1 // Copyright 2006-2016 Nemanja Trifunovic 2 3 /* 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: 10 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. 17 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. 25 */ 26 27 28 #ifndef UTF8_FOR_CPP_CHECKED_H_2675DCD0_9480_4c0c_B92A_CC14C027B731 29 #define UTF8_FOR_CPP_CHECKED_H_2675DCD0_9480_4c0c_B92A_CC14C027B731 30 31 #include "core.h" 32 #include <stdexcept> 33 34 namespace utf8 35 { 36 // Base for the exceptions that may be thrown from the library 37 class exception : public ::std::exception { 38 }; 39 40 // Exceptions that may be thrown from the library functions. 41 class invalid_code_point : public exception { 42 uint32_t cp; 43 public: invalid_code_point(uint32_t codepoint)44 invalid_code_point(uint32_t codepoint) : cp(codepoint) {} what()45 virtual const char* what() const throw() { return "Invalid code point"; } code_point()46 uint32_t code_point() const {return cp;} 47 }; 48 49 class invalid_utf8 : public exception { 50 uint8_t u8; 51 public: invalid_utf8(uint8_t u)52 invalid_utf8 (uint8_t u) : u8(u) {} what()53 virtual const char* what() const throw() { return "Invalid UTF-8"; } utf8_octet()54 uint8_t utf8_octet() const {return u8;} 55 }; 56 57 class invalid_utf16 : public exception { 58 uint16_t u16; 59 public: invalid_utf16(uint16_t u)60 invalid_utf16 (uint16_t u) : u16(u) {} what()61 virtual const char* what() const throw() { return "Invalid UTF-16"; } utf16_word()62 uint16_t utf16_word() const {return u16;} 63 }; 64 65 class not_enough_room : public exception { 66 public: what()67 virtual const char* what() const throw() { return "Not enough space"; } 68 }; 69 70 /// The library API - functions intended to be called by the users 71 72 template <typename octet_iterator> append(uint32_t cp,octet_iterator result)73 octet_iterator append(uint32_t cp, octet_iterator result) 74 { 75 if (!utf8::internal::is_code_point_valid(cp)) 76 throw invalid_code_point(cp); 77 78 if (cp < 0x80) // one octet 79 *(result++) = static_cast<uint8_t>(cp); 80 else if (cp < 0x800) { // two octets 81 *(result++) = static_cast<uint8_t>((cp >> 6) | 0xc0); 82 *(result++) = static_cast<uint8_t>((cp & 0x3f) | 0x80); 83 } 84 else if (cp < 0x10000) { // three octets 85 *(result++) = static_cast<uint8_t>((cp >> 12) | 0xe0); 86 *(result++) = static_cast<uint8_t>(((cp >> 6) & 0x3f) | 0x80); 87 *(result++) = static_cast<uint8_t>((cp & 0x3f) | 0x80); 88 } 89 else { // four octets 90 *(result++) = static_cast<uint8_t>((cp >> 18) | 0xf0); 91 *(result++) = static_cast<uint8_t>(((cp >> 12) & 0x3f) | 0x80); 92 *(result++) = static_cast<uint8_t>(((cp >> 6) & 0x3f) | 0x80); 93 *(result++) = static_cast<uint8_t>((cp & 0x3f) | 0x80); 94 } 95 return result; 96 } 97 98 template <typename octet_iterator, typename output_iterator> replace_invalid(octet_iterator start,octet_iterator end,output_iterator out,uint32_t replacement)99 output_iterator replace_invalid(octet_iterator start, octet_iterator end, output_iterator out, uint32_t replacement) 100 { 101 while (start != end) { 102 octet_iterator sequence_start = start; 103 internal::utf_error err_code = utf8::internal::validate_next(start, end); 104 switch (err_code) { 105 case internal::UTF8_OK : 106 for (octet_iterator it = sequence_start; it != start; ++it) 107 *out++ = *it; 108 break; 109 case internal::NOT_ENOUGH_ROOM: 110 out = utf8::append (replacement, out); 111 start = end; 112 break; 113 case internal::INVALID_LEAD: 114 out = utf8::append (replacement, out); 