Mercurial > hg > orthanc-stone
comparison Resources/Orthanc/Core/Toolbox.cpp @ 200:03afbee0cc7b
integration of Orthanc core into Stone
author | Sebastien Jodogne <s.jodogne@gmail.com> |
---|---|
date | Fri, 23 Mar 2018 11:04:03 +0100 |
parents | |
children | 795d71f66f31 |
comparison
equal
deleted
inserted
replaced
199:dabe9982fca3 | 200:03afbee0cc7b |
---|---|
1 /** | |
2 * Orthanc - A Lightweight, RESTful DICOM Store | |
3 * Copyright (C) 2012-2016 Sebastien Jodogne, Medical Physics | |
4 * Department, University Hospital of Liege, Belgium | |
5 * Copyright (C) 2017-2018 Osimis S.A., Belgium | |
6 * | |
7 * This program is free software: you can redistribute it and/or | |
8 * modify it under the terms of the GNU General Public License as | |
9 * published by the Free Software Foundation, either version 3 of the | |
10 * License, or (at your option) any later version. | |
11 * | |
12 * In addition, as a special exception, the copyright holders of this | |
13 * program give permission to link the code of its release with the | |
14 * OpenSSL project's "OpenSSL" library (or with modified versions of it | |
15 * that use the same license as the "OpenSSL" library), and distribute | |
16 * the linked executables. You must obey the GNU General Public License | |
17 * in all respects for all of the code used other than "OpenSSL". If you | |
18 * modify file(s) with this exception, you may extend this exception to | |
19 * your version of the file(s), but you are not obligated to do so. If | |
20 * you do not wish to do so, delete this exception statement from your | |
21 * version. If you delete this exception statement from all source files | |
22 * in the program, then also delete it here. | |
23 * | |
24 * This program is distributed in the hope that it will be useful, but | |
25 * WITHOUT ANY WARRANTY; without even the implied warranty of | |
26 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU | |
27 * General Public License for more details. | |
28 * | |
29 * You should have received a copy of the GNU General Public License | |
30 * along with this program. If not, see <http://www.gnu.org/licenses/>. | |
31 **/ | |
32 | |
33 | |
34 #include "PrecompiledHeaders.h" | |
35 #include "Toolbox.h" | |
36 | |
37 #include "OrthancException.h" | |
38 #include "Logging.h" | |
39 | |
40 #include <boost/algorithm/string/case_conv.hpp> | |
41 #include <boost/algorithm/string/replace.hpp> | |
42 #include <boost/lexical_cast.hpp> | |
43 #include <boost/regex.hpp> | |
44 #include <boost/uuid/sha1.hpp> | |
45 | |
46 #include <string> | |
47 #include <stdint.h> | |
48 #include <string.h> | |
49 #include <algorithm> | |
50 #include <ctype.h> | |
51 | |
52 | |
53 #if ORTHANC_ENABLE_MD5 == 1 | |
54 # include "../Resources/ThirdParty/md5/md5.h" | |
55 #endif | |
56 | |
57 #if ORTHANC_ENABLE_BASE64 == 1 | |
58 # include "../Resources/ThirdParty/base64/base64.h" | |
59 #endif | |
60 | |
61 #if ORTHANC_ENABLE_LOCALE == 1 | |
62 # include <boost/locale.hpp> | |
63 #endif | |
64 | |
65 | |
66 #if defined(_MSC_VER) && (_MSC_VER < 1800) | |
67 // Patch for the missing "_strtoll" symbol when compiling with Visual Studio < 2013 | |
68 extern "C" | |
69 { | |
70 int64_t _strtoi64(const char *nptr, char **endptr, int base); | |
71 int64_t strtoll(const char *nptr, char **endptr, int base) | |
72 { | |
73 return _strtoi64(nptr, endptr, base); | |
74 } | |
75 } | |
76 #endif | |
77 | |
78 | |
79 #if defined(_WIN32) | |
80 # include <windows.h> // For ::Sleep | |
81 #endif | |
82 | |
83 | |
84 #if ORTHANC_ENABLE_PUGIXML == 1 | |
85 # include "ChunkedBuffer.h" | |
86 # include <pugixml.hpp> | |
87 #endif | |
88 | |
89 | |
90 namespace Orthanc | |
91 { | |
92 void Toolbox::ToUpperCase(std::string& s) | |
93 { | |
94 std::transform(s.begin(), s.end(), s.begin(), toupper); | |
95 } | |
96 | |
97 | |
98 void Toolbox::ToLowerCase(std::string& s) | |
99 { | |
100 std::transform(s.begin(), s.end(), s.begin(), tolower); | |
101 } | |
102 | |
103 | |
104 void Toolbox::ToUpperCase(std::string& result, | |
105 const std::string& source) | |
106 { | |
107 result = source; | |
108 ToUpperCase(result); | |
109 } | |
110 | |
111 void Toolbox::ToLowerCase(std::string& result, | |
112 const std::string& source) | |
113 { | |
114 result = source; | |
115 ToLowerCase(result); | |
116 } | |
117 | |
118 | |
119 void Toolbox::SplitUriComponents(UriComponents& components, | |
120 const std::string& uri) | |
121 { | |
122 static const char URI_SEPARATOR = '/'; | |
123 | |
124 components.clear(); | |
125 | |
