1 /*===--- ConvertUTF.h - Universal Character Names conversions ---------------===
2 *
3 * Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
4 * See https://llvm.org/LICENSE.txt for license information.
5 * SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
6 *
7 *==------------------------------------------------------------------------==*/
8 /*
9 * Copyright © 1991-2015 Unicode, Inc. All rights reserved.
10 * Distributed under the Terms of Use in
11 * http://www.unicode.org/copyright.html.
12 *
13 * Permission is hereby granted, free of charge, to any person obtaining
14 * a copy of the Unicode data files and any associated documentation
15 * (the "Data Files") or Unicode software and any associated documentation
16 * (the "Software") to deal in the Data Files or Software
17 * without restriction, including without limitation the rights to use,
18 * copy, modify, merge, publish, distribute, and/or sell copies of
19 * the Data Files or Software, and to permit persons to whom the Data Files
20 * or Software are furnished to do so, provided that
21 * (a) this copyright and permission notice appear with all copies
22 * of the Data Files or Software,
23 * (b) this copyright and permission notice appear in associated
24 * documentation, and
25 * (c) there is clear notice in each modified Data File or in the Software
26 * as well as in the documentation associated with the Data File(s) or
27 * Software that the data or software has been modified.
28 *
29 * THE DATA FILES AND SOFTWARE ARE PROVIDED "AS IS", WITHOUT WARRANTY OF
30 * ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE
31 * WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
32 * NONINFRINGEMENT OF THIRD PARTY RIGHTS.
33 * IN NO EVENT SHALL THE COPYRIGHT HOLDER OR HOLDERS INCLUDED IN THIS
34 * NOTICE BE LIABLE FOR ANY CLAIM, OR ANY SPECIAL INDIRECT OR CONSEQUENTIAL
35 * DAMAGES, OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE,
36 * DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
37 * TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
38 * PERFORMANCE OF THE DATA FILES OR SOFTWARE.
39 *
40 * Except as contained in this notice, the name of a copyright holder
41 * shall not be used in advertising or otherwise to promote the sale,
42 * use or other dealings in these Data Files or Software without prior
43 * written authorization of the copyright holder.
44 */
45
46 /* ---------------------------------------------------------------------
47
48 Conversions between UTF32, UTF-16, and UTF-8. Header file.
49
50 Several funtions are included here, forming a complete set of
51 conversions between the three formats. UTF-7 is not included
52 here, but is handled in a separate source file.
53
54 Each of these routines takes pointers to input buffers and output
55 buffers. The input buffers are const.
56
57 Each routine converts the text between *sourceStart and sourceEnd,
58 putting the result into the buffer between *targetStart and
59 targetEnd. Note: the end pointers are *after* the last item: e.g.
60 *(sourceEnd - 1) is the last item.
61
62 The return result indicates whether the conversion was successful,
63 and if not, whether the problem was in the source or target buffers.
64 (Only the first encountered problem is indicated.)
65
66 After the conversion, *sourceStart and *targetStart are both
67 updated to point to the end of last text successfully converted in
68 the respective buffers.
69
70 Input parameters:
71 sourceStart - pointer to a pointer to the source buffer.
72 The contents of this are modified on return so that
73 it points at the next thing to be converted.
74 targetStart - similarly, pointer to pointer to the target buffer.
75 sourceEnd, targetEnd - respectively pointers to the ends of the
76 two buffers, for overflow checking only.
77
78 These conversion functions take a ConversionFlags argument. When this
79 flag is set to strict, both irregular sequences and isolated surrogates
80 will cause an error. When the flag is set to lenient, both irregular
81 sequences and isolated surrogates are converted.
82
83 Whether the flag is strict or lenient, all illegal sequences will cause
84 an error return. This includes sequences such as: <F4 90 80 80>, <C0 80>,
85 or <A0> in UTF-8, and values above 0x10FFFF in UTF-32. Conformant code
86 must check for illegal sequences.
87
88 When the flag is set to lenient, characters over 0x10FFFF are converted
89 to the replacement character; otherwise (when the flag is set to strict)
90 they constitute an error.
91
92 Output parameters:
93 The value "sourceIllegal" is returned from some routines if the input
94 sequence is malformed. When "sourceIllegal" is returned, the source
95 value will point to the illegal value that caused the problem. E.g.,
96 in UTF-8 when a sequence is malformed, it points to the start of the
97 malformed sequence.
98
99 Author: Mark E. Davis, 1994.
100 Rev History: Rick McGowan, fixes & updates May 2001.
101 Fixes & updates, Sept 2001.
102
103 ------------------------------------------------------------------------ */
104
105 #ifndef LLVM_SUPPORT_CONVERTUTF_H
106 #define LLVM_SUPPORT_CONVERTUTF_H
107
108 #include <cstddef>
109 #include <string>
110
111 #if defined(_WIN32)
112 #include <system_error>
113 #endif
114
115 // Wrap everything in namespace llvm so that programs can link with llvm and
116 // their own version of the unicode libraries.
