/src/libreoffice/sal/rtl/uri.cxx
Line | Count | Source |
1 | | /* -*- Mode: C++; tab-width: 4; indent-tabs-mode: nil; c-basic-offset: 4 -*- */ |
2 | | /* |
3 | | * This file is part of the LibreOffice project. |
4 | | * |
5 | | * This Source Code Form is subject to the terms of the Mozilla Public |
6 | | * License, v. 2.0. If a copy of the MPL was not distributed with this |
7 | | * file, You can obtain one at http://mozilla.org/MPL/2.0/. |
8 | | * |
9 | | * This file incorporates work covered by the following license notice: |
10 | | * |
11 | | * Licensed to the Apache Software Foundation (ASF) under one or more |
12 | | * contributor license agreements. See the NOTICE file distributed |
13 | | * with this work for additional information regarding copyright |
14 | | * ownership. The ASF licenses this file to you under the Apache |
15 | | * License, Version 2.0 (the "License"); you may not use this file |
16 | | * except in compliance with the License. You may obtain a copy of |
17 | | * the License at http://www.apache.org/licenses/LICENSE-2.0 . |
18 | | */ |
19 | | |
20 | | #include <rtl/character.hxx> |
21 | | #include <rtl/strbuf.hxx> |
22 | | #include <rtl/textenc.h> |
23 | | #include <rtl/textcvt.h> |
24 | | #include <rtl/uri.h> |
25 | | #include <rtl/uri.hxx> |
26 | | #include <rtl/ustrbuf.h> |
27 | | #include <rtl/ustrbuf.hxx> |
28 | | #include <rtl/ustring.h> |
29 | | #include <rtl/ustring.hxx> |
30 | | #include <sal/types.h> |
31 | | #include <sal/macros.h> |
32 | | |
33 | | #include <uri_internal.hxx> |
34 | | |
35 | | #include <algorithm> |
36 | | #include <cstddef> |
37 | | |
38 | | namespace { |
39 | | |
40 | | sal_Unicode const cEscapePrefix = 0x25; // '%' |
41 | | |
42 | | int getHexWeight(sal_uInt32 nUtf32) |
43 | 1.48M | { |
44 | 1.48M | return nUtf32 >= 0x30 && nUtf32 <= 0x39 ? // '0'--'9' |
45 | 765k | static_cast< int >(nUtf32 - 0x30) : |
46 | 1.48M | nUtf32 >= 0x41 && nUtf32 <= 0x46 ? // 'A'--'F' |
47 | 718k | static_cast< int >(nUtf32 - 0x41 + 10) : |
48 | 719k | nUtf32 >= 0x61 && nUtf32 <= 0x66 ? // 'a'--'f' |
49 | 15 | static_cast< int >(nUtf32 - 0x61 + 10) : |
50 | 269 | -1; // not a hex digit |
51 | 1.48M | } |
52 | | |
53 | | bool isValid(sal_Bool const * pCharClass, sal_uInt32 nUtf32) |
54 | 29.3M | { |
55 | 29.3M | return nUtf32 < rtl::UriCharClassSize && pCharClass[nUtf32]; |
56 | 29.3M | } |
57 | | |
58 | | void writeUnicode(rtl_uString ** pBuffer, sal_Int32 * pCapacity, |
59 | | sal_Unicode cChar) |
60 | 33.5M | { |
61 | 33.5M | rtl_uStringbuffer_insert(pBuffer, pCapacity, (*pBuffer)->length, &cChar, 1); |
62 | 33.5M | } |
63 | | |
64 | | } |
65 | | |
66 | | namespace rtl::uri::detail { |
67 | | |
68 | | /** Read any of the following: |
69 | | |
70 | | @li sequence of escape sequences representing character from eCharset, |
71 | | translated to single UCS4 character; or |
72 | | @li pair of UTF-16 surrogates, translated to single UCS4 character; or |
73 | | @li single UTF-16 character, extended to UCS4 character. |
74 | | */ |
75 | | sal_uInt32 readUcs4(sal_Unicode const ** pBegin, sal_Unicode const * pEnd, |
76 | | bool bEncoded, rtl_TextEncoding eCharset, |
77 | | EscapeType * pType) |
78 | 103M | { |
79 | 103M | sal_uInt32 nChar = *(*pBegin)++; |
80 | 103M | int nWeight1; |
81 | 103M | int nWeight2; |
82 | 103M | if (nChar == cEscapePrefix && bEncoded && pEnd - *pBegin >= 2 |
83 | 739k | && (nWeight1 = getHexWeight((*pBegin)[0])) >= 0 |
84 | 738k | && (nWeight2 = getHexWeight((*pBegin)[1])) >= 0) |
85 | 738k | { |