126 if (uri.size() == 0 || | |
127 uri[0] != URI_SEPARATOR) | |
128 { | |
129 throw OrthancException(ErrorCode_UriSyntax); | |
130 } | |
131 | |
132 // Count the number of slashes in the URI to make an assumption | |
133 // about the number of components in the URI | |
134 unsigned int estimatedSize = 0; | |
135 for (unsigned int i = 0; i < uri.size(); i++) | |
136 { | |
137 if (uri[i] == URI_SEPARATOR) | |
138 estimatedSize++; | |
139 } | |
140 | |
141 components.reserve(estimatedSize - 1); | |
142 | |
143 unsigned int start = 1; | |
144 unsigned int end = 1; | |
145 while (end < uri.size()) | |
146 { | |
147 // This is the loop invariant | |
148 assert(uri[start - 1] == '/' && (end >= start)); | |
149 | |
150 if (uri[end] == '/') | |
151 { | |
152 components.push_back(std::string(&uri[start], end - start)); | |
153 end++; | |
154 start = end; | |
155 } | |
156 else | |
157 { | |
158 end++; | |
159 } | |
160 } | |
161 | |
162 if (start < uri.size()) | |
163 { | |
164 components.push_back(std::string(&uri[start], end - start)); | |
165 } | |
166 | |
167 for (size_t i = 0; i < components.size(); i++) | |
168 { | |
169 if (components[i].size() == 0) | |
170 { | |
171 // Empty component, as in: "/coucou//e" | |
172 throw OrthancException(ErrorCode_UriSyntax); | |
173 } | |
174 } | |
175 } | |
176 | |
177 | |
178 void Toolbox::TruncateUri(UriComponents& target, | |
179 const UriComponents& source, | |
180 size_t fromLevel) | |
181 { | |
182 target.clear(); | |
183 | |
184 if (source.size() > fromLevel) | |
185 { | |
186 target.resize(source.size() - fromLevel); | |
187 | |
188 size_t j = 0; | |
189 for (size_t i = fromLevel; i < source.size(); i++, j++) | |
190 { | |
191 target[j] = source[i]; | |
192 } | |
193 | |
194 assert(j == target.size()); | |
195 } | |
196 } | |
197 | |
198 | |
199 | |
200 bool Toolbox::IsChildUri(const UriComponents& baseUri, | |
201 const UriComponents& testedUri) | |
202 { | |
203 if (testedUri.size() < baseUri.size()) | |
204 { | |
205 return false; | |
206 } | |
207 | |
208 for (size_t i = 0; i < baseUri.size(); i++) | |
209 { | |
210 if (baseUri[i] != testedUri[i]) | |
211 return false; | |
212 } | |
213 | |
214 return true; | |
215 } | |
216 | |
217 | |
218 std::string Toolbox::AutodetectMimeType(const std::string& path) | |
219 { | |
220 std::string contentType; | |
221 size_t lastDot = path.rfind('.'); | |
222 size_t lastSlash = path.rfind('/'); | |
223 | |
224 if (lastDot == std::string::npos || | |
225 (lastSlash != std::string::npos && lastDot < lastSlash)) | |
226 { | |
227 // No trailing dot, unable to detect the content type | |
228 } | |
229 else | |
230 { | |
231 const char* extension = &path[lastDot + 1]; | |
232 | |
233 // http://en.wikipedia.org/wiki/Mime_types | |
234 // Text types | |
235 if (!strcmp(extension, "txt")) | |
236 contentType = "text/plain"; | |
237 else if (!strcmp(extension, "html")) | |
238 contentType = "text/html"; | |
239 else if (!strcmp(extension, "xml")) | |
240 contentType = "text/xml"; | |
241 else if (!strcmp(extension, "css")) | |
242 contentType = "text/css"; | |
243 | |
244 // Application types | |
245 else if (!strcmp(extension, "js")) | |
246 contentType = "application/javascript"; | |
247 else if (!strcmp(extension, "json")) | |
248 contentType = "application/json"; | |
249 else if (!strcmp(extension, "pdf")) | |
250 contentType = "application/pdf"; | |
251 | |
252 // Images types | |
253 else if (!strcmp(extension, "jpg") || !strcmp(extension, "jpeg")) | |
254 contentType = "image/jpeg"; | |
255 else if (!strcmp(extension, "gif")) | |
256 contentType = "image/gif"; | |
257 else if (!strcmp(extension, "png")) | |
258 contentType = "image/png"; | |
259 } | |
260 | |
261 return contentType; | |
262 } | |
263 | |
264 | |
265 std::string Toolbox::FlattenUri(const UriComponents& components, | |
266 size_t fromLevel) | |
267 { | |
268 if (components.size() <= fromLevel) | |
269 { | |
270 return "/"; | |
271 } | |
272 else | |
273 { | |
274 std::string r; | |
275 | |
276 for (size_t i = fromLevel; i < components.size(); i++) | |
277 { | |
278 r += "/" + components[i]; | |
279 } | |
280 | |
281 return r; | |
282 } | |
283 } | |
284 | |
285 | |
286 #if ORTHANC_ENABLE_MD5 == 1 | |
287 static char GetHexadecimalCharacter(uint8_t value) | |
288 { | |
289 assert(value < 16); | |
290 | |
291 if (value < 10) | |
292 { | |
293 return value + '0'; | |
294 } | |
295 else | |
296 { | |
297 return (value - 10) + 'a'; | |
298 } | |
299 } | |
300 | |
301 | |
302 void Toolbox::ComputeMD5(std::string& result, | |
