include/llvm/Support/ConvertUTF.h

   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 2001-2004 Unicode, Inc.
  10  *
  11  * Disclaimer
  12  *
  13  * This source code is provided as is by Unicode, Inc. No claims are
  14  * made as to fitness for any particular purpose. No warranties of any
  15  * kind are expressed or implied. The recipient agrees to determine
  16  * applicability of information provided. If this file has been
  17  * purchased on magnetic or optical media from Unicode, Inc., the
  18  * sole remedy for any claim will be exchange of defective media
  19  * within 90 days of receipt.
  20  *
  21  * Limitations on Rights to Redistribute This Code
  22  *
  23  * Unicode, Inc. hereby grants the right to freely use the information
  24  * supplied in this file in the creation of products supporting the
  25  * Unicode Standard, and to make copies of this file in any form
  26  * for internal or external distribution as long as this notice
  27  * remains attached.
  28  */
  29
  30 /* ---------------------------------------------------------------------
  31
  32     Conversions between UTF32, UTF-16, and UTF-8.  Header file.
  33
  34     Several funtions are included here, forming a complete set of
  35     conversions between the three formats.  UTF-7 is not included
  36     here, but is handled in a separate source file.
  37
  38     Each of these routines takes pointers to input buffers and output
  39     buffers.  The input buffers are const.
  40
  41     Each routine converts the text between *sourceStart and sourceEnd,
  42     putting the result into the buffer between *targetStart and
  43     targetEnd. Note: the end pointers are *after* the last item: e.g.
  44     *(sourceEnd - 1) is the last item.
  45
  46     The return result indicates whether the conversion was successful,
  47     and if not, whether the problem was in the source or target buffers.
  48     (Only the first encountered problem is indicated.)
  49
  50     After the conversion, *sourceStart and *targetStart are both
  51     updated to point to the end of last text successfully converted in
  52     the respective buffers.
  53
  54     Input parameters:
  55         sourceStart - pointer to a pointer to the source buffer.
  56                 The contents of this are modified on return so that
  57                 it points at the next thing to be converted.
  58         targetStart - similarly, pointer to pointer to the target buffer.
  59         sourceEnd, targetEnd - respectively pointers to the ends of the
  60                 two buffers, for overflow checking only.
  61
  62     These conversion functions take a ConversionFlags argument. When this
  63     flag is set to strict, both irregular sequences and isolated surrogates
  64     will cause an error.  When the flag is set to lenient, both irregular
  65     sequences and isolated surrogates are converted.
  66
  67     Whether the flag is strict or lenient, all illegal sequences will cause
  68     an error return. This includes sequences such as: <F4 90 80 80>, <C0 80>,
  69     or <A0> in UTF-8, and values above 0x10FFFF in UTF-32. Conformant code
  70     must check for illegal sequences.
  71
  72     When the flag is set to lenient, characters over 0x10FFFF are converted
  73     to the replacement character; otherwise (when the flag is set to strict)
  74     they constitute an error.
  75
  76     Output parameters:
  77         The value "sourceIllegal" is returned from some routines if the input
  78         sequence is malformed.  When "sourceIllegal" is returned, the source
  79         value will point to the illegal value that caused the problem. E.g.,
  80         in UTF-8 when a sequence is malformed, it points to the start of the
  81         malformed sequence.
  82
  83     Author: Mark E. Davis, 1994.
  84     Rev History: Rick McGowan, fixes & updates May 2001.
  85          Fixes & updates, Sept 2001.
  86
  87 ------------------------------------------------------------------------ */
  88
  89 #ifndef LLVM_SUPPORT_CONVERTUTF_H
  90 #define LLVM_SUPPORT_CONVERTUTF_H
  91
  92 #include <cstddef>
  93 #include <string>
  94 #include <system_error>
  95
  96 // Wrap everything in namespace llvm so that programs can link with llvm and
  97 // their own version of the unicode libraries.
  98
  99 namespace llvm {
 100
 101 /* ---------------------------------------------------------------------
 102     The following 4 definitions are compiler-specific.
 103     The C standard does not guarantee that wchar_t has at least
 104     16 bits, so wchar_t is no less portable than unsigned short!
 105     All should be unsigned values to avoid sign extension during
 106     bit mask & shift operations.