86 | 738k | *pBegin += 2; |
87 | 738k | nChar = static_cast< sal_uInt32 >(nWeight1 << 4 | nWeight2); |
88 | 738k | if (nChar <= 0x7F) |
89 | 736k | { |
90 | 736k | *pType = EscapeChar; |
91 | 736k | } |
92 | 2.22k | else if (eCharset == RTL_TEXTENCODING_UTF8) |
93 | 2.21k | { |
94 | 2.21k | if (nChar >= 0xC0 && nChar <= 0xF4) |
95 | 2.20k | { |
96 | 2.20k | sal_uInt32 nEncoded; |
97 | 2.20k | int nShift; |
98 | 2.20k | sal_uInt32 nMin; |
99 | 2.20k | if (nChar <= 0xDF) |
100 | 1.15k | { |
101 | 1.15k | nEncoded = (nChar & 0x1F) << 6; |
102 | 1.15k | nShift = 0; |
103 | 1.15k | nMin = 0x80; |
104 | 1.15k | } |
105 | 1.05k | else if (nChar <= 0xEF) |
106 | 1.00k | { |
107 | 1.00k | nEncoded = (nChar & 0x0F) << 12; |
108 | 1.00k | nShift = 6; |
109 | 1.00k | nMin = 0x800; |
110 | 1.00k | } |
111 | 42 | else |
112 | 42 | { |
113 | 42 | nEncoded = (nChar & 0x07) << 18; |
114 | 42 | nShift = 12; |
115 | 42 | nMin = 0x10000; |
116 | 42 | } |
117 | | |
118 | 2.20k | sal_Unicode const * p = *pBegin; |
119 | 2.20k | bool bUTF8 = true; |
120 | | |
121 | 5.49k | for (; nShift >= 0; nShift -= 6) |
122 | 3.29k | { |
123 | 3.29k | if (pEnd - p < 3 || p[0] != cEscapePrefix |
124 | 3.29k | || (nWeight1 = getHexWeight(p[1])) < 8 |
125 | 3.29k | || nWeight1 > 11 |
126 | 3.29k | || (nWeight2 = getHexWeight(p[2])) < 0) |
127 | 2 | { |
128 | 2 | bUTF8 = false; |
129 | 2 | break; |
130 | 2 | } |
131 | 3.29k | p += 3; |
132 | 3.29k | nEncoded |= ((nWeight1 & 3) << 4 | nWeight2) << nShift; |
133 | 3.29k | } |
134 | 2.20k | if (bUTF8 && rtl::isUnicodeScalarValue(nEncoded) |
135 | 2.19k | && nEncoded >= nMin) |
136 | 2.19k | { |
137 | 2.19k | *pBegin = p; |
138 | 2.19k | *pType = EscapeChar; |
139 | 2.19k | return nEncoded; |
140 | 2.19k | } |
141 | 2.20k | } |
142 | 24 | *pType = EscapeOctet; |
143 | 24 | } |
144 | 6 | else |
145 | 6 | { |
146 | 6 | OStringBuffer aBuf(16); |
147 | 6 | aBuf.append(static_cast< char >(nChar)); |
148 | 6 | rtl_TextToUnicodeConverter aConverter |
149 | 6 | = rtl_createTextToUnicodeConverter(eCharset); |
150 | 6 | sal_Unicode const * p = *pBegin; |
151 | | |
152 | 6 | for (;;) |
153 | 6 | { |
154 | 6 | sal_Unicode aDst[2]; |
155 | 6 | sal_uInt32 nInfo; |
156 | 6 | sal_Size nConverted; |
157 | 6 | sal_Size nDstSize = rtl_convertTextToUnicode( |
158 | 6 | aConverter, nullptr, aBuf.getStr(), aBuf.getLength(), aDst, |
159 | 6 | SAL_N_ELEMENTS( aDst ), |
160 | 6 | (RTL_TEXTTOUNICODE_FLAGS_UNDEFINED_ERROR |
161 | 6 | | RTL_TEXTTOUNICODE_FLAGS_MBUNDEFINED_ERROR |
162 | 6 | | RTL_TEXTTOUNICODE_FLAGS_INVALID_ERROR), |
163 | 6 | &nInfo, &nConverted); |
164 | | |
165 | 6 | if (nInfo == 0) |
166 | 6 | { |
167 | 6 | assert( nConverted |
168 | 6 | == sal::static_int_cast< sal_uInt32 >( |
169 | 6 | aBuf.getLength())); |
170 | | |
171 | 6 | rtl_destroyTextToUnicodeConverter(aConverter); |
172 | 6 | *pBegin = p; |
173 | 6 | *pType = EscapeChar; |
174 | | |
175 | 6 | assert( nDstSize == 1 |
176 | 6 | || (nDstSize == 2 && rtl::isHighSurrogate(aDst[0]) |
177 | 6 | && rtl::isLowSurrogate(aDst[1]))); |
178 | | |
179 | 6 | return nDstSize == 1 |
180 | 6 | ? aDst[0] : rtl::combineSurrogates(aDst[0], aDst[1]); |
181 | 6 | } |
182 | 0 | if (nInfo == RTL_TEXTTOUNICODE_INFO_SRCBUFFERTOOSMALL |
183 | 0 | && pEnd - p >= 3 && p[0] == cEscapePrefix |
184 | 0 | && (nWeight1 = getHexWeight(p[1])) >= 0 |
185 | 0 | && (nWeight2 = getHexWeight(p[2])) >= 0) |
186 | 0 | { |
187 | 0 | p += 3; |
188 | 0 | aBuf.append(static_cast< char >(nWeight1 << 4 | nWeight2)); |
189 | 0 | } |