303 const std::string& data) | |
304 { | |
305 if (data.size() > 0) | |
306 { | |
307 ComputeMD5(result, &data[0], data.size()); | |
308 } | |
309 else | |
310 { | |
311 ComputeMD5(result, NULL, 0); | |
312 } | |
313 } | |
314 | |
315 | |
316 void Toolbox::ComputeMD5(std::string& result, | |
317 const void* data, | |
318 size_t size) | |
319 { | |
320 md5_state_s state; | |
321 md5_init(&state); | |
322 | |
323 if (size > 0) | |
324 { | |
325 md5_append(&state, | |
326 reinterpret_cast<const md5_byte_t*>(data), | |
327 static_cast<int>(size)); | |
328 } | |
329 | |
330 md5_byte_t actualHash[16]; | |
331 md5_finish(&state, actualHash); | |
332 | |
333 result.resize(32); | |
334 for (unsigned int i = 0; i < 16; i++) | |
335 { | |
336 result[2 * i] = GetHexadecimalCharacter(static_cast<uint8_t>(actualHash[i] / 16)); | |
337 result[2 * i + 1] = GetHexadecimalCharacter(static_cast<uint8_t>(actualHash[i] % 16)); | |
338 } | |
339 } | |
340 #endif | |
341 | |
342 | |
343 #if ORTHANC_ENABLE_BASE64 == 1 | |
344 void Toolbox::EncodeBase64(std::string& result, | |
345 const std::string& data) | |
346 { | |
347 result = base64_encode(data); | |
348 } | |
349 | |
350 void Toolbox::DecodeBase64(std::string& result, | |
351 const std::string& data) | |
352 { | |
353 for (size_t i = 0; i < data.length(); i++) | |
354 { | |
355 if (!isalnum(data[i]) && | |
356 data[i] != '+' && | |
357 data[i] != '/' && | |
358 data[i] != '=') | |
359 { | |
360 // This is not a valid character for a Base64 string | |
361 throw OrthancException(ErrorCode_BadFileFormat); | |
362 } | |
363 } | |
364 | |
365 result = base64_decode(data); | |
366 } | |
367 | |
368 | |
369 bool Toolbox::DecodeDataUriScheme(std::string& mime, | |
370 std::string& content, | |
371 const std::string& source) | |
372 { | |
373 boost::regex pattern("data:([^;]+);base64,([a-zA-Z0-9=+/]*)", | |
374 boost::regex::icase /* case insensitive search */); | |
375 | |
376 boost::cmatch what; | |
377 if (regex_match(source.c_str(), what, pattern)) | |
378 { | |
379 mime = what[1]; | |
380 DecodeBase64(content, what[2]); | |
381 return true; | |
382 } | |
383 else | |
384 { | |
385 return false; | |
386 } | |
387 } | |
388 | |
389 | |
390 void Toolbox::EncodeDataUriScheme(std::string& result, | |
391 const std::string& mime, | |
392 const std::string& content) | |
393 { | |
394 result = "data:" + mime + ";base64," + base64_encode(content); | |
395 } | |
396 | |
397 #endif | |
398 | |
399 | |
400 #if ORTHANC_ENABLE_LOCALE == 1 | |
401 static const char* GetBoostLocaleEncoding(const Encoding sourceEncoding) | |
402 { | |
403 switch (sourceEncoding) | |
404 { | |
405 case Encoding_Utf8: | |
406 return "UTF-8"; | |
407 | |
408 case Encoding_Ascii: | |
409 return "ASCII"; | |
410 | |
411 case Encoding_Latin1: | |
412 return "ISO-8859-1"; | |
413 break; | |
414 | |
415 case Encoding_Latin2: | |
416 return "ISO-8859-2"; | |
417 break; | |
418 | |
419 case Encoding_Latin3: | |
420 return "ISO-8859-3"; | |
421 break; | |
422 | |
423 case Encoding_Latin4: | |
424 return "ISO-8859-4"; | |
425 break; | |
426 | |
427 case Encoding_Latin5: | |
428 return "ISO-8859-9"; | |
429 break; | |
430 | |
431 case Encoding_Cyrillic: | |
432 return "ISO-8859-5"; | |
433 break; | |
434 | |
435 case Encoding_Windows1251: | |
436 return "WINDOWS-1251"; | |
437 break; | |
438 | |
439 case Encoding_Arabic: | |
440 return "ISO-8859-6"; | |
441 break; | |
442 | |
443 case Encoding_Greek: | |
444 return "ISO-8859-7"; | |
445 break; | |
446 | |
447 case Encoding_Hebrew: | |
448 return "ISO-8859-8"; | |
449 break; | |
450 | |
451 case Encoding_Japanese: | |
452 return "SHIFT-JIS"; | |
453 break; | |
454 | |
455 case Encoding_Chinese: | |
456 return "GB18030"; | |
457 break; | |
458 | |
459 case Encoding_Thai: | |
460 return "TIS620.2533-0"; | |
461 break; | |
462 | |
463 default: | |
464 throw OrthancException(ErrorCode_NotImplemented); | |
465 } | |
466 } | |
467 #endif | |
468 | |
469 | |
470 #if ORTHANC_ENABLE_LOCALE == 1 | |
471 std::string Toolbox::ConvertToUtf8(const std::string& source, | |
472 Encoding sourceEncoding) | |
473 { | |
474 if (sourceEncoding == Encoding_Utf8) | |
475 { | |
476 // Already in UTF-8: No conversion is required | |
477 return source; | |
478 } | |
479 | |
480 if (sourceEncoding == Encoding_Ascii) | |
481 { | |
482 return ConvertToAscii(source); | |
483 } | |
484 | |
485 const char* encoding = GetBoostLocaleEncoding(sourceEncoding); | |
486 | |
487 try | |
488 { | |
489 return boost::locale::conv::to_utf<char>(source, encoding); | |
490 } | |