 107 ------------------------------------------------------------------------ */
 108
 109 typedef unsigned int    UTF32;  /* at least 32 bits */
 110 typedef unsigned short  UTF16;  /* at least 16 bits */
 111 typedef unsigned char   UTF8;   /* typically 8 bits */
 112 typedef unsigned char   Boolean; /* 0 or 1 */
 113
 114 /* Some fundamental constants */
 115 #define UNI_REPLACEMENT_CHAR (UTF32)0x0000FFFD
 116 #define UNI_MAX_BMP (UTF32)0x0000FFFF
 117 #define UNI_MAX_UTF16 (UTF32)0x0010FFFF
 118 #define UNI_MAX_UTF32 (UTF32)0x7FFFFFFF
 119 #define UNI_MAX_LEGAL_UTF32 (UTF32)0x0010FFFF
 120
 121 #define UNI_MAX_UTF8_BYTES_PER_CODE_POINT 4
 122
 123 #define UNI_UTF16_BYTE_ORDER_MARK_NATIVE  0xFEFF
 124 #define UNI_UTF16_BYTE_ORDER_MARK_SWAPPED 0xFFFE
 125
 126 typedef enum {
 127   conversionOK,           /* conversion successful */
 128   sourceExhausted,        /* partial character in source, but hit end */
 129   targetExhausted,        /* insuff. room in target for conversion */
 130   sourceIllegal           /* source sequence is illegal/malformed */
 131 } ConversionResult;
 132
 133 typedef enum {
 134   strictConversion = 0,
 135   lenientConversion
 136 } ConversionFlags;
 137
 138 ConversionResult ConvertUTF8toUTF16 (
 139   const UTF8** sourceStart, const UTF8* sourceEnd,
 140   UTF16** targetStart, UTF16* targetEnd, ConversionFlags flags);
 141
 142 /**
 143  * Convert a partial UTF8 sequence to UTF32.  If the sequence ends in an
 144  * incomplete code unit sequence, returns \c sourceExhausted.
 145  */
 146 ConversionResult ConvertUTF8toUTF32Partial(
 147   const UTF8** sourceStart, const UTF8* sourceEnd,
 148   UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags);
 149
 150 /**
 151  * Convert a partial UTF8 sequence to UTF32.  If the sequence ends in an
 152  * incomplete code unit sequence, returns \c sourceIllegal.
 153  */
 154 ConversionResult ConvertUTF8toUTF32(
 155   const UTF8** sourceStart, const UTF8* sourceEnd,
 156   UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags);
 157
 158 ConversionResult ConvertUTF16toUTF8 (
 159   const UTF16** sourceStart, const UTF16* sourceEnd,
 160   UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags);
 161
 162 ConversionResult ConvertUTF32toUTF8 (
 163   const UTF32** sourceStart, const UTF32* sourceEnd,
 164   UTF8** targetStart, UTF8* targetEnd, ConversionFlags flags);
 165
 166 ConversionResult ConvertUTF16toUTF32 (
 167   const UTF16** sourceStart, const UTF16* sourceEnd,
 168   UTF32** targetStart, UTF32* targetEnd, ConversionFlags flags);
 169
 170 ConversionResult ConvertUTF32toUTF16 (
 171   const UTF32** sourceStart, const UTF32* sourceEnd,
 172   UTF16** targetStart, UTF16* targetEnd, ConversionFlags flags);
 173
 174 Boolean isLegalUTF8Sequence(const UTF8 *source, const UTF8 *sourceEnd);
 175
 176 Boolean isLegalUTF8String(const UTF8 **source, const UTF8 *sourceEnd);
 177
 178 unsigned getNumBytesForUTF8(UTF8 firstByte);
 179
 180 /*************************************************************************/
 181 /* Below are LLVM-specific wrappers of the functions above. */
 182
 183 template <typename T> class ArrayRef;
 184 template <typename T> class SmallVectorImpl;
 185 class StringRef;
 186
 187 /**
 188  * Convert an UTF8 StringRef to UTF8, UTF16, or UTF32 depending on
 189  * WideCharWidth. The converted data is written to ResultPtr, which needs to
 190  * point to at least WideCharWidth * (Source.Size() + 1) bytes. On success,
 191  * ResultPtr will point one after the end of the copied string. On failure,
 192  * ResultPtr will not be changed, and ErrorPtr will be set to the location of
 193  * the first character which could not be converted.
 194  * \return true on success.
 195  */
 196 bool ConvertUTF8toWide(unsigned WideCharWidth, llvm::StringRef Source,
 197                        char *&ResultPtr, const UTF8 *&ErrorPtr);
 198
 199 /**
 200 * Converts a UTF-8 StringRef to a std::wstring.
 201 * \return true on success.