190 | 0 | else if (nInfo == RTL_TEXTTOUNICODE_INFO_SRCBUFFERTOOSMALL |
191 | 0 | && p != pEnd && *p <= 0x7F) |
192 | 0 | { |
193 | 0 | aBuf.append(static_cast< char >(*p++)); |
194 | 0 | } |
195 | 0 | else |
196 | 0 | { |
197 | 0 | assert( |
198 | 0 | (nInfo & RTL_TEXTTOUNICODE_INFO_DESTBUFFERTOOSMALL) |
199 | 0 | == 0); |
200 | 0 | break; |
201 | 0 | } |
202 | 0 | } |
203 | 0 | rtl_destroyTextToUnicodeConverter(aConverter); |
204 | 0 | *pType = EscapeOctet; |
205 | 0 | } |
206 | 736k | return nChar; |
207 | 738k | } |
208 | | |
209 | 103M | *pType = EscapeNo; |
210 | 103M | return rtl::isHighSurrogate(nChar) && *pBegin < pEnd |
211 | 3.11k | && rtl::isLowSurrogate(**pBegin) ? |
212 | 103M | rtl::combineSurrogates(nChar, *(*pBegin)++) : nChar; |
213 | 103M | } |
214 | | |
215 | | } |
216 | | |
217 | | namespace { |
218 | | |
219 | | void writeUcs4(rtl_uString ** pBuffer, sal_Int32 * pCapacity, sal_uInt32 nUtf32) |
220 | 25.2M | { |
221 | 25.2M | rtl_uStringbuffer_insertUtf32(pBuffer, pCapacity, (*pBuffer)->length, nUtf32); |
222 | 25.2M | } |
223 | | |
224 | | void writeEscapeOctet(rtl_uString ** pBuffer, sal_Int32 * pCapacity, |
225 | | sal_uInt32 nOctet) |
226 | 1.81M | { |
227 | 1.81M | assert(nOctet <= 0xFF); // bad octet |
228 | | |
229 | 1.81M | static sal_Unicode const aHex[16] |
230 | 1.81M | = { 0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, |
231 | 1.81M | 0x41, 0x42, 0x43, 0x44, 0x45, 0x46 }; /* '0'--'9', 'A'--'F' */ |
232 | | |
233 | 1.81M | writeUnicode(pBuffer, pCapacity, cEscapePrefix); |
234 | 1.81M | writeUnicode(pBuffer, pCapacity, aHex[nOctet >> 4]); |
235 | 1.81M | writeUnicode(pBuffer, pCapacity, aHex[nOctet & 15]); |
236 | 1.81M | } |
237 | | |
238 | | bool writeEscapeChar(rtl_uString ** pBuffer, sal_Int32 * pCapacity, |
239 | | sal_uInt32 nUtf32, rtl_TextEncoding eCharset, bool bStrict) |
240 | 1.25M | { |
241 | 1.25M | assert(rtl::isUnicodeCodePoint(nUtf32)); |
242 | 1.25M | if (eCharset == RTL_TEXTENCODING_UTF8) |
243 | 1.25M | { |
244 | 1.25M | if (nUtf32 < 0x80) |
245 | 950k | { |
246 | 950k | writeEscapeOctet(pBuffer, pCapacity, nUtf32); |
247 | 950k | } |
248 | 307k | else if (nUtf32 < 0x800) |
249 | 54.0k | { |
250 | 54.0k | writeEscapeOctet(pBuffer, pCapacity, nUtf32 >> 6 | 0xC0); |
251 | 54.0k | writeEscapeOctet(pBuffer, pCapacity, (nUtf32 & 0x3F) | 0x80); |
252 | 54.0k | } |
253 | 253k | else if (nUtf32 < 0x10000) |
254 | 253k | { |
255 | 253k | writeEscapeOctet(pBuffer, pCapacity, nUtf32 >> 12 | 0xE0); |
256 | 253k | writeEscapeOctet(pBuffer, pCapacity, (nUtf32 >> 6 & 0x3F) | 0x80); |
257 | 253k | writeEscapeOctet(pBuffer, pCapacity, (nUtf32 & 0x3F) | 0x80); |
258 | 253k | } |
259 | 350 | else |
260 | 350 | { |
261 | 350 | writeEscapeOctet(pBuffer, pCapacity, nUtf32 >> 18 | 0xF0); |
262 | 350 | writeEscapeOctet(pBuffer, pCapacity, (nUtf32 >> 12 & 0x3F) | 0x80); |
263 | 350 | writeEscapeOctet(pBuffer, pCapacity, (nUtf32 >> 6 & 0x3F) | 0x80); |
264 | 350 | writeEscapeOctet(pBuffer, pCapacity, (nUtf32 & 0x3F) | 0x80); |
265 | 350 | } |
266 | 1.25M | } |
267 | 0 | else |
268 | 0 | { |
269 | 0 | rtl_UnicodeToTextConverter aConverter |
270 | 0 | = rtl_createUnicodeToTextConverter(eCharset); |
271 | 0 | sal_Unicode aSrc[2]; |
272 | 0 | sal_Size nSrcSize = rtl::splitSurrogates(nUtf32, aSrc); |
273 | |
|
274 | 0 | char aDst[32]; // FIXME random value |
275 | 0 | sal_uInt32 nInfo; |
276 | 0 | sal_Size nConverted; |
277 | 0 | sal_Size nDstSize = rtl_convertUnicodeToText( |
278 | 0 | aConverter, nullptr, aSrc, nSrcSize, aDst, sizeof aDst, |