491 catch (std::runtime_error&) | |
492 { | |
493 // Bad input string or bad encoding | |
494 return ConvertToAscii(source); | |
495 } | |
496 } | |
497 #endif | |
498 | |
499 | |
500 #if ORTHANC_ENABLE_LOCALE == 1 | |
501 std::string Toolbox::ConvertFromUtf8(const std::string& source, | |
502 Encoding targetEncoding) | |
503 { | |
504 if (targetEncoding == Encoding_Utf8) | |
505 { | |
506 // Already in UTF-8: No conversion is required | |
507 return source; | |
508 } | |
509 | |
510 if (targetEncoding == Encoding_Ascii) | |
511 { | |
512 return ConvertToAscii(source); | |
513 } | |
514 | |
515 const char* encoding = GetBoostLocaleEncoding(targetEncoding); | |
516 | |
517 try | |
518 { | |
519 return boost::locale::conv::from_utf<char>(source, encoding); | |
520 } | |
521 catch (std::runtime_error&) | |
522 { | |
523 // Bad input string or bad encoding | |
524 return ConvertToAscii(source); | |
525 } | |
526 } | |
527 #endif | |
528 | |
529 | |
530 bool Toolbox::IsAsciiString(const void* data, | |
531 size_t size) | |
532 { | |
533 const uint8_t* p = reinterpret_cast<const uint8_t*>(data); | |
534 | |
535 for (size_t i = 0; i < size; i++, p++) | |
536 { | |
537 if (*p > 127 || *p == 0 || iscntrl(*p)) | |
538 { | |
539 return false; | |
540 } | |
541 } | |
542 | |
543 return true; | |
544 } | |
545 | |
546 | |
547 bool Toolbox::IsAsciiString(const std::string& s) | |
548 { | |
549 return IsAsciiString(s.c_str(), s.size()); | |
550 } | |
551 | |
552 | |
553 std::string Toolbox::ConvertToAscii(const std::string& source) | |
554 { | |
555 std::string result; | |
556 | |
557 result.reserve(source.size() + 1); | |
558 for (size_t i = 0; i < source.size(); i++) | |
559 { | |
560 if (source[i] <= 127 && source[i] >= 0 && !iscntrl(source[i])) | |
561 { | |
562 result.push_back(source[i]); | |
563 } | |
564 } | |
565 | |
566 return result; | |
567 } | |
568 | |
569 | |
570 void Toolbox::ComputeSHA1(std::string& result, | |
571 const void* data, | |
572 size_t size) | |
573 { | |
574 boost::uuids::detail::sha1 sha1; | |
575 | |
576 if (size > 0) | |
577 { | |
578 sha1.process_bytes(data, size); | |
579 } | |
580 | |
581 unsigned int digest[5]; | |
582 | |
583 // Sanity check for the memory layout: A SHA-1 digest is 160 bits wide | |
584 assert(sizeof(unsigned int) == 4 && sizeof(digest) == (160 / 8)); | |
585 | |
586 sha1.get_digest(digest); | |
587 | |
588 result.resize(8 * 5 + 4); | |
589 sprintf(&result[0], "%08x-%08x-%08x-%08x-%08x", | |
590 digest[0], | |
591 digest[1], | |
592 digest[2], | |
593 digest[3], | |
594 digest[4]); | |
595 } | |
596 | |
597 void Toolbox::ComputeSHA1(std::string& result, | |
598 const std::string& data) | |
599 { | |
600 if (data.size() > 0) | |
601 { | |
602 ComputeSHA1(result, data.c_str(), data.size()); | |
603 } | |
604 else | |
605 { | |
606 ComputeSHA1(result, NULL, 0); | |
607 } | |
608 } | |
609 | |
610 | |
611 bool Toolbox::IsSHA1(const char* str, | |
612 size_t size) | |
613 { | |
614 if (size == 0) | |
615 { | |
616 return false; | |
617 } | |
618 | |
619 const char* start = str; | |
620 const char* end = str + size; | |
621 | |
622 // Trim the beginning of the string | |
623 while (start < end) | |
624 { | |
625 if (*start == '\0' || | |
626 isspace(*start)) | |
627 { | |
628 start++; | |
629 } | |
630 else | |
631 { | |
632 break; | |
633 } | |
634 } | |
635 | |
636 // Trim the trailing of the string | |
637 while (start < end) | |
638 { | |
639 if (*(end - 1) == '\0' || | |
640 isspace(*(end - 1))) | |
641 { | |
642 end--; | |
643 } | |
644 else | |
645 { | |
646 break; | |
647 } | |
648 } | |
649 | |
650 if (end - start != 44) | |
651 { | |
652 return false; | |
653 } | |
654 | |
655 for (unsigned int i = 0; i < 44; i++) | |
656 { | |
657 if (i == 8 || | |
658 i == 17 || | |
659 i == 26 || | |
660 i == 35) | |
661 { | |
662 if (start[i] != '-') | |
663 return false; | |
664 } | |
665 else | |
666 { | |
667 if (!isalnum(start[i])) | |
668 return false; | |
669 } | |
670 } | |
671 | |
672 return true; | |
673 } | |
674 | |
675 | |
676 bool Toolbox::IsSHA1(const std::string& s) | |
677 { | |
678 if (s.size() == 0) | |
679 { | |
680 return false; | |
681 } | |
682 else | |
683 { | |
684 return IsSHA1(s.c_str(), s.size()); | |
685 } | |
686 } | |
687 | |
688 | |
689 std::string Toolbox::StripSpaces(const std::string& source) | |
690 { | |
691 size_t first = 0; | |
692 | |
693 while (first < source.length() && | |
694 isspace(source[first])) | |
695 { | |
696 first++; | |
697 } | |
698 | |
699 if (first == source.length()) | |