117
118 namespace llvm {
119
120 /* ---------------------------------------------------------------------
121 The following 4 definitions are compiler-specific.
122 The C standard does not guarantee that wchar_t has at least
123 16 bits, so wchar_t is no less portable than unsigned short!
124 All should be unsigned values to avoid sign extension during
125 bit mask & shift operations.
126 ------------------------------------------------------------------------ */
127
128 typedef unsigned int UTF32; /* at least 32 bits */
129 typedef unsigned short UTF16; /* at least 16 bits */
130 typedef unsigned char UTF8; /* typically 8 bits */
131 typedef unsigned char Boolean; /* 0 or 1 */
132
133 /* Some fundamental constants */
134 #define UNI_REPLACEMENT_CHAR (UTF32)0x0000FFFD
135 #define UNI_MAX_BMP (UTF32)0x0000FFFF
136 #define UNI_MAX_UTF16 (UTF32)0x0010FFFF
137 #define UNI_MAX_UTF32 (UTF32)0x7FFFFFFF
138 #define UNI_MAX_LEGAL_UTF32 (UTF32)0x0010FFFF
139
140 #define UNI_MAX_UTF8_BYTES_PER_CODE_POINT 4
141
142 #define UNI_UTF16_BYTE_ORDER_MARK_NATIVE 0xFEFF
143 #define UNI_UTF16_BYTE_ORDER_MARK_SWAPPED 0xFFFE
144
145 #define UNI_UTF32_BYTE_ORDER_MARK_NATIVE 0x0000FEFF
146 #define UNI_UTF32_BYTE_ORDER_MARK_SWAPPED 0xFFFE0000
147
148 typedef enum {
149 conversionOK, /* conversion successful */
150 sourceExhausted, /* partial character in source, but hit end */
151 targetExhausted, /* insuff. room in target for conversion */
152 sourceIllegal /* source sequence is illegal/malformed */
153 } ConversionResult;
154
155 typedef enum {
156 strictConversion = 0,
157 lenientConversion
158 } ConversionFlags;
159
160 ConversionResult ConvertUTF8toUTF16 (
161 const UTF8** sourceStart, const UTF8* sourceEnd,
162 UTF16** targetStart, UTF16* targetEnd, ConversionFlags flags);
163
164 /**
165 * Convert a partial UTF8 sequence to UTF32. If the sequence ends in an
166 * incomplete code unit sequence, returns \c sourceExhausted.
167 */
168 ConversionResult ConvertUTF8toUTF32Partial(
169 const UTF8** sourceStart, const UTF8* sourceEnd,
170 UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags);
171
172 /**
173 * Convert a partial UTF8 sequence to UTF32. If the sequence ends in an
174 * incomplete code unit sequence, returns \c sourceIllegal.
175 */
176 ConversionResult ConvertUTF8toUTF32(
177 const UTF8** sourceStart, const UTF8* sourceEnd,
178 UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags);
179
180 ConversionResult ConvertUTF16toUTF8 (
181 const UTF16** sourceStart, const UTF16* sourceEnd,
182 UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags);
183
184 ConversionResult ConvertUTF32toUTF8 (
185 const UTF32** sourceStart, const UTF32* sourceEnd,
186 UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags);
187
188 ConversionResult ConvertUTF16toUTF32 (
189 const UTF16** sourceStart, const UTF16* sourceEnd,
190 UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags);
191
192 ConversionResult ConvertUTF32toUTF16 (
193 const UTF32** sourceStart, const UTF32* sourceEnd,
194 UTF16** targetStart, UTF16* targetEnd, ConversionFlags flags);
195
196 Boolean isLegalUTF8Sequence(const UTF8 *source, const UTF8 *sourceEnd);
197
198 Boolean isLegalUTF8String(const UTF8 **source, const UTF8 *sourceEnd);
199
200 unsigned getUTF8SequenceSize(const UTF8 *source, const UTF8 *sourceEnd);
201
202 unsigned getNumBytesForUTF8(UTF8 firstByte);
203
204 /*************************************************************************/
205 /* Below are LLVM-specific wrappers of the functions above. */
206
207 template <typename T> class ArrayRef;
208 template <typename T> class SmallVectorImpl;
209 class StringRef;
210
211 /**
212 * Convert an UTF8 StringRef to UTF8, UTF16, or UTF32 depending on
213 * WideCharWidth. The converted data is written to ResultPtr, which needs to
214 * point to at least WideCharWidth * (Source.Size() + 1) bytes. On success,
215 * ResultPtr will point one after the end of the copied string. On failure,
216 * ResultPtr will not be changed, and ErrorPtr will be set to the location of
217 * the first character which could not be converted.
218 * \return true on success.
219 */
220 bool ConvertUTF8toWide(unsigned WideCharWidth, llvm::StringRef Source,
221 char *&ResultPtr, const UTF8 *&ErrorPtr);
222
223 /**
224 * Converts a UTF-8 StringRef to a std::wstring.
225 * \return true on success.
226 */
227 bool ConvertUTF8toWide(llvm::StringRef Source, std::wstring &Result);
228
229 /**
230 * Converts a UTF-8 C-string to a std::wstring.