279 | 0 | RTL_UNICODETOTEXT_FLAGS_UNDEFINED_ERROR |
280 | 0 | | RTL_UNICODETOTEXT_FLAGS_INVALID_ERROR |
281 | 0 | | RTL_UNICODETOTEXT_FLAGS_FLUSH, |
282 | 0 | &nInfo, &nConverted); |
283 | 0 | assert((nInfo & RTL_UNICODETOTEXT_INFO_DESTBUFFERTOSMALL) == 0); |
284 | 0 | rtl_destroyUnicodeToTextConverter(aConverter); |
285 | |
|
286 | 0 | if (nInfo == 0) |
287 | 0 | { |
288 | 0 | assert(nConverted == nSrcSize); // bad rtl_convertUnicodeToText |
289 | |
|
290 | 0 | for (sal_Size i = 0; i < nDstSize; ++i) |
291 | 0 | { |
292 | 0 | writeEscapeOctet(pBuffer, pCapacity, |
293 | 0 | static_cast< unsigned char >(aDst[i])); |
294 | | // FIXME all octets are escaped, even if there is no need |
295 | 0 | } |
296 | 0 | } |
297 | 0 | else |
298 | 0 | { |
299 | 0 | if (bStrict) |
300 | 0 | return false; |
301 | | |
302 | 0 | writeUcs4(pBuffer, pCapacity, nUtf32); |
303 | 0 | } |
304 | 0 | } |
305 | 1.25M | return true; |
306 | 1.25M | } |
307 | | |
308 | | struct Component |
309 | | { |
310 | | sal_Unicode const * pBegin; |
311 | | sal_Unicode const * pEnd; |
312 | | |
313 | 120k | Component(): pBegin(nullptr), pEnd(nullptr) {} |
314 | | |
315 | 37.8k | bool isPresent() const { return pBegin != nullptr; } |
316 | | |
317 | | sal_Int32 getLength() const; |
318 | | }; |
319 | | |
320 | | sal_Int32 Component::getLength() const |
321 | 8.11k | { |
322 | 8.11k | assert(isPresent()); // taking length of non-present component |
323 | 8.11k | return static_cast< sal_Int32 >(pEnd - pBegin); |
324 | 8.11k | } |
325 | | |
326 | | struct Components |
327 | | { |
328 | | Component aScheme; |
329 | | Component aAuthority; |
330 | | Component aPath; |
331 | | Component aQuery; |
332 | | Component aFragment; |
333 | | }; |
334 | | |
335 | | void parseUriRef(rtl_uString const * pUriRef, Components * pComponents) |
336 | 24.1k | { |
337 | | // This algorithm is liberal and accepts various forms of illegal input. |
338 | | |
339 | 24.1k | sal_Unicode const * pBegin = pUriRef->buffer; |
340 | 24.1k | sal_Unicode const * pEnd = pBegin + pUriRef->length; |
341 | 24.1k | sal_Unicode const * pPos = pBegin; |
342 | | |
343 | 24.1k | if (pPos != pEnd && rtl::isAsciiAlpha(*pPos)) |
344 | 4.18k | { |
345 | 20.5k | for (sal_Unicode const * p = pPos + 1; p != pEnd; ++p) |
346 | 20.4k | { |
347 | 20.4k | if (*p == ':') |
348 | 4.04k | { |
349 | 4.04k | pComponents->aScheme.pBegin = pBegin; |
350 | 4.04k | pComponents->aScheme.pEnd = ++p; |
351 | 4.04k | pPos = p; |
352 | 4.04k | break; |
353 | 4.04k | } |
354 | | |
355 | 16.4k | if (!rtl::isAsciiAlphanumeric(*p) && *p != '+' && *p != '-' |
356 | 123 | && *p != '.') |
357 | 114 | { |
358 | 114 | break; |
359 | 114 | } |
360 | 16.4k | } |
361 | 4.18k | } |
362 | | |
363 | 24.1k | if (pEnd - pPos >= 2 && pPos[0] == '/' && pPos[1] == '/') |
364 | 2.80k | { |
365 | 2.80k | pComponents->aAuthority.pBegin = pPos; |
366 | 2.80k | pPos += 2; |
367 | 49.5k | while (pPos != pEnd && *pPos != '/' && *pPos != '?' && *pPos != '#') |
368 | 46.7k | { |
369 | 46.7k | ++pPos; |
370 | 46.7k | } |
371 | | |
372 | 2.80k | pComponents->aAuthority.pEnd = pPos; |
373 | 2.80k | } |
374 | | |
375 | 24.1k | pComponents->aPath.pBegin = pPos; |
376 | 162k | while (pPos != pEnd && *pPos != '?' && * pPos != '#') |
377 | 138k | { |
378 | 138k | ++pPos; |
379 | 138k | } |
380 | | |
381 | 24.1k | pComponents->aPath.pEnd = pPos; |
382 | | |
383 | 24.1k | if (pPos != pEnd && *pPos == '?') |
384 | 581 | { |
385 | 581 | pComponents->aQuery.pBegin = pPos++; |
386 | 25.3k | while (pPos != pEnd && * pPos != '#') |
