pszAuthority;
300	pParsed->offAuthorityPassword = &pszColon[1] - pszUri;
301	pParsed->cchAuthorityPassword = pszAt - &pszColon[1];
302	}
303	else
304	{
305	pParsed->cchAuthorityUsername = cchTmp;
306	pParsed->offAuthorityPassword = off + cchTmp;
307	}
308	}
309
310	/*
311	* If there is a port part, its after the last colon in the host part.
312	*/
313	const char pszColon = (const char )memrchr(&pszUri[pParsed->offAuthorityHost], ':', pParsed->cchAuthorityHost);
314	if (pszColon)
315	{
316	size_t cchTmp = &pszUri[pParsed->offAuthorityHost + pParsed->cchAuthorityHost] - &pszColon[1];
317	pParsed->cchAuthorityHost -= cchTmp + 1;
318
319	pParsed->uAuthorityPort = 0;
320	while (cchTmp-- > 0)
321	{
322	ch = *++pszColon;
323	if ( RT_C_IS_DIGIT(ch)
324	&& pParsed->uAuthorityPort < UINT32_MAX / UINT32_C(10))
325	{
326	pParsed->uAuthorityPort *= 10;
327	pParsed->uAuthorityPort += ch - '0';
328	}
329	else
330	return VERR_URI_INVALID_PORT_NUMBER;
331	}
332	}
333	}
334
335	/* Skip past the authority. */
336	off += cchAuthority;
337	}
338	else
339	pParsed->offAuthority = pParsed->offAuthorityUsername = pParsed->offAuthorityPassword = pParsed->offAuthorityHost = off;
340
341	/*
342	* RFC-3986, section 3.3: Path
343	* The path is terminated by the first question mark ("?")
344	* or number sign ("#") character, or by the end of the URI.
345	*/
346	pParsed->offPath = off;
347	pParsed->cchPath = RT_MIN(offHash, offQuestionMark) - off;
348	off += pParsed->cchPath;
349
350	/*
351	* RFC-3986, section 3.4: Query
352	* The query component is indicated by the first question mark ("?")
353	* character and terminated by a number sign ("#") character or by the
354	* end of the URI.
355	*/
356	if ( off == offQuestionMark
357	&& off < cchUri)
358	{
359	Assert(pszUri[offQuestionMark] == '?');
360	pParsed->offQuery = ++off;
361	pParsed->cchQuery = offHash - off;
362	off = offHash;
363	}
364	else
365	{
366	Assert(!pszUri[offQuestionMark]);
367	pParsed->offQuery = off;
368	}
369
370	/*
371	* RFC-3986, section 3.5: Fragment
372	* A fragment identifier component is indicated by the presence of a
373	* number sign ("#") character and terminated by the end of the URI.
374	*/
375	if ( off == offHash
376	&& off < cchUri)
377	{
378	pParsed->offFragment = ++off;
379	pParsed->cchFragment = cchUri - off;
380	}
381	else
382	{
383	Assert(!pszUri[offHash]);
384	pParsed->offFragment = off;
385	}
386
387	/*
388	* If there are any escape sequences, validate them.
389	*
390	* This is reasonably simple as we already know that the string is valid UTF-8
391	* before they get decoded. Thus we only have to validate the escaped sequences.