231 * \return true on success.
232 */
233 bool ConvertUTF8toWide(const char *Source, std::wstring &Result);
234
235 /**
236 * Converts a std::wstring to a UTF-8 encoded std::string.
237 * \return true on success.
238 */
239 bool convertWideToUTF8(const std::wstring &Source, std::string &Result);
240
241
242 /**
243 * Convert an Unicode code point to UTF8 sequence.
244 *
245 * \param Source a Unicode code point.
246 * \param [in,out] ResultPtr pointer to the output buffer, needs to be at least
247 * \c UNI_MAX_UTF8_BYTES_PER_CODE_POINT bytes. On success \c ResultPtr is
248 * updated one past end of the converted sequence.
249 *
250 * \returns true on success.
251 */
252 bool ConvertCodePointToUTF8(unsigned Source, char *&ResultPtr);
253
254 /**
255 * Convert the first UTF8 sequence in the given source buffer to a UTF32
256 * code point.
257 *
258 * \param [in,out] source A pointer to the source buffer. If the conversion
259 * succeeds, this pointer will be updated to point to the byte just past the
260 * end of the converted sequence.
261 * \param sourceEnd A pointer just past the end of the source buffer.
262 * \param [out] target The converted code
263 * \param flags Whether the conversion is strict or lenient.
264 *
265 * \returns conversionOK on success
266 *
267 * \sa ConvertUTF8toUTF32
268 */
convertUTF8Sequence(const UTF8 ** source,const UTF8 * sourceEnd,UTF32 * target,ConversionFlags flags)269 inline ConversionResult convertUTF8Sequence(const UTF8 **source,
270 const UTF8 *sourceEnd,
271 UTF32 *target,
272 ConversionFlags flags) {
273 if (*source == sourceEnd)
274 return sourceExhausted;
275 unsigned size = getNumBytesForUTF8(**source);
276 if ((ptrdiff_t)size > sourceEnd - *source)
277 return sourceExhausted;
278 return ConvertUTF8toUTF32(source, *source + size, &target, target + 1, flags);
279 }
280
281 /**
282 * Returns true if a blob of text starts with a UTF-16 big or little endian byte
283 * order mark.
284 */
285 bool hasUTF16ByteOrderMark(ArrayRef<char> SrcBytes);
286
287 /**
288 * Converts a stream of raw bytes assumed to be UTF16 into a UTF8 std::string.
289 *
290 * \param [in] SrcBytes A buffer of what is assumed to be UTF-16 encoded text.
291 * \param [out] Out Converted UTF-8 is stored here on success.
292 * \returns true on success
293 */
294 bool convertUTF16ToUTF8String(ArrayRef<char> SrcBytes, std::string &Out);
295
296 /**
297 * Converts a UTF16 string into a UTF8 std::string.
298 *
299 * \param [in] Src A buffer of UTF-16 encoded text.
300 * \param [out] Out Converted UTF-8 is stored here on success.
301 * \returns true on success
302 */
303 bool convertUTF16ToUTF8String(ArrayRef<UTF16> Src, std::string &Out);
304
305 /**
306 * Converts a stream of raw bytes assumed to be UTF32 into a UTF8 std::string.
307 *
308 * \param [in] SrcBytes A buffer of what is assumed to be UTF-32 encoded text.
309 * \param [out] Out Converted UTF-8 is stored here on success.
310 * \returns true on success
311 */
312 bool convertUTF32ToUTF8String(ArrayRef<char> SrcBytes, std::string &Out);
313
314 /**
315 * Converts a UTF32 string into a UTF8 std::string.
316 *
317 * \param [in] Src A buffer of UTF-32 encoded text.
318 * \param [out] Out Converted UTF-8 is stored here on success.
319 * \returns true on success
320 */
321 bool convertUTF32ToUTF8String(ArrayRef<UTF32> Src, std::string &Out);
322
323 /**
324 * Converts a UTF-8 string into a UTF-16 string with native endianness.
325 *
326 * \returns true on success
327 */
328 bool convertUTF8ToUTF16String(StringRef SrcUTF8,
329 SmallVectorImpl<UTF16> &DstUTF16);
330
331 #if defined(_WIN32)
332 namespace sys {
333 namespace windows {
334 std::error_code UTF8ToUTF16(StringRef utf8, SmallVectorImpl<wchar_t> &utf16);
335 /// Convert to UTF16 from the current code page used in the system
336 std::error_code CurCPToUTF16(StringRef utf8, SmallVectorImpl<wchar_t> &utf16);
337 std::error_code UTF16ToUTF8(const wchar_t *utf16, size_t utf16_len,
338 SmallVectorImpl<char> &utf8);
339 /// Convert from UTF16 to the current code page used in the system
340 std::error_code UTF16ToCurCP(const wchar_t *utf16, size_t utf16_len,
341 SmallVectorImpl<char> &utf8);
342 } // namespace windows
343 } // namespace sys
344 #endif
345
346 } /* end namespace llvm */
347
348 #endif
349