115 ++start; 116 break; 117 case internal::INCOMPLETE_SEQUENCE: 118 case internal::OVERLONG_SEQUENCE: 119 case internal::INVALID_CODE_POINT: 120 out = utf8::append (replacement, out); 121 ++start; 122 // just one replacement mark for the sequence 123 while (start != end && utf8::internal::is_trail(*start)) 124 ++start; 125 break; 126 } 127 } 128 return out; 129 } 130 131 template <typename octet_iterator, typename output_iterator> replace_invalid(octet_iterator start,octet_iterator end,output_iterator out)132 inline output_iterator replace_invalid(octet_iterator start, octet_iterator end, output_iterator out) 133 { 134 static const uint32_t replacement_marker = utf8::internal::mask16(0xfffd); 135 return utf8::replace_invalid(start, end, out, replacement_marker); 136 } 137 138 template <typename octet_iterator> next(octet_iterator & it,octet_iterator end)139 uint32_t next(octet_iterator& it, octet_iterator end) 140 { 141 uint32_t cp = 0; 142 internal::utf_error err_code = utf8::internal::validate_next(it, end, cp); 143 switch (err_code) { 144 case internal::UTF8_OK : 145 break; 146 case internal::NOT_ENOUGH_ROOM : 147 throw not_enough_room(); 148 case internal::INVALID_LEAD : 149 case internal::INCOMPLETE_SEQUENCE : 150 case internal::OVERLONG_SEQUENCE : 151 throw invalid_utf8(*it); 152 case internal::INVALID_CODE_POINT : 153 throw invalid_code_point(cp); 154 } 155 return cp; 156 } 157 158 template <typename octet_iterator> peek_next(octet_iterator it,octet_iterator end)159 uint32_t peek_next(octet_iterator it, octet_iterator end) 160 { 161 return utf8::next(it, end); 162 } 163 164 template <typename octet_iterator> prior(octet_iterator & it,octet_iterator start)165 uint32_t prior(octet_iterator& it, octet_iterator start) 166 { 167 // can't do much if it == start 168 if (it == start) 169 throw not_enough_room(); 170 171 octet_iterator end = it; 172 // Go back until we hit either a lead octet or start 173 while (utf8::internal::is_trail(*(--it))) 174 if (it == start) 175 throw invalid_utf8(*it); // error - no lead byte in the sequence 176 return utf8::peek_next(it, end); 177 } 178 179 template <typename octet_iterator, typename distance_type> advance(octet_iterator & it,distance_type n,octet_iterator end)180 void advance (octet_iterator& it, distance_type n, octet_iterator end) 181 { 182 const distance_type zero(0); 183 if (n < zero) { 184 // backward 185 for (distance_type i = n; i < zero; ++i) 186 utf8::prior(it, end); 187 } else { 188 // forward 189 for (distance_type i = zero; i < n; ++i) 190 utf8::next(it, end); 191 } 192 } 193 194 template <typename octet_iterator> 195 typename std::iterator_traits<octet_iterator>::difference_type distance(octet_iterator first,octet_iterator last)196 distance (octet_iterator first, octet_iterator last) 197 { 198 typename std::iterator_traits<octet_iterator>::difference_type dist; 199 for (dist = 0; first < last; ++dist) 200 utf8::next(first, last); 201 return dist; 202 } 203 204 template <typename u16bit_iterator, typename octet_iterator> utf16to8(u16bit_iterator start,u16bit_iterator end,octet_iterator result)205 octet_iterator utf16to8 (u16bit_iterator start, u16bit_iterator end, octet_iterator result) 206 { 207 while (start != end) { 208 uint32_t cp = utf8::internal::mask16(*start++); 209 // Take care of surrogate pairs first 210 if (utf8::internal::is_lead_surrogate(cp)) { 211 if (start != end) { 212 uint32_t trail_surrogate = utf8::internal::mask16(*start++); 213 if (utf8::internal::is_trail_surrogate(trail_surrogate)) 