392	*/
393	if (pParsed->fFlags & RTURIPARSED_F_CONTAINS_ESCAPED_CHARS)
394	{
395	const char pchSrc = (const char )memchr(pszUri, '%', cchUri);
396	AssertReturn(pchSrc, VERR_INTERNAL_ERROR);
397	do
398	{
399	char szUtf8Seq[8];
400	unsigned cchUtf8Seq = 0;
401	unsigned cchNeeded = 0;
402	size_t cchLeft = &pszUri[cchUri] - pchSrc;
403	do
404	{
405	if (cchLeft >= 3)
406	{
407	char chHigh = pchSrc[1];
408	char chLow = pchSrc[2];
409	if ( RT_C_IS_XDIGIT(chHigh)
410	&& RT_C_IS_XDIGIT(chLow))
411	{
412	uint8_t b = RT_C_IS_DIGIT(chHigh) ? chHigh - '0' : (chHigh & ~0x20) - 'A' + 10;
413	b <<= 4;
414	b \|= RT_C_IS_DIGIT(chLow) ? chLow - '0' : (chLow & ~0x20) - 'A' + 10;
415
416	if (!(b & 0x80))
417	{
418	/* We don't want the string to be terminated prematurely. */
419	if (RT_LIKELY(b != 0)) { /* likely */ }
420	else return VERR_URI_ESCAPED_ZERO;
421
422	/* Check that we're not expecting more UTF-8 bytes. */
423	if (RT_LIKELY(cchNeeded == 0)) { /* likely */ }
424	else return VERR_URI_MISSING_UTF8_CONTINUATION_BYTE;
425	}
426	/* Are we waiting UTF-8 bytes? */
427	else if (cchNeeded > 0)
428	{
429	if (RT_LIKELY(!(b & 0x40))) { /* likely */ }
430	else return VERR_URI_INVALID_ESCAPED_UTF8_CONTINUATION_BYTE;
431
432	szUtf8Seq[cchUtf8Seq++] = (char)b;
433	if (--cchNeeded == 0)
434	{
435	szUtf8Seq[cchUtf8Seq] = '\0';
436	int rc = RTStrValidateEncoding(szUtf8Seq);
437	if (RT_FAILURE(rc))
438	return VERR_URI_ESCAPED_CHARS_NOT_VALID_UTF8;
439	cchUtf8Seq = 0;
440	}
441	}
442	/* Start a new UTF-8 sequence. */
443	else
444	{
445	if ((b & 0xf8) == 0xf0)
446	cchNeeded = 3;
447	else if ((b & 0xf0) == 0xe0)
448	cchNeeded = 2;
449	else if ((b & 0xe0) == 0xc0)
450	cchNeeded = 1;
451	else
452	return VERR_URI_INVALID_ESCAPED_UTF8_LEAD_BYTE;
453	szUtf8Seq[0] = (char)b;
454	cchUtf8Seq = 1;
455	}
456	pchSrc += 3;
457	cchLeft -= 3;
458	}
459	else
460	return VERR_URI_INVALID_ESCAPE_SEQ;
461	}
462	else
463	return VERR_URI_INVALID_ESCAPE_SEQ;
464	} while (cchLeft > 0 && pchSrc[0] == '%');
465
466	/* Check that we're not expecting more UTF-8 bytes. */
467	if (RT_LIKELY(cchNeeded == 0)) { /* likely */ }
468	else return VERR_URI_MISSING_UTF8_CONTINUATION_BYTE;
469
470	/* next */
471	pchSrc = (const char *)memchr(pchSrc, '%', cchLeft);
472	} while (pchSrc);
473	}
474
475	pParsed->u32Magic = RTURIPARSED_MAGIC;
476	return VINF_SUCCESS;
477	}
478
479
480
481	RTDECL(int) RTUriParse(const char *pszUri, PRTURIPARSED pParsed)
482	{
483	return rtUriParse(pszUri, pParsed);
484	}
485
486
487	RTDECL(char ) RTUriParsedScheme(const char pszUri, PCRTURIPARSED pParsed)
488	{
489	AssertPtrReturn(pszUri, NULL);
490	AssertPtrReturn(pParsed, NULL);
491	AssertReturn(pParsed->u32Magic == RTURIPARSED_MAGIC, NULL);
492	return RTStrDupN(pszUri, pParsed->cchScheme);
493	}