700 { | |
701 // String containing only spaces | |
702 return ""; | |
703 } | |
704 | |
705 size_t last = source.length(); | |
706 while (last > first && | |
707 isspace(source[last - 1])) | |
708 { | |
709 last--; | |
710 } | |
711 | |
712 assert(first <= last); | |
713 return source.substr(first, last - first); | |
714 } | |
715 | |
716 | |
717 static char Hex2Dec(char c) | |
718 { | |
719 return ((c >= '0' && c <= '9') ? c - '0' : | |
720 ((c >= 'a' && c <= 'f') ? c - 'a' + 10 : c - 'A' + 10)); | |
721 } | |
722 | |
723 void Toolbox::UrlDecode(std::string& s) | |
724 { | |
725 // http://en.wikipedia.org/wiki/Percent-encoding | |
726 // http://www.w3schools.com/tags/ref_urlencode.asp | |
727 // http://stackoverflow.com/questions/154536/encode-decode-urls-in-c | |
728 | |
729 if (s.size() == 0) | |
730 { | |
731 return; | |
732 } | |
733 | |
734 size_t source = 0; | |
735 size_t target = 0; | |
736 | |
737 while (source < s.size()) | |
738 { | |
739 if (s[source] == '%' && | |
740 source + 2 < s.size() && | |
741 isalnum(s[source + 1]) && | |
742 isalnum(s[source + 2])) | |
743 { | |
744 s[target] = (Hex2Dec(s[source + 1]) << 4) | Hex2Dec(s[source + 2]); | |
745 source += 3; | |
746 target += 1; | |
747 } | |
748 else | |
749 { | |
750 if (s[source] == '+') | |
751 s[target] = ' '; | |
752 else | |
753 s[target] = s[source]; | |
754 | |
755 source++; | |
756 target++; | |
757 } | |
758 } | |
759 | |
760 s.resize(target); | |
761 } | |
762 | |
763 | |
764 Endianness Toolbox::DetectEndianness() | |
765 { | |
766 // http://sourceforge.net/p/predef/wiki/Endianness/ | |
767 | |
768 uint8_t buffer[4]; | |
769 | |
770 buffer[0] = 0x00; | |
771 buffer[1] = 0x01; | |
772 buffer[2] = 0x02; | |
773 buffer[3] = 0x03; | |
774 | |
775 switch (*((uint32_t *)buffer)) | |
776 { | |
777 case 0x00010203: | |
778 return Endianness_Big; | |
779 | |
780 case 0x03020100: | |
781 return Endianness_Little; | |
782 | |
783 default: | |
784 throw OrthancException(ErrorCode_NotImplemented); | |
785 } | |
786 } | |
787 | |
788 | |
789 std::string Toolbox::WildcardToRegularExpression(const std::string& source) | |
790 { | |
791 // TODO - Speed up this with a regular expression | |
792 | |
793 std::string result = source; | |
794 | |
795 // Escape all special characters | |
796 boost::replace_all(result, "\\", "\\\\"); | |
797 boost::replace_all(result, "^", "\\^"); | |
798 boost::replace_all(result, ".", "\\."); | |
799 boost::replace_all(result, "$", "\\$"); | |
800 boost::replace_all(result, "|", "\\|"); | |
801 boost::replace_all(result, "(", "\\("); | |
802 boost::replace_all(result, ")", "\\)"); | |
803 boost::replace_all(result, "[", "\\["); | |
804 boost::replace_all(result, "]", "\\]"); | |
805 boost::replace_all(result, "+", "\\+"); | |
806 boost::replace_all(result, "/", "\\/"); | |
807 boost::replace_all(result, "{", "\\{"); | |
808 boost::replace_all(result, "}", "\\}"); | |
809 | |
810 // Convert wildcards '*' and '?' to their regex equivalents | |
811 boost::replace_all(result, "?", "."); | |
812 boost::replace_all(result, "*", ".*"); | |
813 | |
814 return result; | |
815 } | |
816 | |
817 | |
818 void Toolbox::TokenizeString(std::vector<std::string>& result, | |
819 const std::string& value, | |
820 char separator) | |
821 { | |
822 result.clear(); | |
823 | |
824 std::string currentItem; | |
825 | |
826 for (size_t i = 0; i < value.size(); i++) | |
827 { | |
828 if (value[i] == separator) | |
829 { | |
830 result.push_back(currentItem); | |
831 currentItem.clear(); | |
832 } | |
833 else | |
834 { | |
835 currentItem.push_back(value[i]); | |
836 } | |
837 } | |
838 | |
839 result.push_back(currentItem); | |
840 } | |
841 | |
842 | |
843 #if ORTHANC_ENABLE_PUGIXML == 1 | |
844 class ChunkedBufferWriter : public pugi::xml_writer | |
845 { | |
846 private: | |
847 ChunkedBuffer buffer_; | |
848 | |
849 public: | |
850 virtual void write(const void *data, size_t size) | |
851 { | |
852 if (size > 0) | |
853 { | |
854 buffer_.AddChunk(reinterpret_cast<const char*>(data), size); | |
855 } | |
856 } | |
857 | |
858 void Flatten(std::string& s) | |
859 { | |
860 buffer_.Flatten(s); | |
861 } | |
862 }; | |
863 | |
864 | |
865 static void JsonToXmlInternal(pugi::xml_node& target, | |
866 const Json::Value& source, | |
867 const std::string& arrayElement) | |
868 { | |
869 // http://jsoncpp.sourceforge.net/value_8h_source.html#l00030 | |
870 | |
871 switch (source.type()) | |
872 { | |
873 case Json::nullValue: | |
874 { | |
875 target.append_child(pugi::node_pcdata).set_value("null"); | |