214 cp = (cp << 10) + trail_surrogate + internal::SURROGATE_OFFSET; 215 else 216 throw invalid_utf16(static_cast<uint16_t>(trail_surrogate)); 217 } 218 else 219 throw invalid_utf16(static_cast<uint16_t>(cp)); 220 221 } 222 // Lone trail surrogate 223 else if (utf8::internal::is_trail_surrogate(cp)) 224 throw invalid_utf16(static_cast<uint16_t>(cp)); 225 226 result = utf8::append(cp, result); 227 } 228 return result; 229 } 230 231 template <typename u16bit_iterator, typename octet_iterator> utf8to16(octet_iterator start,octet_iterator end,u16bit_iterator result)232 u16bit_iterator utf8to16 (octet_iterator start, octet_iterator end, u16bit_iterator result) 233 { 234 while (start < end) { 235 uint32_t cp = utf8::next(start, end); 236 if (cp > 0xffff) { //make a surrogate pair 237 *result++ = static_cast<uint16_t>((cp >> 10) + internal::LEAD_OFFSET); 238 *result++ = static_cast<uint16_t>((cp & 0x3ff) + internal::TRAIL_SURROGATE_MIN); 239 } 240 else 241 *result++ = static_cast<uint16_t>(cp); 242 } 243 return result; 244 } 245 246 template <typename octet_iterator, typename u32bit_iterator> utf32to8(u32bit_iterator start,u32bit_iterator end,octet_iterator result)247 octet_iterator utf32to8 (u32bit_iterator start, u32bit_iterator end, octet_iterator result) 248 { 249 while (start != end) 250 result = utf8::append(*(start++), result); 251 252 return result; 253 } 254 255 template <typename octet_iterator, typename u32bit_iterator> utf8to32(octet_iterator start,octet_iterator end,u32bit_iterator result)256 u32bit_iterator utf8to32 (octet_iterator start, octet_iterator end, u32bit_iterator result) 257 { 258 while (start < end) 259 (*result++) = utf8::next(start, end); 260 261 return result; 262 } 263 264 // The iterator class 265 template <typename octet_iterator> 266 class iterator : public std::iterator <std::bidirectional_iterator_tag, uint32_t> { 267 octet_iterator it; 268 octet_iterator range_start; 269 octet_iterator range_end; 270 public: iterator()271 iterator () {} iterator(const octet_iterator & octet_it,const octet_iterator & rangestart,const octet_iterator & rangeend)272 explicit iterator (const octet_iterator& octet_it, 273 const octet_iterator& rangestart, 274 const octet_iterator& rangeend) : 275 it(octet_it), range_start(rangestart), range_end(rangeend) 276 { 277 if (it < range_start || it > range_end) 278 throw std::out_of_range("Invalid utf-8 iterator position"); 279 } 280 // the default "big three" are OK base()281 octet_iterator base () const { return it; } 282 uint32_t operator * () const 283 { 284 octet_iterator temp = it; 285 return utf8::next(temp, range_end); 286 } 287 bool operator == (const iterator& rhs) const 288 { 289 if (range_start != rhs.range_start || range_end != rhs.range_end) 290 throw std::logic_error("Comparing utf-8 iterators defined with different ranges"); 291 return (it == rhs.it); 292 } 293 bool operator != (const iterator& rhs) const 294 { 295 return !(operator == (rhs)); 296 } 297 iterator& operator ++ () 298 { 299 utf8::next(it, range_end); 300 return *this; 301 } 302 iterator operator ++ (int) 303 { 304 iterator temp = *this; 305 utf8::next(it, range_end); 306 return temp; 307 } 308 iterator& operator -- () 309 { 310 utf8::prior(it, range_start); 311 return *this; 312 } 313 iterator operator -- (int) 314 { 315 iterator temp = *this; 316 utf8::prior(it, range_start); 317 return temp; 318 } 319 }; // class iterator 320 321 } // namespace utf8 322 323 #endif //header guard 324 325