387 | 24.7k | { |
388 | 24.7k | ++pPos; |
389 | 24.7k | } |
390 | | |
391 | 581 | pComponents->aQuery.pEnd = pPos; |
392 | 581 | } |
393 | | |
394 | 24.1k | if (pPos != pEnd) |
395 | 64 | { |
396 | 64 | assert(*pPos == '#'); |
397 | 64 | pComponents->aFragment.pBegin = pPos; |
398 | 64 | pComponents->aFragment.pEnd = pEnd; |
399 | 64 | } |
400 | 24.1k | } |
401 | | |
402 | | void appendPath( |
403 | | OUStringBuffer & buffer, sal_Int32 bufferStart, bool precedingSlash, |
404 | | sal_Unicode const * pathBegin, sal_Unicode const * pathEnd) |
405 | 3.29k | { |
406 | 13.7k | while (precedingSlash || pathBegin != pathEnd) |
407 | 10.4k | { |
408 | 10.4k | sal_Unicode const * p = pathBegin; |
409 | 122k | while (p != pathEnd && *p != '/') |
410 | 112k | { |
411 | 112k | ++p; |
412 | 112k | } |
413 | | |
414 | 10.4k | std::size_t n = p - pathBegin; |
415 | 10.4k | if (n == 1 && pathBegin[0] == '.') |
416 | 0 | { |
417 | | // input begins with "." -> remove from input (and done): |
418 | | // i.e., !precedingSlash -> !precedingSlash |
419 | | // input begins with "./" -> remove from input: |
420 | | // i.e., !precedingSlash -> !precedingSlash |
421 | | // input begins with "/." -> replace with "/" in input (and not yet |
422 | | // done): |
423 | | // i.e., precedingSlash -> precedingSlash |
424 | | // input begins with "/./" -> replace with "/" in input: |
425 | | // i.e., precedingSlash -> precedingSlash |
426 | 0 | } |
427 | 10.4k | else if (n == 2 && pathBegin[0] == '.' && pathBegin[1] == '.') |
428 | 0 | { |
429 | | // input begins with ".." -> remove from input (and done): |
430 | | // i.e., !precedingSlash -> !precedingSlash |
431 | | // input begins with "../" -> remove from input |
432 | | // i.e., !precedingSlash -> !precedingSlash |
433 | | // input begins with "/.." -> replace with "/" in input, and shrink |
434 | | // output (not yet done): |
435 | | // i.e., precedingSlash -> precedingSlash |
436 | | // input begins with "/../" -> replace with "/" in input, and shrink |
437 | | // output: |
438 | | // i.e., precedingSlash -> precedingSlash |
439 | 0 | if (precedingSlash) |
440 | 0 | { |
441 | 0 | buffer.truncate( |
442 | 0 | bufferStart |
443 | 0 | + std::max<sal_Int32>( |
444 | 0 | rtl_ustr_lastIndexOfChar_WithLength( |
445 | 0 | buffer.getStr() + bufferStart, |
446 | 0 | buffer.getLength() - bufferStart, '/'), |
447 | 0 | 0)); |
448 | 0 | } |
449 | 0 | } |
450 | 10.4k | else |
451 | 10.4k | { |
452 | 10.4k | if (precedingSlash) |
453 | 7.17k | buffer.append('/'); |
454 | | |
455 | 10.4k | buffer.append(pathBegin, n); |
456 | 10.4k | precedingSlash = p != pathEnd; |
457 | 10.4k | } |
458 | 10.4k | pathBegin = p + (p == pathEnd ? 0 : 1); |
459 | 10.4k | } |
460 | 3.29k | } |
461 | | |
462 | | } |
463 | | |
464 | | sal_Bool const * SAL_CALL rtl_getUriCharClass(rtl_UriCharClass eCharClass) noexcept |
465 | 492 | { |
466 | 492 | static constexpr std::array<sal_Bool, rtl::UriCharClassSize> aCharClass[] = { |
467 | 492 | rtl::createUriCharClass(u8""), // None |
468 | 492 | rtl::createUriCharClass( |
469 | 492 | u8"!$&'()*+,-./:;=?@[]_~" |
470 | 492 | "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"), // Uric |
471 | 492 | rtl::createUriCharClass( |
472 | 492 | u8"!$&'()*+,-.:;=?@_~" |
473 | 492 | "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"), // UricNoSlash |
474 | 492 | rtl::createUriCharClass( |
475 | 492 | u8"!$&'()*+,-.;=@_~" |
476 | 492 | "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"), // RelSegment |