494
495
496	RTDECL(char ) RTUriParsedAuthority(const char pszUri, PCRTURIPARSED pParsed)
497	{
498	AssertPtrReturn(pszUri, NULL);
499	AssertPtrReturn(pParsed, NULL);
500	AssertReturn(pParsed->u32Magic == RTURIPARSED_MAGIC, NULL);
501	if (pParsed->cchAuthority \|\| (pParsed->fFlags & RTURIPARSED_F_HAVE_AUTHORITY))
502	return rtUriPercentDecodeN(&pszUri[pParsed->offAuthority], pParsed->cchAuthority);
503	return NULL;
504	}
505
506
507	RTDECL(char ) RTUriParsedAuthorityUsername(const char pszUri, PCRTURIPARSED pParsed)
508	{
509	AssertPtrReturn(pszUri, NULL);
510	AssertPtrReturn(pParsed, NULL);
511	AssertReturn(pParsed->u32Magic == RTURIPARSED_MAGIC, NULL);
512	if (pParsed->cchAuthorityUsername)
513	return rtUriPercentDecodeN(&pszUri[pParsed->offAuthorityUsername], pParsed->cchAuthorityUsername);
514	return NULL;
515	}
516
517
518	RTDECL(char ) RTUriParsedAuthorityPassword(const char pszUri, PCRTURIPARSED pParsed)
519	{
520	AssertPtrReturn(pszUri, NULL);
521	AssertPtrReturn(pParsed, NULL);
522	AssertReturn(pParsed->u32Magic == RTURIPARSED_MAGIC, NULL);
523	if (pParsed->cchAuthorityPassword)
524	return rtUriPercentDecodeN(&pszUri[pParsed->offAuthorityPassword], pParsed->cchAuthorityPassword);
525	return NULL;
526	}
527
528
529	RTDECL(char ) RTUriParsedAuthorityHost(const char pszUri, PCRTURIPARSED pParsed)
530	{
531	AssertPtrReturn(pszUri, NULL);
532	AssertPtrReturn(pParsed, NULL);
533	AssertReturn(pParsed->u32Magic == RTURIPARSED_MAGIC, NULL);
534	if (pParsed->cchAuthorityHost)
535	return rtUriPercentDecodeN(&pszUri[pParsed->offAuthorityHost], pParsed->cchAuthorityHost);
536	return NULL;
537	}
538
539
540	RTDECL(uint32_t) RTUriParsedAuthorityPort(const char *pszUri, PCRTURIPARSED pParsed)
541	{
542	AssertPtrReturn(pszUri, UINT32_MAX);
543	AssertPtrReturn(pParsed, UINT32_MAX);
544	AssertReturn(pParsed->u32Magic == RTURIPARSED_MAGIC, UINT32_MAX);
545	return pParsed->uAuthorityPort;
546	}
547
548
549	RTDECL(char ) RTUriParsedPath(const char pszUri, PCRTURIPARSED pParsed)
550	{
551	AssertPtrReturn(pszUri, NULL);
552	AssertPtrReturn(pParsed, NULL);
553	AssertReturn(pParsed->u32Magic == RTURIPARSED_MAGIC, NULL);
554	if (pParsed->cchPath)
555	return rtUriPercentDecodeN(&pszUri[pParsed->offPath], pParsed->cchPath);
556	return NULL;
557	}
558
559
560	RTDECL(char ) RTUriParsedQuery(const char pszUri, PCRTURIPARSED pParsed)
561	{
562	AssertPtrReturn(pszUri, NULL);
563	AssertPtrReturn(pParsed, NULL);
564	AssertReturn(pParsed->u32Magic == RTURIPARSED_MAGIC, NULL);
565	if (pParsed->cchQuery)
566	return rtUriPercentDecodeN(&pszUri[pParsed->offQuery], pParsed->cchQuery);
567	return NULL;
568	}
569
570
571	RTDECL(char ) RTUriParsedFragment(const char pszUri, PCRTURIPARSED pParsed)
572	{
573	AssertPtrReturn(pszUri, NULL);
