1 // Copyright 2006 Nemanja Trifunovic
4 Permission is hereby granted, free of charge, to any person or organization
5 obtaining a copy of the software and accompanying documentation covered by
6 this license (the "Software") to use, reproduce, display, distribute,
7 execute, and transmit the Software, and to prepare derivative works of the
8 Software, and to permit third-parties to whom the Software is furnished to
9 do so, all subject to the following:
11 The copyright notices in the Software and this entire statement, including
12 the above license grant, this restriction and the following disclaimer,
13 must be included in all copies of the Software, in whole or in part, and
14 all derivative works of the Software, unless such copies or derivative
15 works are solely in the form of machine-executable object code generated by
16 a source language processor.
18 THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
19 IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
20 FITNESS FOR A PARTICULAR PURPOSE, TITLE AND NON-INFRINGEMENT. IN NO EVENT
21 SHALL THE COPYRIGHT HOLDERS OR ANYONE DISTRIBUTING THE SOFTWARE BE LIABLE
22 FOR ANY DAMAGES OR OTHER LIABILITY, WHETHER IN CONTRACT, TORT OR OTHERWISE,
23 ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
24 DEALINGS IN THE SOFTWARE.
28 #ifndef UTF8_FOR_CPP_CORE_H_2675DCD0_9480_4c0c_B92A_CC14C027B731
29 #define UTF8_FOR_CPP_CORE_H_2675DCD0_9480_4c0c_B92A_CC14C027B731
32 #include "../../BoostParts/boost/cstdint.hpp"
36 // The typedefs for 8-bit, 16-bit and 32-bit unsigned integers
37 // You may need to change them to match your system.
38 // These typedefs have the same names as ones from cstdint, or boost/cstdint
39 //typedef unsigned char uint8_t;
40 //typedef unsigned short uint16_t;
41 //typedef unsigned int uint32_t;
43 // Changed by Strahinja Markovic to use the boost versions
44 // of these (for portability).
45 typedef boost::uint8_t uint8_t;
46 typedef boost::uint16_t uint16_t;
47 typedef boost::uint32_t uint32_t;
49 // Helper code - not intended to be directly called by the library users. May be changed at any time
53 // Leading (high) surrogates: 0xd800 - 0xdbff
54 // Trailing (low) surrogates: 0xdc00 - 0xdfff
55 const uint16_t LEAD_SURROGATE_MIN = 0xd800u;
56 const uint16_t LEAD_SURROGATE_MAX = 0xdbffu;
57 const uint16_t TRAIL_SURROGATE_MIN = 0xdc00u;
58 const uint16_t TRAIL_SURROGATE_MAX = 0xdfffu;
59 const uint16_t LEAD_OFFSET = LEAD_SURROGATE_MIN - (0x10000 >> 10);
60 const uint32_t SURROGATE_OFFSET = 0x10000u - (LEAD_SURROGATE_MIN << 10) - TRAIL_SURROGATE_MIN;
62 // Maximum valid value for a Unicode code point
63 const uint32_t CODE_POINT_MAX = 0x0010ffffu;
65 template<typename octet_type>
66 inline uint8_t mask8(octet_type oc)
68 return static_cast<uint8_t>(0xff & oc);
70 template<typename u16_type>
71 inline uint16_t mask16(u16_type oc)
73 return static_cast<uint16_t>(0xffff & oc);
75 template<typename octet_type>
76 inline bool is_trail(octet_type oc)
78 return ((mask8(oc) >> 6) == 0x2);
81 template <typename u16>
82 inline bool is_lead_surrogate(u16 cp)
84 return (cp >= LEAD_SURROGATE_MIN && cp <= LEAD_SURROGATE_MAX);
87 template <typename u16>
88 inline bool is_trail_surrogate(u16 cp)
90 return (cp >= TRAIL_SURROGATE_MIN && cp <= TRAIL_SURROGATE_MAX);
93 template <typename u16>
94 inline bool is_surrogate(u16 cp)
96 return (cp >= LEAD_SURROGATE_MIN && cp <= TRAIL_SURROGATE_MAX);
99 template <typename u32>
100 inline bool is_code_point_valid(u32 cp)
102 return (cp <= CODE_POINT_MAX && !is_surrogate(cp) && cp != 0xfffe && cp != 0xffff);
105 template <typename octet_iterator>
106 inline typename std::iterator_traits<octet_iterator>::difference_type
107 sequence_length(octet_iterator lead_it)
109 uint8_t lead = mask8(*lead_it);
112 else if ((lead >> 5) == 0x6)
114 else if ((lead >> 4) == 0xe)
116 else if ((lead >> 3) == 0x1e)
122 template <typename octet_difference_type>
123 inline bool is_overlong_sequence(uint32_t cp, octet_difference_type length)
129 else if (cp < 0x800) {