477 | 492 | rtl::createUriCharClass( |
478 | 492 | u8"!$&'()*+,-.:;=@_~" |
479 | 492 | "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"), // RegName |
480 | 492 | rtl::createUriCharClass( |
481 | 492 | u8"!$&'()*+,-.:;=_~" |
482 | 492 | "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"), // Userinfo |
483 | 492 | rtl::createUriCharClass( |
484 | 492 | u8"!$&'()*+,-.:=@_~" |
485 | 492 | "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz"), // Pchar |
486 | 492 | rtl::createUriCharClass( |
487 | 492 | u8"!$&'()*+-./:?@_~" |
488 | 492 | "0123456789ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz")}; // UnoParamValue |
489 | | |
490 | 492 | assert( |
491 | 492 | (eCharClass >= 0 |
492 | 492 | && (sal::static_int_cast< std::size_t >(eCharClass) |
493 | 492 | < SAL_N_ELEMENTS(aCharClass)))); // bad eCharClass |
494 | 492 | return aCharClass[eCharClass].data(); |
495 | 492 | } |
496 | | |
497 | | void SAL_CALL rtl_uriEncode(rtl_uString * pText, sal_Bool const * pCharClass, |
498 | | rtl_UriEncodeMechanism eMechanism, |
499 | | rtl_TextEncoding eCharset, rtl_uString ** pResult) noexcept |
500 | 890k | { |
501 | 890k | assert(!pCharClass[0x25]); // make sure the percent sign is encoded... |
502 | | |
503 | 890k | sal_Unicode const * p = pText->buffer; |
504 | 890k | sal_Unicode const * pEnd = p + pText->length; |
505 | 890k | sal_Int32 nCapacity = 256; |
506 | 890k | rtl_uString_new_WithLength(pResult, nCapacity); |
507 | | |
508 | 30.2M | while (p < pEnd) |
509 | 29.3M | { |
510 | 29.3M | rtl::uri::detail::EscapeType eType; |
511 | 29.3M | sal_uInt32 nUtf32 = rtl::uri::detail::readUcs4( |
512 | 29.3M | &p, pEnd, |
513 | 29.3M | (eMechanism == rtl_UriEncodeKeepEscapes |
514 | 29.3M | || eMechanism == rtl_UriEncodeCheckEscapes |
515 | 29.3M | || eMechanism == rtl_UriEncodeStrictKeepEscapes), |
516 | 29.3M | eMechanism == rtl_UriEncodeStrictKeepEscapes ? RTL_TEXTENCODING_UTF8 : eCharset, |
517 | 29.3M | &eType); |
518 | | |
519 | 29.3M | switch (eType) |
520 | 29.3M | { |
521 | 29.3M | case rtl::uri::detail::EscapeNo: |
522 | 29.3M | if (isValid(pCharClass, nUtf32)) // implies nUtf32 <= 0x7F |
523 | 28.1M | { |
524 | 28.1M | writeUnicode(pResult, &nCapacity, |
525 | 28.1M | static_cast< sal_Unicode >(nUtf32)); |
526 | 28.1M | } |
527 | 1.25M | else if (!writeEscapeChar( |
528 | 1.25M | pResult, &nCapacity, nUtf32, eCharset, |
529 | 1.25M | (eMechanism == rtl_UriEncodeStrict |
530 | 1.25M | || eMechanism == rtl_UriEncodeStrictKeepEscapes))) |
531 | 0 | { |
532 | 0 | rtl_uString_new(pResult); |
533 | 0 | return; |
534 | 0 | } |
535 | 29.3M | break; |
536 | | |
537 | 29.3M | case rtl::uri::detail::EscapeChar: |
538 | 0 | if (eMechanism == rtl_UriEncodeCheckEscapes |
539 | 0 | && isValid(pCharClass, nUtf32)) // implies nUtf32 <= 0x7F |
540 | 0 | { |
541 | 0 | writeUnicode(pResult, &nCapacity, |
542 | 0 | static_cast< sal_Unicode >(nUtf32)); |
543 | 0 | } |
544 | 0 | else if (!writeEscapeChar( |
545 | 0 | pResult, &nCapacity, nUtf32, eCharset, |
546 | 0 | (eMechanism == rtl_UriEncodeStrict |
547 | 0 | || eMechanism == rtl_UriEncodeStrictKeepEscapes))) |
548 | 0 | { |
549 | 0 | rtl_uString_new(pResult); |
550 | 0 | return; |
551 | 0 | } |
552 | 0 | break; |
553 | | |
554 | 0 | case rtl::uri::detail::EscapeOctet: |
555 | 0 | writeEscapeOctet(pResult, &nCapacity, nUtf32); |
556 | 0 | break; |
557 | 29.3M | } |
558 | 29.3M | } |
559 | 890k | *pResult = rtl_uStringBuffer_makeStringAndClear(pResult, &nCapacity); |
560 | 890k | } |
561 | | |