574	AssertPtrReturn(pParsed, NULL);
575	AssertReturn(pParsed->u32Magic == RTURIPARSED_MAGIC, NULL);
576	if (pParsed->cchFragment)
577	return rtUriPercentDecodeN(&pszUri[pParsed->offFragment], pParsed->cchFragment);
578	return NULL;
579	}
580
581
582	RTDECL(char ) RTUriCreate(const char pszScheme, const char pszAuthority, const char pszPath, const char *pszQuery,
583	const char *pszFragment)
584	{
585	if (!pszScheme) /* Scheme is minimum requirement */
586	return NULL;
587
588	char *pszResult = 0;
589	char *pszAuthority1 = 0;
590	char *pszPath1 = 0;
591	char *pszQuery1 = 0;
592	char *pszFragment1 = 0;
593
594	do
595	{
596	/* Create the percent encoded strings and calculate the necessary uri
597	* length. */
598	size_t cbSize = strlen(pszScheme) + 1 + 1; /* plus zero byte */
599	if (pszAuthority)
600	{
601	pszAuthority1 = rtUriPercentEncodeN(pszAuthority, RTSTR_MAX);
602	if (!pszAuthority1)
603	break;
604	cbSize += strlen(pszAuthority1) + 2;
605	}
606	if (pszPath)
607	{
608	pszPath1 = rtUriPercentEncodeN(pszPath, RTSTR_MAX);
609	if (!pszPath1)
610	break;
611	cbSize += strlen(pszPath1);
612	}
613	if (pszQuery)
614	{
615	pszQuery1 = rtUriPercentEncodeN(pszQuery, RTSTR_MAX);
616	if (!pszQuery1)
617	break;
618	cbSize += strlen(pszQuery1) + 1;
619	}
620	if (pszFragment)
621	{
622	pszFragment1 = rtUriPercentEncodeN(pszFragment, RTSTR_MAX);
623	if (!pszFragment1)
624	break;
625	cbSize += strlen(pszFragment1) + 1;
626	}
627
628	char pszTmp = pszResult = (char )RTStrAlloc(cbSize);
629	if (!pszResult)
630	break;
631	RT_BZERO(pszTmp, cbSize);
632
633	/* Compose the target uri string. */
634	RTStrCatP(&pszTmp, &cbSize, pszScheme);
635	RTStrCatP(&pszTmp, &cbSize, ":");
636	if (pszAuthority1)
637	{
638	RTStrCatP(&pszTmp, &cbSize, "//");
639	RTStrCatP(&pszTmp, &cbSize, pszAuthority1);
640	}
641	if (pszPath1)
642	{
643	RTStrCatP(&pszTmp, &cbSize, pszPath1);
644	}
645	if (pszQuery1)
646	{
647	RTStrCatP(&pszTmp, &cbSize, "?");
648	RTStrCatP(&pszTmp, &cbSize, pszQuery1);
649	}
650	if (pszFragment1)
651	{
652	RTStrCatP(&pszTmp, &cbSize, "#");
653	RTStrCatP(&pszTmp, &cbSize, pszFragment1);
654	}
655	} while (0);
656
657	/* Cleanup */
658	if (pszAuthority1)
659	RTStrFree(pszAuthority1);
660	if (pszPath1)
661	RTStrFree(pszPath1);
662	if (pszQuery1)
663	RTStrFree(pszQuery1);
664	if (pszFragment1)
665	RTStrFree(pszFragment1);
666
667	return pszResult;
668	}
669
670
671	RTDECL(bool) RTUriIsSchemeMatch(const char pszUri, const char pszScheme)
672	{
673	AssertPtrReturn(pszUri, false);
674	size_t const cchScheme = strlen(pszScheme);
675	return RTStrNICmp(pszUri, pszScheme, cchScheme) == 0
676	&& pszScheme[cchScheme] == ':';
677	}
678
679
680	/* temporarily kept around till Andy stops using it. */
681	RTDECL(char ) RTUriPath(const char pszUri)
