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
33 // Determine the C++ standard version.
34 // If the user defines UTF_CPP_CPLUSPLUS, use that.
35 // Otherwise, trust the unreliable predefined macro __cplusplus
37 #if !defined UTF_CPP_CPLUSPLUS
38 #define UTF_CPP_CPLUSPLUS __cplusplus
41 #if UTF_CPP_CPLUSPLUS >= 201103L // C++ 11 or later
42 #define UTF_CPP_OVERRIDE override
43 #define UTF_CPP_NOEXCEPT noexcept
45 #define UTF_CPP_OVERRIDE
46 #define UTF_CPP_NOEXCEPT throw()
47 #endif // C++ 11 or later
52 // The typedefs for 8-bit, 16-bit and 32-bit unsigned integers
53 // You may need to change them to match your system.
54 // These typedefs have the same names as ones from cstdint, or boost/cstdint
55 typedef unsigned char uint8_t;
56 typedef unsigned short uint16_t;
57 typedef unsigned int uint32_t;
59 // Helper code - not intended to be directly called by the library users. May be changed at any time
63 // Leading (high) surrogates: 0xd800 - 0xdbff
64 // Trailing (low) surrogates: 0xdc00 - 0xdfff
65 const uint16_t LEAD_SURROGATE_MIN = 0xd800u;
66 const uint16_t LEAD_SURROGATE_MAX = 0xdbffu;
67 const uint16_t TRAIL_SURROGATE_MIN = 0xdc00u;
68 const uint16_t TRAIL_SURROGATE_MAX = 0xdfffu;
69 const uint16_t LEAD_OFFSET = 0xd7c0u; // LEAD_SURROGATE_MIN - (0x10000 >> 10)
70 const uint32_t SURROGATE_OFFSET = 0xfca02400u; // 0x10000u - (LEAD_SURROGATE_MIN << 10) - TRAIL_SURROGATE_MIN
72 // Maximum valid value for a Unicode code point
73 const uint32_t CODE_POINT_MAX = 0x0010ffffu;
75 template<typename octet_type>
76 inline uint8_t mask8(octet_type oc)
78 return static_cast<uint8_t>(0xff & oc);
80 template<typename u16_type>
81 inline uint16_t mask16(u16_type oc)
83 return static_cast<uint16_t>(0xffff & oc);
85 template<typename octet_type>
86 inline bool is_trail(octet_type oc)
88 return ((utf8::internal::mask8(oc) >> 6) == 0x2);
91 template <typename u16>
92 inline bool is_lead_surrogate(u16 cp)
94 return (cp >= LEAD_SURROGATE_MIN && cp <= LEAD_SURROGATE_MAX);
97 template <typename u16>
98 inline bool is_trail_surrogate(u16 cp)
100 return (cp >= TRAIL_SURROGATE_MIN && cp <= TRAIL_SURROGATE_MAX);
103 template <typename u16>
104 inline bool is_surrogate(u16 cp)
106 return (cp >= LEAD_SURROGATE_MIN && cp <= TRAIL_SURROGATE_MAX);
109 template <typename u32>
110 inline bool is_code_point_valid(u32 cp)
112 return (cp <= CODE_POINT_MAX && !utf8::internal::is_surrogate(cp));
115 template <typename octet_iterator>
116 inline typename std::iterator_traits<octet_iterator>::difference_type
117 sequence_length(octet_iterator lead_it)
119 uint8_t lead = utf8::internal::mask8(*lead_it);
122 else if ((lead >> 5) == 0x6)
124 else if ((lead >> 4) == 0xe)
126 else if ((lead >> 3) == 0x1e)
132 template <typename octet_difference_type>
133 inline bool is_overlong_sequence(uint32_t cp, octet_difference_type length)
139 else if (cp < 0x800) {
143 else if (cp < 0x10000) {
151 enum utf_error {UTF8_OK, NOT_ENOUGH_ROOM, INVALID_LEAD, INCOMPLETE_SEQUENCE, OVERLONG_SEQUENCE, INVALID_CODE_POINT};
153 /// Helper for get_sequence_x
154 template <typename octet_iterator>
155 utf_error increase_safely(octet_iterator& it, octet_iterator end)
158 return NOT_ENOUGH_ROOM;
160 if (!utf8::internal::is_trail(*it))
161 return INCOMPLETE_SEQUENCE;
166 #define UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(IT, END) {utf_error ret = increase_safely(IT, END); if (ret != UTF8_OK) return ret;}
168 /// get_sequence_x functions decode utf-8 sequences of the length x
169 template <typename octet_iterator>
170 utf_error get_sequence_1(octet_iterator& it, octet_iterator end, uint32_t& code_point)
173 return NOT_ENOUGH_ROOM;
175 code_point = utf8::internal::mask8(*it);
180 template <typename octet_iterator>