562 | | void SAL_CALL rtl_uriDecode(rtl_uString * pText, |
563 | | rtl_UriDecodeMechanism eMechanism, |
564 | | rtl_TextEncoding eCharset, rtl_uString ** pResult) noexcept |
565 | 799k | { |
566 | 799k | switch (eMechanism) |
567 | 799k | { |
568 | 0 | case rtl_UriDecodeNone: |
569 | 0 | rtl_uString_assign(pResult, pText); |
570 | 0 | break; |
571 | | |
572 | 0 | case rtl_UriDecodeToIuri: |
573 | 0 | eCharset = RTL_TEXTENCODING_UTF8; |
574 | 0 | [[fallthrough]]; |
575 | 799k | default: // rtl_UriDecodeWithCharset, rtl_UriDecodeStrict |
576 | 799k | { |
577 | 799k | sal_Unicode const * p = pText->buffer; |
578 | 799k | sal_Unicode const * pEnd = p + pText->length; |
579 | 799k | sal_Int32 nCapacity = pText->length; |
580 | 799k | rtl_uString_new_WithLength(pResult, nCapacity); |
581 | | |
582 | 26.0M | while (p < pEnd) |
583 | 25.2M | { |
584 | 25.2M | rtl::uri::detail::EscapeType eType; |
585 | 25.2M | sal_uInt32 nUtf32 = rtl::uri::detail::readUcs4(&p, pEnd, true, eCharset, &eType); |
586 | 25.2M | switch (eType) |
587 | 25.2M | { |
588 | 18.0k | case rtl::uri::detail::EscapeChar: |
589 | 18.0k | if (nUtf32 <= 0x7F && eMechanism == rtl_UriDecodeToIuri) |
590 | 0 | { |
591 | 0 | writeEscapeOctet(pResult, &nCapacity, nUtf32); |
592 | 0 | break; |
593 | 0 | } |
594 | 18.0k | [[fallthrough]]; |
595 | | |
596 | 25.2M | case rtl::uri::detail::EscapeNo: |
597 | 25.2M | writeUcs4(pResult, &nCapacity, nUtf32); |
598 | 25.2M | break; |
599 | | |
600 | 15 | case rtl::uri::detail::EscapeOctet: |
601 | 15 | if (eMechanism == rtl_UriDecodeStrict) |
602 | 0 | { |
603 | 0 | rtl_uString_new(pResult); |
604 | 0 | return; |
605 | 0 | } |
606 | 15 | writeEscapeOctet(pResult, &nCapacity, nUtf32); |
607 | 15 | break; |
608 | 25.2M | } |
609 | 25.2M | } |
610 | | |
611 | 799k | *pResult = rtl_uStringBuffer_makeStringAndClear( pResult, &nCapacity ); |
612 | 799k | } |
613 | 0 | break; |
614 | 799k | } |
615 | 799k | } |
616 | | |
617 | | sal_Bool SAL_CALL rtl_uriConvertRelToAbs(rtl_uString * pBaseUriRef, |
618 | | rtl_uString * pRelUriRef, |
619 | | rtl_uString ** pResult, |
620 | | rtl_uString ** pException) noexcept |
621 | 13.7k | { |
622 | | // Use the strict parser algorithm from RFC 3986, section 5.2, to turn the |
623 | | // relative URI into an absolute one: |
624 | 13.7k | Components aRelComponents; |
625 | 13.7k | parseUriRef(pRelUriRef, &aRelComponents); |
626 | 13.7k | OUStringBuffer aBuffer(256); |
627 | | |
628 | 13.7k | if (aRelComponents.aScheme.isPresent()) |
629 | 3.28k | { |
630 | 3.28k | aBuffer.append(aRelComponents.aScheme.pBegin, |
631 | 3.28k | aRelComponents.aScheme.getLength()); |
632 | | |
633 | 3.28k | if (aRelComponents.aAuthority.isPresent()) |
634 | 2.42k | { |
635 | 2.42k | aBuffer.append(aRelComponents.aAuthority.pBegin, |
636 | 2.42k | aRelComponents.aAuthority.getLength()); |
637 | 2.42k | } |
638 | | |
639 | 3.28k | appendPath( |
640 | 3.28k | aBuffer, aBuffer.getLength(), false, aRelComponents.aPath.pBegin, |
641 | 3.28k | aRelComponents.aPath.pEnd); |
642 | | |
643 | 3.28k | if (aRelComponents.aQuery.isPresent()) |
644 | 421 | { |
645 | 421 | aBuffer.append(aRelComponents.aQuery.pBegin, |
646 | 421 | aRelComponents.aQuery.getLength()); |
647 | 421 | } |
648 | 3.28k | } |
649 | 10.4k | else |
650 | 10.4k | { |
651 | 10.4k | Components aBaseComponents; |
652 | 10.4k | parseUriRef(pBaseUriRef, &aBaseComponents); |
653 | 10.4k | if (!aBaseComponents.aScheme.isPresent()) |
654 | 9.69k | { |
655 | 9.69k | rtl_uString_assign( |
656 | 9.69k | pException, |
