ConvertUTF.h File Reference

#include <span>
#include <cstddef>
#include <string>
#include <string_view>
#include <system_error>

Go to the source code of this file.

Namespaces
namespace	wpi
	/file This file defines the SmallVector class.

Macros
#define	UNI_REPLACEMENT_CHAR   (UTF32)0x0000FFFD
#define	UNI_MAX_BMP   (UTF32)0x0000FFFF
#define	UNI_MAX_UTF16   (UTF32)0x0010FFFF
#define	UNI_MAX_UTF32   (UTF32)0x7FFFFFFF
#define	UNI_MAX_LEGAL_UTF32   (UTF32)0x0010FFFF
#define	UNI_MAX_UTF8_BYTES_PER_CODE_POINT   4
#define	UNI_UTF16_BYTE_ORDER_MARK_NATIVE   0xFEFF
#define	UNI_UTF16_BYTE_ORDER_MARK_SWAPPED   0xFFFE

Typedefs
typedef unsigned int	wpi::UTF32
typedef unsigned short	wpi::UTF16
typedef unsigned char	wpi::UTF8
typedef bool	wpi::Boolean

Enumerations
enum	wpi::ConversionResult { wpi::conversionOK , wpi::sourceExhausted , wpi::targetExhausted , wpi::sourceIllegal }
enum	wpi::ConversionFlags { wpi::strictConversion = 0 , wpi::lenientConversion }

Functions
ConversionResult	wpi::ConvertUTF8toUTF16 (const UTF8 **sourceStart, const UTF8 *sourceEnd, UTF16 **targetStart, UTF16 *targetEnd, ConversionFlags flags)
ConversionResult	wpi::ConvertUTF8toUTF32Partial (const UTF8 **sourceStart, const UTF8 *sourceEnd, UTF32 **targetStart, UTF32 *targetEnd, ConversionFlags flags)
	Convert a partial UTF8 sequence to UTF32. More...
ConversionResult	wpi::ConvertUTF8toUTF32 (const UTF8 **sourceStart, const UTF8 *sourceEnd, UTF32 **targetStart, UTF32 *targetEnd, ConversionFlags flags)
	Convert a partial UTF8 sequence to UTF32. More...
ConversionResult	wpi::ConvertUTF16toUTF8 (const UTF16 **sourceStart, const UTF16 *sourceEnd, UTF8 **targetStart, UTF8 *targetEnd, ConversionFlags flags)
ConversionResult	wpi::ConvertUTF32toUTF8 (const UTF32 **sourceStart, const UTF32 *sourceEnd, UTF8 **targetStart, UTF8 *targetEnd, ConversionFlags flags)
ConversionResult	wpi::ConvertUTF16toUTF32 (const UTF16 **sourceStart, const UTF16 *sourceEnd, UTF32 **targetStart, UTF32 *targetEnd, ConversionFlags flags)
ConversionResult	wpi::ConvertUTF32toUTF16 (const UTF32 **sourceStart, const UTF32 *sourceEnd, UTF16 **targetStart, UTF16 *targetEnd, ConversionFlags flags)
Boolean	wpi::isLegalUTF8Sequence (const UTF8 *source, const UTF8 *sourceEnd)
Boolean	wpi::isLegalUTF8String (const UTF8 **source, const UTF8 *sourceEnd)
unsigned	wpi::getNumBytesForUTF8 (UTF8 firstByte)
bool	wpi::ConvertUTF8toWide (unsigned WideCharWidth, std::string_view Source, char *&ResultPtr, const UTF8 *&ErrorPtr)
	Convert an UTF8 string_view to UTF8, UTF16, or UTF32 depending on WideCharWidth. More...
bool	wpi::ConvertUTF8toWide (std::string_view Source, std::wstring &Result)
	Converts a UTF-8 string_view to a std::wstring. More...
bool	wpi::ConvertUTF8toWide (const char *Source, std::wstring &Result)
	Converts a UTF-8 C-string to a std::wstring. More...
bool	wpi::convertWideToUTF8 (const std::wstring &Source, SmallVectorImpl< char > &Result)
	Converts a std::wstring to a UTF-8 encoded std::string. More...
bool	wpi::ConvertCodePointToUTF8 (unsigned Source, char *&ResultPtr)
	Convert an Unicode code point to UTF8 sequence. More...
ConversionResult	wpi::convertUTF8Sequence (const UTF8 **source, const UTF8 *sourceEnd, UTF32 *target, ConversionFlags flags)
	Convert the first UTF8 sequence in the given source buffer to a UTF32 code point. More...
bool	wpi::hasUTF16ByteOrderMark (std::span< const char > SrcBytes)
	Returns true if a blob of text starts with a UTF-16 big or little endian byte order mark. More...
bool	wpi::convertUTF16ToUTF8String (std::span< const char > SrcBytes, SmallVectorImpl< char > &Out)
	Converts a stream of raw bytes assumed to be UTF16 into a UTF8 std::string. More...
bool	wpi::convertUTF16ToUTF8String (std::span< const UTF16 > Src, SmallVectorImpl< char > &Out)
	Converts a UTF16 string into a UTF8 std::string. More...
bool	wpi::convertUTF8ToUTF16String (std::string_view SrcUTF8, SmallVectorImpl< UTF16 > &DstUTF16)
	Converts a UTF-8 string into a UTF-16 string with native endianness. More...

Macro Definition Documentation

UNI_MAX_BMP

#define UNI_MAX_BMP   (UTF32)0x0000FFFF

UNI_MAX_LEGAL_UTF32

#define UNI_MAX_LEGAL_UTF32   (UTF32)0x0010FFFF

UNI_MAX_UTF16

#define UNI_MAX_UTF16   (UTF32)0x0010FFFF

UNI_MAX_UTF32

#define UNI_MAX_UTF32   (UTF32)0x7FFFFFFF

UNI_MAX_UTF8_BYTES_PER_CODE_POINT

#define UNI_MAX_UTF8_BYTES_PER_CODE_POINT   4

UNI_REPLACEMENT_CHAR

#define UNI_REPLACEMENT_CHAR   (UTF32)0x0000FFFD

UNI_UTF16_BYTE_ORDER_MARK_NATIVE

#define UNI_UTF16_BYTE_ORDER_MARK_NATIVE   0xFEFF

UNI_UTF16_BYTE_ORDER_MARK_SWAPPED

#define UNI_UTF16_BYTE_ORDER_MARK_SWAPPED   0xFFFE