133 else if (cp < 0x10000) {
141 enum utf_error {UTF8_OK, NOT_ENOUGH_ROOM, INVALID_LEAD, INCOMPLETE_SEQUENCE, OVERLONG_SEQUENCE, INVALID_CODE_POINT};
143 /// get_sequence_x functions decode utf-8 sequences of the length x
145 template <typename octet_iterator>
146 utf_error get_sequence_1(octet_iterator& it, octet_iterator end, uint32_t* code_point)
150 *code_point = mask8(*it);
153 return NOT_ENOUGH_ROOM;
156 template <typename octet_iterator>
157 utf_error get_sequence_2(octet_iterator& it, octet_iterator end, uint32_t* code_point)
159 utf_error ret_code = NOT_ENOUGH_ROOM;
162 uint32_t cp = mask8(*it);
165 cp = ((cp << 6) & 0x7ff) + ((*it) & 0x3f);
172 ret_code = INCOMPLETE_SEQUENCE;
175 ret_code = NOT_ENOUGH_ROOM;
181 template <typename octet_iterator>
182 utf_error get_sequence_3(octet_iterator& it, octet_iterator end, uint32_t* code_point)
184 utf_error ret_code = NOT_ENOUGH_ROOM;
187 uint32_t cp = mask8(*it);
190 cp = ((cp << 12) & 0xffff) + ((mask8(*it) << 6) & 0xfff);
200 ret_code = INCOMPLETE_SEQUENCE;
203 ret_code = NOT_ENOUGH_ROOM;
206 ret_code = INCOMPLETE_SEQUENCE;
209 ret_code = NOT_ENOUGH_ROOM;
215 template <typename octet_iterator>
216 utf_error get_sequence_4(octet_iterator& it, octet_iterator end, uint32_t* code_point)
218 utf_error ret_code = NOT_ENOUGH_ROOM;
221 uint32_t cp = mask8(*it);
224 cp = ((cp << 18) & 0x1fffff) + ((mask8(*it) << 12) & 0x3ffff);
227 cp += (mask8(*it) << 6) & 0xfff;
237 ret_code = INCOMPLETE_SEQUENCE;
240 ret_code = NOT_ENOUGH_ROOM;
243 ret_code = INCOMPLETE_SEQUENCE;
246 ret_code = NOT_ENOUGH_ROOM;
249 ret_code = INCOMPLETE_SEQUENCE;
252 ret_code = NOT_ENOUGH_ROOM;
258 template <typename octet_iterator>
259 utf_error validate_next(octet_iterator& it, octet_iterator end, uint32_t* code_point)
261 // Save the original value of it so we can go back in case of failure
262 // Of course, it does not make much sense with i.e. stream iterators
263 octet_iterator original_it = it;
266 // Determine the sequence length based on the lead octet
267 typedef typename std::iterator_traits<octet_iterator>::difference_type octet_difference_type;
268 octet_difference_type length = sequence_length(it);
272 // Now that we have a valid sequence length, get trail octets and calculate the code point
273 utf_error err = UTF8_OK;
276 err = get_sequence_1(it, end, &cp);
279 err = get_sequence_2(it, end, &cp);
282 err = get_sequence_3(it, end, &cp);
285 err = get_sequence_4(it, end, &cp);
289 if (err == UTF8_OK) {
290 // Decoding succeeded. Now, security checks...
291 if (is_code_point_valid(cp)) {
292 if (!is_overlong_sequence(cp, length)){
293 // Passed! Return here.
300 err = OVERLONG_SEQUENCE;
303 err = INVALID_CODE_POINT;
306 // Failure branch - restore the original value of the iterator
311 template <typename octet_iterator>
312 inline utf_error validate_next(octet_iterator& it, octet_iterator end) {
313 return validate_next(it, end, 0);
316 } // namespace internal
318 /// The library API - functions intended to be called by the users
321 const uint8_t bom[] = {0xef, 0xbb, 0xbf};
323 template <typename octet_iterator>
324 octet_iterator find_invalid(octet_iterator start, octet_iterator end)
326 octet_iterator result = start;
327 while (result != end) {
328 internal::utf_error err_code = internal::validate_next(result, end);
329 if (err_code != internal::UTF8_OK)
335 template <typename octet_iterator>
336 inline bool is_valid(octet_iterator start, octet_iterator end)
338 return (find_invalid(start, end) == end);
341 template <typename octet_iterator>
342 inline bool starts_with_bom (octet_iterator it, octet_iterator end)
345 ((it != end) && (internal::mask8(*it++)) == bom[0]) &&
346 ((it != end) && (internal::mask8(*it++)) == bom[1]) &&
347 ((it != end) && (internal::mask8(*it)) == bom[2])
351 //Deprecated in release 2.3
352 template <typename octet_iterator>
353 inline bool is_bom (octet_iterator it)
356 (internal::mask8(*it++)) == bom[0] &&
357 (internal::mask8(*it++)) == bom[1] &&
358 (internal::mask8(*it)) == bom[2]
363 #endif // header guard