876 break; | |
877 } | |
878 | |
879 case Json::intValue: | |
880 { | |
881 std::string s = boost::lexical_cast<std::string>(source.asInt()); | |
882 target.append_child(pugi::node_pcdata).set_value(s.c_str()); | |
883 break; | |
884 } | |
885 | |
886 case Json::uintValue: | |
887 { | |
888 std::string s = boost::lexical_cast<std::string>(source.asUInt()); | |
889 target.append_child(pugi::node_pcdata).set_value(s.c_str()); | |
890 break; | |
891 } | |
892 | |
893 case Json::realValue: | |
894 { | |
895 std::string s = boost::lexical_cast<std::string>(source.asFloat()); | |
896 target.append_child(pugi::node_pcdata).set_value(s.c_str()); | |
897 break; | |
898 } | |
899 | |
900 case Json::stringValue: | |
901 { | |
902 target.append_child(pugi::node_pcdata).set_value(source.asString().c_str()); | |
903 break; | |
904 } | |
905 | |
906 case Json::booleanValue: | |
907 { | |
908 target.append_child(pugi::node_pcdata).set_value(source.asBool() ? "true" : "false"); | |
909 break; | |
910 } | |
911 | |
912 case Json::arrayValue: | |
913 { | |
914 for (Json::Value::ArrayIndex i = 0; i < source.size(); i++) | |
915 { | |
916 pugi::xml_node node = target.append_child(); | |
917 node.set_name(arrayElement.c_str()); | |
918 JsonToXmlInternal(node, source[i], arrayElement); | |
919 } | |
920 break; | |
921 } | |
922 | |
923 case Json::objectValue: | |
924 { | |
925 Json::Value::Members members = source.getMemberNames(); | |
926 | |
927 for (size_t i = 0; i < members.size(); i++) | |
928 { | |
929 pugi::xml_node node = target.append_child(); | |
930 node.set_name(members[i].c_str()); | |
931 JsonToXmlInternal(node, source[members[i]], arrayElement); | |
932 } | |
933 | |
934 break; | |
935 } | |
936 | |
937 default: | |
938 throw OrthancException(ErrorCode_NotImplemented); | |
939 } | |
940 } | |
941 | |
942 | |
943 void Toolbox::JsonToXml(std::string& target, | |
944 const Json::Value& source, | |
945 const std::string& rootElement, | |
946 const std::string& arrayElement) | |
947 { | |
948 pugi::xml_document doc; | |
949 | |
950 pugi::xml_node n = doc.append_child(rootElement.c_str()); | |
951 JsonToXmlInternal(n, source, arrayElement); | |
952 | |
953 pugi::xml_node decl = doc.prepend_child(pugi::node_declaration); | |
954 decl.append_attribute("version").set_value("1.0"); | |
955 decl.append_attribute("encoding").set_value("utf-8"); | |
956 | |
957 ChunkedBufferWriter writer; | |
958 doc.save(writer, " ", pugi::format_default, pugi::encoding_utf8); | |
959 writer.Flatten(target); | |
960 } | |
961 | |
962 #endif | |
963 | |
964 | |
965 | |
966 bool Toolbox::IsInteger(const std::string& str) | |
967 { | |
968 std::string s = StripSpaces(str); | |
969 | |
970 if (s.size() == 0) | |
971 { | |
972 return false; | |
973 } | |
974 | |
975 size_t pos = 0; | |
976 if (s[0] == '-') | |
977 { | |
978 if (s.size() == 1) | |
979 { | |
980 return false; | |
981 } | |
982 | |
983 pos = 1; | |
984 } | |
985 | |
986 while (pos < s.size()) | |
987 { | |
988 if (!isdigit(s[pos])) | |
989 { | |
990 return false; | |
991 } | |
992 | |
993 pos++; | |
994 } | |
995 | |
996 return true; | |
997 } | |
998 | |
999 | |
1000 void Toolbox::CopyJsonWithoutComments(Json::Value& target, | |
1001 const Json::Value& source) | |
1002 { | |
1003 switch (source.type()) | |
1004 { | |
1005 case Json::nullValue: | |
1006 target = Json::nullValue; | |
1007 break; | |
1008 | |
1009 case Json::intValue: | |
1010 target = source.asInt64(); | |
1011 break; | |
1012 | |
1013 case Json::uintValue: | |
1014 target = source.asUInt64(); | |
1015 break; | |
1016 | |
1017 case Json::realValue: | |
1018 target = source.asDouble(); | |
1019 break; | |
1020 | |
1021 case Json::stringValue: | |
1022 target = source.asString(); | |
1023 break; | |
1024 | |
1025 case Json::booleanValue: | |
1026 target = source.asBool(); | |
1027 break; | |
1028 | |
1029 case Json::arrayValue: | |
1030 { | |
1031 target = Json::arrayValue; | |
1032 for (Json::Value::ArrayIndex i = 0; i < source.size(); i++) | |
1033 { | |
1034 Json::Value& item = target.append(Json::nullValue); | |
1035 CopyJsonWithoutComments(item, source[i]); | |
1036 } | |
1037 | |
1038 break; | |
1039 } | |
1040 | |
1041 case Json::objectValue: | |
1042 { | |
1043 target = Json::objectValue; | |
1044 Json::Value::Members members = source.getMemberNames(); | |
1045 for (Json::Value::ArrayIndex i = 0; i < members.size(); i++) | |
1046 { | |
1047 const std::string item = members[i]; | |
1048 CopyJsonWithoutComments(target[item], source[item]); | |
1049 } | |
1050 | |