 202 */
 203 bool ConvertUTF8toWide(llvm::StringRef Source, std::wstring &Result);
 204
 205 /**
 206 * Converts a UTF-8 C-string to a std::wstring.
 207 * \return true on success.
 208 */
 209 bool ConvertUTF8toWide(const char *Source, std::wstring &Result);
 210
 211 /**
 212 * Converts a std::wstring to a UTF-8 encoded std::string.
 213 * \return true on success.
 214 */
 215 bool convertWideToUTF8(const std::wstring &Source, std::string &Result);
 216
 217
 218 /**
 219  * Convert an Unicode code point to UTF8 sequence.
 220  *
 221  * \param Source a Unicode code point.
 222  * \param [in,out] ResultPtr pointer to the output buffer, needs to be at least
 223  * \c UNI_MAX_UTF8_BYTES_PER_CODE_POINT bytes.  On success \c ResultPtr is
 224  * updated one past end of the converted sequence.
 225  *
 226  * \returns true on success.
 227  */
 228 bool ConvertCodePointToUTF8(unsigned Source, char *&ResultPtr);
 229
 230 /**
 231  * Convert the first UTF8 sequence in the given source buffer to a UTF32
 232  * code point.
 233  *
 234  * \param [in,out] source A pointer to the source buffer. If the conversion
 235  * succeeds, this pointer will be updated to point to the byte just past the
 236  * end of the converted sequence.
 237  * \param sourceEnd A pointer just past the end of the source buffer.
 238  * \param [out] target The converted code
 239  * \param flags Whether the conversion is strict or lenient.
 240  *
 241  * \returns conversionOK on success
 242  *
 243  * \sa ConvertUTF8toUTF32
 244  */
 245 inline ConversionResult convertUTF8Sequence(const UTF8 **source,
 246                                             const UTF8 *sourceEnd,
 247                                             UTF32 *target,
 248                                             ConversionFlags flags) {
 249   if (*source == sourceEnd)
 250     return sourceExhausted;
 251   unsigned size = getNumBytesForUTF8(**source);
 252   if ((ptrdiff_t)size > sourceEnd - *source)
 253     return sourceExhausted;
 254   return ConvertUTF8toUTF32(source, *source + size, &target, target + 1, flags);
 255 }
 256
 257 /**
 258  * Returns true if a blob of text starts with a UTF-16 big or little endian byte
 259  * order mark.
 260  */
 261 bool hasUTF16ByteOrderMark(ArrayRef<char> SrcBytes);
 262
 263 /**
 264  * Converts a stream of raw bytes assumed to be UTF16 into a UTF8 std::string.
 265  *
 266  * \param [in] SrcBytes A buffer of what is assumed to be UTF-16 encoded text.
 267  * \param [out] Out Converted UTF-8 is stored here on success.
 268  * \returns true on success
 269  */
 270 bool convertUTF16ToUTF8String(ArrayRef<char> SrcBytes, std::string &Out);
 271
 272 /**
 273 * Converts a UTF16 string into a UTF8 std::string.
 274 *
 275 * \param [in] Src A buffer of UTF-16 encoded text.
 276 * \param [out] Out Converted UTF-8 is stored here on success.
 277 * \returns true on success
 278 */
 279 bool convertUTF16ToUTF8String(ArrayRef<UTF16> Src, std::string &Out);
 280
 281 /**
 282  * Converts a UTF-8 string into a UTF-16 string with native endianness.
 283  *
 284  * \returns true on success
 285  */
 286 bool convertUTF8ToUTF16String(StringRef SrcUTF8,
 287                               SmallVectorImpl<UTF16> &DstUTF16);
 288
 289 #if defined(_WIN32)
 290 namespace sys {
 291 namespace windows {
 292 std::error_code UTF8ToUTF16(StringRef utf8, SmallVectorImpl<wchar_t> &utf16);
 293 /// Convert to UTF16 from the current code page used in the system
 294 std::error_code CurCPToUTF16(StringRef utf8, SmallVectorImpl<wchar_t> &utf16);
 295 std::error_code UTF16ToUTF8(const wchar_t *utf16, size_t utf16_len,
 296                             SmallVectorImpl<char> &utf8);
 297 /// Convert from UTF16 to the current code page used in the system
 298 std::error_code UTF16ToCurCP(const wchar_t *utf16, size_t utf16_len,
 299                              SmallVectorImpl<char> &utf8);
 300 } // namespace windows
 301 } // namespace sys
 302 #endif
 303
 304 } /* end namespace llvm */
 305
 306 #endif