682	{
683	RTURIPARSED Parsed;
684	int rc = rtUriParse(pszUri, &Parsed);
685	if (RT_SUCCESS(rc))
686	if (Parsed.cchPath)
687	return rtUriPercentDecodeN(&pszUri[Parsed.offPath], Parsed.cchPath);
688	return NULL;
689	}
690
691
692	RTDECL(char ) RTUriFileCreate(const char pszPath)
693	{
694	char *pszResult = NULL;
695	if (pszPath)
696	{
697	/* Create the percent encoded strings and calculate the necessary uri length. */
698	char *pszPath1 = rtUriPercentEncodeN(pszPath, RTSTR_MAX);
699	if (pszPath1)
700	{
701	size_t cbSize = 7 /* file:// / + strlen(pszPath1) + 1; / plus zero byte */
702	if (pszPath1[0] != '/')
703	++cbSize;
704	char *pszTmp = pszResult = RTStrAlloc(cbSize);
705	if (pszResult)
706	{
707	/* Compose the target uri string. */
708	*pszTmp = '\0';
709	RTStrCatP(&pszTmp, &cbSize, "file://");
710	if (pszPath1[0] != '/')
711	RTStrCatP(&pszTmp, &cbSize, "/");
712	RTStrCatP(&pszTmp, &cbSize, pszPath1);
713	}
714	RTStrFree(pszPath1);
715	}
716	}
717	return pszResult;
718	}
719
720
721	RTDECL(char ) RTUriFilePath(const char pszUri, uint32_t uFormat)
722	{
723	return RTUriFileNPath(pszUri, uFormat, RTSTR_MAX);
724	}
725
726
727	RTDECL(char ) RTUriFileNPath(const char pszUri, uint32_t uFormat, size_t cchMax)
728	{
729	AssertPtrReturn(pszUri, NULL);
730	AssertReturn(uFormat == URI_FILE_FORMAT_AUTO \|\| uFormat == URI_FILE_FORMAT_UNIX \|\| uFormat == URI_FILE_FORMAT_WIN, NULL);
731
732	/* Auto is based on the current OS. */
733	if (uFormat == URI_FILE_FORMAT_AUTO)
734	#if defined(RT_OS_WINDOWS) \|\| defined(RT_OS_OS2)
735	uFormat = URI_FILE_FORMAT_WIN;
736	#else
737	uFormat = URI_FILE_FORMAT_UNIX;
738	#endif
739
740	/* Check that this is a file Uri */
741	if (RTStrNICmp(pszUri, RT_STR_TUPLE("file:")) != 0)
742	return NULL;
743
744	RTURIPARSED Parsed;
745	int rc = rtUriParse(pszUri, &Parsed);
746	if (RT_SUCCESS(rc) && Parsed.cchPath)
747	{
748	/* Special hack for DOS path like file:///c:/WINDOWS/clock.avi where we
749	have to drop the leading slash that was used to separate the authority
750	from the path. */
751	if ( uFormat == URI_FILE_FORMAT_WIN
752	&& Parsed.cchPath >= 3
753	&& pszUri[Parsed.offPath] == '/'
754	&& pszUri[Parsed.offPath + 2] == ':'
755	&& RT_C_IS_ALPHA(pszUri[Parsed.offPath + 1]) )
756	{
757	Parsed.offPath++;
758	Parsed.cchPath--;
759	}
760
761	char *pszPath = rtUriPercentDecodeN(&pszUri[Parsed.offPath], Parsed.cchPath);
762	if (uFormat == URI_FILE_FORMAT_UNIX)
763	return RTPathChangeToUnixSlashes(pszPath, true);
764	Assert(uFormat == URI_FILE_FORMAT_WIN);
765	return RTPathChangeToDosSlashes(pszPath, true);
766	}
767	return NULL;
768	}
769

Note: See TracBrowser for help on using the repository browser.

source: vbox/trunk/src/VBox/Runtime/common/misc/uri.cpp@ 57723

Download in other formats: