1 | // Copyright 2006 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: |
44 | invalid_code_point(uint32_t cp) : cp(cp) {} |
45 | virtual const char* what() const throw() { return "Invalid code point" ; } |
46 | uint32_t code_point() const {return cp;} |
47 | }; |
48 | |
49 | class invalid_utf8 : public exception { |
50 | uint8_t u8; |
51 | public: |
52 | invalid_utf8 (uint8_t u) : u8(u) {} |
53 | virtual const char* what() const throw() { return "Invalid UTF-8" ; } |
54 | uint8_t utf8_octet() const {return u8;} |
55 | }; |
56 | |
57 | class invalid_utf16 : public exception { |
58 | uint16_t u16; |
59 | public: |
60 | invalid_utf16 (uint16_t u) : u16(u) {} |
61 | virtual const char* what() const throw() { return "Invalid UTF-16" ; } |
62 | uint16_t utf16_word() const {return u16;} |
63 | }; |
64 | |
65 | class not_enough_room : public exception { |
66 | public: |
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> |
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> |
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 | throw not_enough_room(); |
111 | case internal::INVALID_LEAD: |
112 | out = utf8::append (replacement, out); |
113 | ++start; |
114 | break; |
115 | case internal::INCOMPLETE_SEQUENCE: |
116 | case internal::OVERLONG_SEQUENCE: |
117 | case internal::INVALID_CODE_POINT: |
118 | out = utf8::append (replacement, out); |
119 | ++start; |
120 | // just one replacement mark for the sequence |
121 | while (start != end && utf8::internal::is_trail(*start)) |
122 | ++start; |
123 | break; |
124 | } |
125 | } |
126 | return out; |
127 | } |
128 | |
129 | template <typename octet_iterator, typename output_iterator> |
130 | inline output_iterator replace_invalid(octet_iterator start, octet_iterator end, output_iterator out) |
131 | { |
132 | static const uint32_t replacement_marker = utf8::internal::mask16(0xfffd); |
133 | return utf8::replace_invalid(start, end, out, replacement_marker); |
134 | } |
135 | |
136 | template <typename octet_iterator> |
137 | uint32_t next(octet_iterator& it, octet_iterator end) |
138 | { |
139 | uint32_t cp = 0; |
140 | internal::utf_error err_code = utf8::internal::validate_next(it, end, cp); |
141 | switch (err_code) { |
142 | case internal::UTF8_OK : |
143 | break; |
144 | case internal::NOT_ENOUGH_ROOM : |
145 | throw not_enough_room(); |
146 | case internal::INVALID_LEAD : |
147 | case internal::INCOMPLETE_SEQUENCE : |
148 | case internal::OVERLONG_SEQUENCE : |
149 | throw invalid_utf8(*it); |
150 | case internal::INVALID_CODE_POINT : |
151 | throw invalid_code_point(cp); |
152 | } |
153 | return cp; |
154 | } |
155 | |
156 | template <typename octet_iterator> |
157 | uint32_t peek_next(octet_iterator it, octet_iterator end) |
158 | { |
159 | return utf8::next(it, end); |
160 | } |
161 | |
162 | template <typename octet_iterator> |
163 | uint32_t prior(octet_iterator& it, octet_iterator start) |
164 | { |
165 | // can't do much if it == start |
166 | if (it == start) |
167 | throw not_enough_room(); |
168 | |
169 | octet_iterator end = it; |
170 | // Go back until we hit either a lead octet or start |
171 | while (utf8::internal::is_trail(*(--it))) |
172 | if (it == start) |
173 | throw invalid_utf8(*it); // error - no lead byte in the sequence |
174 | return utf8::peek_next(it, end); |
175 | } |
176 | |
177 | /// Deprecated in versions that include "prior" |
178 | template <typename octet_iterator> |
179 | uint32_t previous(octet_iterator& it, octet_iterator pass_start) |
180 | { |
181 | octet_iterator end = it; |
182 | while (utf8::internal::is_trail(*(--it))) |
183 | if (it == pass_start) |
184 | throw invalid_utf8(*it); // error - no lead byte in the sequence |
185 | octet_iterator temp = it; |
186 | return utf8::next(temp, end); |
187 | } |
188 | |
189 | template <typename octet_iterator, typename distance_type> |
190 | void advance (octet_iterator& it, distance_type n, octet_iterator end) |
191 | { |
192 | for (distance_type i = 0; i < n; ++i) |
193 | utf8::next(it, end); |
194 | } |
195 | |
196 | template <typename octet_iterator> |
197 | typename std::iterator_traits<octet_iterator>::difference_type |
198 | distance (octet_iterator first, octet_iterator last) |
199 | { |
200 | typename std::iterator_traits<octet_iterator>::difference_type dist; |