1051 break; | |
1052 } | |
1053 | |
1054 default: | |
1055 break; | |
1056 } | |
1057 } | |
1058 | |
1059 | |
1060 bool Toolbox::StartsWith(const std::string& str, | |
1061 const std::string& prefix) | |
1062 { | |
1063 if (str.size() < prefix.size()) | |
1064 { | |
1065 return false; | |
1066 } | |
1067 else | |
1068 { | |
1069 return str.compare(0, prefix.size(), prefix) == 0; | |
1070 } | |
1071 } | |
1072 | |
1073 | |
1074 static bool IsUnreservedCharacter(char c) | |
1075 { | |
1076 // This function checks whether "c" is an unserved character | |
1077 // wrt. an URI percent-encoding | |
1078 // https://en.wikipedia.org/wiki/Percent-encoding#Percent-encoding%5Fin%5Fa%5FURI | |
1079 | |
1080 return ((c >= 'A' && c <= 'Z') || | |
1081 (c >= 'a' && c <= 'z') || | |
1082 (c >= '0' && c <= '9') || | |
1083 c == '-' || | |
1084 c == '_' || | |
1085 c == '.' || | |
1086 c == '~'); | |
1087 } | |
1088 | |
1089 void Toolbox::UriEncode(std::string& target, | |
1090 const std::string& source) | |
1091 { | |
1092 // Estimate the length of the percent-encoded URI | |
1093 size_t length = 0; | |
1094 | |
1095 for (size_t i = 0; i < source.size(); i++) | |
1096 { | |
1097 if (IsUnreservedCharacter(source[i])) | |
1098 { | |
1099 length += 1; | |
1100 } | |
1101 else | |
1102 { | |
1103 // This character must be percent-encoded | |
1104 length += 3; | |
1105 } | |
1106 } | |
1107 | |
1108 target.clear(); | |
1109 target.reserve(length); | |
1110 | |
1111 for (size_t i = 0; i < source.size(); i++) | |
1112 { | |
1113 if (IsUnreservedCharacter(source[i])) | |
1114 { | |
1115 target.push_back(source[i]); | |
1116 } | |
1117 else | |
1118 { | |
1119 // This character must be percent-encoded | |
1120 uint8_t byte = static_cast<uint8_t>(source[i]); | |
1121 uint8_t a = byte >> 4; | |
1122 uint8_t b = byte & 0x0f; | |
1123 | |
1124 target.push_back('%'); | |
1125 target.push_back(a < 10 ? a + '0' : a - 10 + 'A'); | |
1126 target.push_back(b < 10 ? b + '0' : b - 10 + 'A'); | |
1127 } | |
1128 } | |
1129 } | |
1130 | |
1131 | |
1132 static bool HasField(const Json::Value& json, | |
1133 const std::string& key, | |
1134 Json::ValueType expectedType) | |
1135 { | |
1136 if (json.type() != Json::objectValue || | |
1137 !json.isMember(key)) | |
1138 { | |
1139 return false; | |
1140 } | |
1141 else if (json[key].type() == expectedType) | |
1142 { | |
1143 return true; | |
1144 } | |
1145 else | |
1146 { | |
1147 throw OrthancException(ErrorCode_BadParameterType); | |
1148 } | |
1149 } | |
1150 | |
1151 | |
1152 std::string Toolbox::GetJsonStringField(const Json::Value& json, | |
1153 const std::string& key, | |
1154 const std::string& defaultValue) | |
1155 { | |
1156 if (HasField(json, key, Json::stringValue)) | |
1157 { | |
1158 return json[key].asString(); | |
1159 } | |
1160 else | |
1161 { | |
1162 return defaultValue; | |
1163 } | |
1164 } | |
1165 | |
1166 | |
1167 bool Toolbox::GetJsonBooleanField(const ::Json::Value& json, | |
1168 const std::string& key, | |
1169 bool defaultValue) | |
1170 { | |
1171 if (HasField(json, key, Json::booleanValue)) | |
1172 { | |
1173 return json[key].asBool(); | |
1174 } | |
1175 else | |
1176 { | |
1177 return defaultValue; | |
1178 } | |
1179 } | |
1180 | |
1181 | |
1182 int Toolbox::GetJsonIntegerField(const ::Json::Value& json, | |
1183 const std::string& key, | |
1184 int defaultValue) | |
1185 { | |
1186 if (HasField(json, key, Json::intValue)) | |
1187 { | |
1188 return json[key].asInt(); | |
1189 } | |
1190 else | |
1191 { | |
1192 return defaultValue; | |
1193 } | |
1194 } | |
1195 | |
1196 | |
1197 unsigned int Toolbox::GetJsonUnsignedIntegerField(const ::Json::Value& json, | |
1198 const std::string& key, | |
1199 unsigned int defaultValue) | |
1200 { | |
1201 int v = GetJsonIntegerField(json, key, defaultValue); | |
1202 | |
1203 if (v < 0) | |
1204 { | |
1205 throw OrthancException(ErrorCode_ParameterOutOfRange); | |
1206 } | |
1207 else | |
1208 { | |
1209 return static_cast<unsigned int>(v); | |
1210 } | |
1211 } | |
1212 | |
1213 | |
1214 bool Toolbox::IsUuid(const std::string& str) | |
1215 { | |
1216 if (str.size() != 36) | |
1217 { | |
1218 return false; | |
1219 } | |
1220 | |
1221 for (size_t i = 0; i < str.length(); i++) | |
1222 { | |
1223 if (i == 8 || i == 13 || i == 18 || i == 23) | |
1224 { | |
1225 if (str[i] != '-') | |
1226 return false; | |
1227 } | |
1228 else | |
1229 { | |
1230 if (!isalnum(str[i])) | |
1231 return false; | |
1232 } | |
1233 } | |
1234 | |
1235 return true; | |
1236 } | |
1237 | |
1238 | |