657 | 9.69k | (OUString( |
658 | 9.69k | "<" + OUString::unacquired(&pBaseUriRef) |
659 | 9.69k | + "> does not start with a scheme component") |
660 | 9.69k | .pData)); |
661 | 9.69k | return false; |
662 | 9.69k | } |
663 | | |
664 | 754 | aBuffer.append(aBaseComponents.aScheme.pBegin, |
665 | 754 | aBaseComponents.aScheme.getLength()); |
666 | 754 | if (aRelComponents.aAuthority.isPresent()) |
667 | 0 | { |
668 | 0 | aBuffer.append(aRelComponents.aAuthority.pBegin, |
669 | 0 | aRelComponents.aAuthority.getLength()); |
670 | 0 | appendPath( |
671 | 0 | aBuffer, aBuffer.getLength(), false, |
672 | 0 | aRelComponents.aPath.pBegin, aRelComponents.aPath.pEnd); |
673 | |
|
674 | 0 | if (aRelComponents.aQuery.isPresent()) |
675 | 0 | { |
676 | 0 | aBuffer.append(aRelComponents.aQuery.pBegin, |
677 | 0 | aRelComponents.aQuery.getLength()); |
678 | 0 | } |
679 | 0 | } |
680 | 754 | else |
681 | 754 | { |
682 | 754 | if (aBaseComponents.aAuthority.isPresent()) |
683 | 350 | { |
684 | 350 | aBuffer.append(aBaseComponents.aAuthority.pBegin, |
685 | 350 | aBaseComponents.aAuthority.getLength()); |
686 | 350 | } |
687 | | |
688 | 754 | if (aRelComponents.aPath.pBegin == aRelComponents.aPath.pEnd) |
689 | 751 | { |
690 | 751 | aBuffer.append(aBaseComponents.aPath.pBegin, |
691 | 751 | aBaseComponents.aPath.getLength()); |
692 | 751 | if (aRelComponents.aQuery.isPresent()) |
693 | 0 | { |
694 | 0 | aBuffer.append(aRelComponents.aQuery.pBegin, |
695 | 0 | aRelComponents.aQuery.getLength()); |
696 | 0 | } |
697 | 751 | else if (aBaseComponents.aQuery.isPresent()) |
698 | 86 | { |
699 | 86 | aBuffer.append(aBaseComponents.aQuery.pBegin, |
700 | 86 | aBaseComponents.aQuery.getLength()); |
701 | 86 | } |
702 | 751 | } |
703 | 3 | else |
704 | 3 | { |
705 | 3 | if (*aRelComponents.aPath.pBegin == '/') |
706 | 0 | { |
707 | 0 | appendPath( |
708 | 0 | aBuffer, aBuffer.getLength(), false, |
709 | 0 | aRelComponents.aPath.pBegin, aRelComponents.aPath.pEnd); |
710 | 0 | } |
711 | 3 | else if (aBaseComponents.aAuthority.isPresent() |
712 | 3 | && aBaseComponents.aPath.pBegin |
713 | 3 | == aBaseComponents.aPath.pEnd) |
714 | 3 | { |
715 | 3 | appendPath( |
716 | 3 | aBuffer, aBuffer.getLength(), true, |
717 | 3 | aRelComponents.aPath.pBegin, aRelComponents.aPath.pEnd); |
718 | 3 | } |
719 | 0 | else |
720 | 0 | { |
721 | 0 | sal_Int32 n = aBuffer.getLength(); |
722 | 0 | sal_Int32 i = rtl_ustr_lastIndexOfChar_WithLength( |
723 | 0 | aBaseComponents.aPath.pBegin, |
724 | 0 | aBaseComponents.aPath.getLength(), '/'); |
725 | |
|
726 | 0 | if (i >= 0) |
727 | 0 | { |
728 | 0 | appendPath( |
729 | 0 | aBuffer, n, false, aBaseComponents.aPath.pBegin, |
730 | 0 | aBaseComponents.aPath.pBegin + i); |
731 | 0 | } |
732 | |
|
733 | 0 | appendPath( |
734 | 0 | aBuffer, n, i >= 0, aRelComponents.aPath.pBegin, |
735 | 0 | aRelComponents.aPath.pEnd); |
736 | 0 | } |
737 | | |
738 | 3 | if (aRelComponents.aQuery.isPresent()) |
739 | 0 | { |
740 | 0 | aBuffer.append(aRelComponents.aQuery.pBegin, |
741 | 0 | aRelComponents.aQuery.getLength()); |
742 | 0 | } |
743 | 3 | } |
744 | 754 | } |
745 | 754 | } |
746 | 4.04k | if (aRelComponents.aFragment.isPresent()) |
747 | 36 | { |
748 | 36 | aBuffer.append(aRelComponents.aFragment.pBegin, |
749 | 36 | aRelComponents.aFragment.getLength()); |
750 | 36 | } |
751 | | |
752 | 4.04k | rtl_uString_assign(pResult, aBuffer.makeStringAndClear().pData); |
753 | 4.04k | return true; |
754 | 13.7k | } |
755 | | |
756 | | /* vim:set shiftwidth=4 softtabstop=4 expandtab: */ |