201 | for (dist = 0; first < last; ++dist) |
202 | utf8::next(first, last); |
203 | return dist; |
204 | } |
205 | |
206 | template <typename u16bit_iterator, typename octet_iterator> |
207 | octet_iterator utf16to8 (u16bit_iterator start, u16bit_iterator end, octet_iterator result) |
208 | { |
209 | while (start != end) { |
210 | uint32_t cp = utf8::internal::mask16(*start++); |
211 | // Take care of surrogate pairs first |
212 | if (utf8::internal::is_lead_surrogate(cp)) { |
213 | if (start != end) { |
214 | uint32_t trail_surrogate = utf8::internal::mask16(*start++); |
215 | if (utf8::internal::is_trail_surrogate(trail_surrogate)) |
216 | cp = (cp << 10) + trail_surrogate + internal::SURROGATE_OFFSET; |
217 | else |
218 | throw invalid_utf16(static_cast<uint16_t>(trail_surrogate)); |
219 | } |
220 | else |
221 | throw invalid_utf16(static_cast<uint16_t>(cp)); |
222 | |
223 | } |
224 | // Lone trail surrogate |
225 | else if (utf8::internal::is_trail_surrogate(cp)) |
226 | throw invalid_utf16(static_cast<uint16_t>(cp)); |
227 | |
228 | result = utf8::append(cp, result); |
229 | } |
230 | return result; |
231 | } |
232 | |
233 | template <typename u16bit_iterator, typename octet_iterator> |
234 | u16bit_iterator utf8to16 (octet_iterator start, octet_iterator end, u16bit_iterator result) |
235 | { |
236 | while (start != end) { |
237 | uint32_t cp = utf8::next(start, end); |
238 | if (cp > 0xffff) { //make a surrogate pair |
239 | *result++ = static_cast<uint16_t>((cp >> 10) + internal::LEAD_OFFSET); |
240 | *result++ = static_cast<uint16_t>((cp & 0x3ff) + internal::TRAIL_SURROGATE_MIN); |
241 | } |
242 | else |
243 | *result++ = static_cast<uint16_t>(cp); |
244 | } |
245 | return result; |
246 | } |
247 | |
248 | template <typename octet_iterator, typename u32bit_iterator> |
249 | octet_iterator utf32to8 (u32bit_iterator start, u32bit_iterator end, octet_iterator result) |
250 | { |
251 | while (start != end) |
252 | result = utf8::append(*(start++), result); |
253 | |
254 | return result; |
255 | } |
256 | |
257 | template <typename octet_iterator, typename u32bit_iterator> |
258 | u32bit_iterator utf8to32 (octet_iterator start, octet_iterator end, u32bit_iterator result) |
259 | { |
260 | while (start != end) |
261 | (*result++) = utf8::next(start, end); |
262 | |
263 | return result; |
264 | } |
265 | |
266 | // The iterator class |
267 | template <typename octet_iterator> |
268 | class iterator : public std::iterator <std::bidirectional_iterator_tag, uint32_t> { |
269 | octet_iterator it; |
270 | octet_iterator range_start; |
271 | octet_iterator range_end; |
272 | public: |
273 | iterator () {} |
274 | explicit iterator (const octet_iterator& octet_it, |
275 | const octet_iterator& range_start, |
276 | const octet_iterator& range_end) : |
277 | it(octet_it), range_start(range_start), range_end(range_end) |
278 | { |
279 | if (it < range_start || it > range_end) |
280 | throw std::out_of_range("Invalid utf-8 iterator position" ); |
281 | } |
282 | // the default "big three" are OK |
283 | octet_iterator base () const { return it; } |
284 | uint32_t operator * () const |
285 | { |
286 | octet_iterator temp = it; |
287 | return utf8::next(temp, range_end); |
288 | } |
289 | bool operator == (const iterator& rhs) const |
290 | { |
291 | if (range_start != rhs.range_start || range_end != rhs.range_end) |
292 | throw std::logic_error("Comparing utf-8 iterators defined with different ranges" ); |
293 | return (it == rhs.it); |
294 | } |
295 | bool operator != (const iterator& rhs) const |
296 | { |
297 | return !(operator == (rhs)); |
298 | } |
299 | iterator& operator ++ () |
300 | { |
301 | utf8::next(it, range_end); |
302 | return *this; |
303 | } |
304 | iterator operator ++ (int) |
305 | { |
306 | iterator temp = *this; |
307 | utf8::next(it, range_end); |
308 | return temp; |
309 | } |
310 | iterator& operator -- () |
311 | { |
312 | utf8::prior(it, range_start); |
313 | return *this; |
314 | } |
315 | iterator operator -- (int) |
316 | { |
317 | iterator temp = *this; |
318 | utf8::prior(it, range_start); |
319 | return temp; |
320 | } |
321 | }; // class iterator |
322 | |
323 | } // namespace utf8 |
324 | |
325 | #endif //header guard |
326 | |
327 | |
328 | |