1239 bool Toolbox::StartsWithUuid(const std::string& str) | |
1240 { | |
1241 if (str.size() < 36) | |
1242 { | |
1243 return false; | |
1244 } | |
1245 | |
1246 if (str.size() == 36) | |
1247 { | |
1248 return IsUuid(str); | |
1249 } | |
1250 | |
1251 assert(str.size() > 36); | |
1252 if (!isspace(str[36])) | |
1253 { | |
1254 return false; | |
1255 } | |
1256 | |
1257 return IsUuid(str.substr(0, 36)); | |
1258 } | |
1259 | |
1260 | |
1261 #if ORTHANC_ENABLE_LOCALE == 1 | |
1262 static std::auto_ptr<std::locale> globalLocale_; | |
1263 | |
1264 static bool SetGlobalLocale(const char* locale) | |
1265 { | |
1266 globalLocale_.reset(NULL); | |
1267 | |
1268 try | |
1269 { | |
1270 if (locale == NULL) | |
1271 { | |
1272 LOG(WARNING) << "Falling back to system-wide default locale"; | |
1273 globalLocale_.reset(new std::locale()); | |
1274 } | |
1275 else | |
1276 { | |
1277 LOG(INFO) << "Using locale: \"" << locale << "\" for case-insensitive comparison of strings"; | |
1278 globalLocale_.reset(new std::locale(locale)); | |
1279 } | |
1280 } | |
1281 catch (std::runtime_error&) | |
1282 { | |
1283 } | |
1284 | |
1285 return (globalLocale_.get() != NULL); | |
1286 } | |
1287 | |
1288 void Toolbox::InitializeGlobalLocale(const char* locale) | |
1289 { | |
1290 // Make Orthanc use English, United States locale | |
1291 // Linux: use "en_US.UTF-8" | |
1292 // Windows: use "" | |
1293 // Wine: use NULL | |
1294 | |
1295 #if defined(__MINGW32__) | |
1296 // Visibly, there is no support of locales in MinGW yet | |
1297 // http://mingw.5.n7.nabble.com/How-to-use-std-locale-global-with-MinGW-correct-td33048.html | |
1298 static const char* DEFAULT_LOCALE = NULL; | |
1299 #elif defined(_WIN32) | |
1300 // For Windows: use default locale (using "en_US" does not work) | |
1301 static const char* DEFAULT_LOCALE = ""; | |
1302 #else | |
1303 // For Linux & cie | |
1304 static const char* DEFAULT_LOCALE = "en_US.UTF-8"; | |
1305 #endif | |
1306 | |
1307 bool ok; | |
1308 | |
1309 if (locale == NULL) | |
1310 { | |
1311 ok = SetGlobalLocale(DEFAULT_LOCALE); | |
1312 | |
1313 #if defined(__MINGW32__) | |
1314 LOG(WARNING) << "This is a MinGW build, case-insensitive comparison of " | |
1315 << "strings with accents will not work outside of Wine"; | |
1316 #endif | |
1317 } | |
1318 else | |
1319 { | |
1320 ok = SetGlobalLocale(locale); | |
1321 } | |
1322 | |
1323 if (!ok && | |
1324 !SetGlobalLocale(NULL)) | |
1325 { | |
1326 LOG(ERROR) << "Cannot initialize global locale"; | |
1327 throw OrthancException(ErrorCode_InternalError); | |
1328 } | |
1329 | |
1330 } | |
1331 | |
1332 | |
1333 void Toolbox::FinalizeGlobalLocale() | |
1334 { | |
1335 globalLocale_.reset(); | |
1336 } | |
1337 | |
1338 | |
1339 std::string Toolbox::ToUpperCaseWithAccents(const std::string& source) | |
1340 { | |
1341 if (globalLocale_.get() == NULL) | |
1342 { | |
1343 LOG(ERROR) << "No global locale was set, call Toolbox::InitializeGlobalLocale()"; | |
1344 throw OrthancException(ErrorCode_BadSequenceOfCalls); | |
1345 } | |
1346 | |
1347 /** | |
1348 * A few notes about locales: | |
1349 * | |
1350 * (1) We don't use "case folding": | |
1351 * http://www.boost.org/doc/libs/1_64_0/libs/locale/doc/html/conversions.html | |
1352 * | |
1353 * Characters are made uppercase one by one. This is because, in | |
1354 * static builds, we are using iconv, which is visibly not | |
1355 * supported correctly (TODO: Understand why). Case folding seems | |
1356 * to be working correctly if using the default backend under | |
1357 * Linux (ICU or POSIX?). If one wishes to use case folding, one | |
1358 * would use: | |
1359 * | |
1360 * boost::locale::generator gen; | |
1361 * std::locale::global(gen(DEFAULT_LOCALE)); | |
1362 * return boost::locale::to_upper(source); | |
1363 * | |
1364 * (2) The function "boost::algorithm::to_upper_copy" does not | |
1365 * make use of the "std::locale::global()". We therefore create a | |
1366 * global variable "globalLocale_". | |
1367 * | |
1368 * (3) The variant of "boost::algorithm::to_upper_copy()" that | |
1369 * uses std::string does not work properly. We need to apply it | |
1370 * one wide strings (std::wstring). This explains the two calls to | |
1371 * "utf_to_utf" in order to convert to/from std::wstring. | |
1372 **/ | |
1373 | |
1374 std::wstring w = boost::locale::conv::utf_to_utf<wchar_t>(source); | |
1375 w = boost::algorithm::to_upper_copy<std::wstring>(w, *globalLocale_); | |
1376 return boost::locale::conv::utf_to_utf<char>(w); | |
1377 } | |
1378 #endif | |
1379 } |