181 utf_error get_sequence_2(octet_iterator& it, octet_iterator end, uint32_t& code_point)
184 return NOT_ENOUGH_ROOM;
186 code_point = utf8::internal::mask8(*it);
188 UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end)
190 code_point = ((code_point << 6) & 0x7ff) + ((*it) & 0x3f);
195 template <typename octet_iterator>
196 utf_error get_sequence_3(octet_iterator& it, octet_iterator end, uint32_t& code_point)
199 return NOT_ENOUGH_ROOM;
201 code_point = utf8::internal::mask8(*it);
203 UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end)
205 code_point = ((code_point << 12) & 0xffff) + ((utf8::internal::mask8(*it) << 6) & 0xfff);
207 UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end)
209 code_point += (*it) & 0x3f;
214 template <typename octet_iterator>
215 utf_error get_sequence_4(octet_iterator& it, octet_iterator end, uint32_t& code_point)
218 return NOT_ENOUGH_ROOM;
220 code_point = utf8::internal::mask8(*it);
222 UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end)
224 code_point = ((code_point << 18) & 0x1fffff) + ((utf8::internal::mask8(*it) << 12) & 0x3ffff);
226 UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end)
228 code_point += (utf8::internal::mask8(*it) << 6) & 0xfff;
230 UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR(it, end)
232 code_point += (*it) & 0x3f;
237 #undef UTF8_CPP_INCREASE_AND_RETURN_ON_ERROR
239 template <typename octet_iterator>
240 utf_error validate_next(octet_iterator& it, octet_iterator end, uint32_t& code_point)
243 return NOT_ENOUGH_ROOM;
245 // Save the original value of it so we can go back in case of failure
246 // Of course, it does not make much sense with i.e. stream iterators
247 octet_iterator original_it = it;
250 // Determine the sequence length based on the lead octet
251 typedef typename std::iterator_traits<octet_iterator>::difference_type octet_difference_type;
252 const octet_difference_type length = utf8::internal::sequence_length(it);
254 // Get trail octets and calculate the code point
255 utf_error err = UTF8_OK;
260 err = utf8::internal::get_sequence_1(it, end, cp);
263 err = utf8::internal::get_sequence_2(it, end, cp);
266 err = utf8::internal::get_sequence_3(it, end, cp);
269 err = utf8::internal::get_sequence_4(it, end, cp);
273 if (err == UTF8_OK) {
274 // Decoding succeeded. Now, security checks...
275 if (utf8::internal::is_code_point_valid(cp)) {
276 if (!utf8::internal::is_overlong_sequence(cp, length)){
277 // Passed! Return here.
283 err = OVERLONG_SEQUENCE;
286 err = INVALID_CODE_POINT;
289 // Failure branch - restore the original value of the iterator
294 template <typename octet_iterator>
295 inline utf_error validate_next(octet_iterator& it, octet_iterator end) {
297 return utf8::internal::validate_next(it, end, ignored);
300 } // namespace internal
302 /// The library API - functions intended to be called by the users
305 const uint8_t bom[] = {0xef, 0xbb, 0xbf};
307 template <typename octet_iterator>
308 octet_iterator find_invalid(octet_iterator start, octet_iterator end)
310 octet_iterator result = start;
311 while (result != end) {
312 utf8::internal::utf_error err_code = utf8::internal::validate_next(result, end);
313 if (err_code != internal::UTF8_OK)
319 template <typename octet_iterator>
320 inline bool is_valid(octet_iterator start, octet_iterator end)
322 return (utf8::find_invalid(start, end) == end);
325 template <typename octet_iterator>
326 inline bool starts_with_bom (octet_iterator it, octet_iterator end)
329 ((it != end) && (utf8::internal::mask8(*it++)) == bom[0]) &&
330 ((it != end) && (utf8::internal::mask8(*it++)) == bom[1]) &&
331 ((it != end) && (utf8::internal::mask8(*it)) == bom[2])
336 #endif // header guard