Standard library header <cinttypes>
From cppreference.com
This header was originally in the C standard library as <inttypes.h>.
Provides conversions from C byte and wide strings to std::intmax_t and std::uintmax_t, overloads some math functions for std::intmax_t and provide C style input/output format macros for the types declared in <cstdint>.
Contents |
Includes
<cstdint> |
Types
(C++11) |
structure type, returned by std::imaxdiv (typedef) |
Functions
(C++11)(C++11) |
computes absolute value of an integral value (|x|) (function) |
(C++11)(C++11) |
computes quotient and remainder of integer division (function) |
(C++11)(C++11) |
converts a byte string to std::intmax_t or std::uintmax_t (function) |
(C++11)(C++11) |
converts a wide string to std::intmax_t or std::uintmax_t (function) |
Macros
Format constants for the std::fprintf family of functions | |
PRId8PRId16PRId32PRId64PRIdLEAST8PRIdLEAST16PRIdLEAST32PRIdLEAST64PRIdFAST8PRIdFAST16PRIdFAST32PRIdFAST64PRIdMAXPRIdPTR |
format conversion specifier to output a signed decimal integer value of type std::int8_t, std::int16_t, std::int32_t, std::int64_t, std::int_least8_t, std::int_least16_t, std::int_least32_t, std::int_least64_t, std::int_fast8_t, std::int_fast16_t, std::int_fast32_t, std::int_fast64_t, std::intmax_t, std::intptr_t respectively, equivalent to d for int (macro constant) |
PRIi8PRIi16PRIi32PRIi64PRIiLEAST8PRIiLEAST16PRIiLEAST32PRIiLEAST64PRIiFAST8PRIiFAST16PRIiFAST32PRIiFAST64PRIiMAXPRIiPTR |
format conversion specifier to output a signed decimal integer value of type std::int8_t, std::int16_t, std::int32_t, std::int64_t, std::int_least8_t, std::int_least16_t, std::int_least32_t, std::int_least64_t, std::int_fast8_t, std::int_fast16_t, std::int_fast32_t, std::int_fast64_t, std::intmax_t, std::intptr_t respectively, equivalent to i for int (macro constant) |
PRIu8PRIu16PRIu32PRIu64PRIuLEAST8PRIuLEAST16PRIuLEAST32PRIuLEAST64PRIuFAST8PRIuFAST16PRIuFAST32PRIuFAST64PRIuMAXPRIuPTR |
format conversion specifier to output an unsigned decimal integer value of type std::uint8_t, std::uint16_t, std::uint32_t, std::uint64_t, std::uint_least8_t, std::uint_least16_t, std::uint_least32_t, std::uint_least64_t, std::uint_fast8_t, std::uint_fast16_t, std::uint_fast32_t, std::uint_fast64_t, std::uintmax_t, std::uintptr_t respectively, equivalent to u for unsigned int (macro constant) |
PRIo8PRIo16PRIo32PRIo64PRIoLEAST8PRIoLEAST16PRIoLEAST32PRIoLEAST64PRIoFAST8PRIoFAST16PRIoFAST32PRIoFAST64PRIoMAXPRIoPTR |
format conversion specifier to output an unsigned octal integer value of type std::uint8_t, std::uint16_t, std::uint32_t, std::uint64_t, std::uint_least8_t, std::uint_least16_t, std::uint_least32_t, std::uint_least64_t, std::uint_fast8_t, std::uint_fast16_t, std::uint_fast32_t, std::uint_fast64_t, std::uintmax_t, std::uintptr_t respectively, equivalent to o for unsigned int (macro constant) |
PRIx8PRIx16PRIx32PRIx64PRIxLEAST8PRIxLEAST16PRIxLEAST32PRIxLEAST64PRIxFAST8PRIxFAST16PRIxFAST32PRIxFAST64PRIxMAXPRIxPTR |
format conversion specifier to output an unsigned lowercase hexadecimal integer value of type std::uint8_t, std::uint16_t, std::uint32_t, std::uint64_t, std::uint_least8_t, std::uint_least16_t, std::uint_least32_t, std::uint_least64_t, std::uint_fast8_t, std::uint_fast16_t, std::uint_fast32_t, std::uint_fast64_t, std::uintmax_t, std::uintptr_t respectively, equivalent to x for unsigned int (macro constant) |
PRIX8PRIX16PRIX32PRIX64PRIXLEAST8PRIXLEAST16PRIXLEAST32PRIXLEAST64PRIXFAST8PRIXFAST16PRIXFAST32PRIXFAST64PRIXMAXPRIXPTR |
format conversion specifier to output an unsigned uppercase hexadecimal integer value of type std::uint8_t, std::uint16_t, std::uint32_t, std::uint64_t, std::uint_least8_t, std::uint_least16_t, std::uint_least32_t, std::uint_least64_t, std::uint_fast8_t, std::uint_fast16_t, std::uint_fast32_t, std::uint_fast64_t, std::uintmax_t, std::uintptr_t respectively, equivalent to X for unsigned int (macro constant) |
Format constants for the std::fscanf family of functions | |
SCNd8SCNd16SCNd32SCNd64SCNdLEAST8SCNdLEAST16SCNdLEAST32SCNdLEAST64SCNdFAST8SCNdFAST16SCNdFAST32SCNdFAST64SCNdMAXSCNdPTR |
format conversion specifier to input a signed decimal integer value of type std::int8_t, std::int16_t, std::int32_t, std::int64_t, std::int_least8_t, std::int_least16_t, std::int_least32_t, std::int_least64_t, std::int_fast8_t, std::int_fast16_t, std::int_fast32_t, std::int_fast64_t, std::intmax_t, std::intptr_t respectively, equivalent to d for int (macro constant) |
SCNi8SCNi16SCNi32SCNi64SCNiLEAST8SCNiLEAST16SCNiLEAST32SCNiLEAST64SCNiFAST8SCNiFAST16SCNiFAST32SCNiFAST64SCNiMAXSCNiPTR |
format conversion specifier to input a signed decimal/octal/hexadecimal integer value of type std::int8_t, std::int16_t, std::int32_t, std::int64_t, std::int_least8_t, std::int_least16_t, std::int_least32_t, std::int_least64_t, std::int_fast8_t, std::int_fast16_t, std::int_fast32_t, std::int_fast64_t, std::intmax_t, std::intptr_t respectively, equivalent to i for int (macro constant) |
SCNu8SCNu16SCNu32SCNu64SCNuLEAST8SCNuLEAST16SCNuLEAST32SCNuLEAST64SCNuFAST8SCNuFAST16SCNuFAST32SCNuFAST64SCNuMAXSCNuPTR |
format conversion specifier to input an unsigned decimal integer value of type std::uint8_t, std::uint16_t, std::uint32_t, std::uint64_t, std::uint_least8_t, std::uint_least16_t, std::uint_least32_t, std::uint_least64_t, std::uint_fast8_t, std::uint_fast16_t, std::uint_fast32_t, std::uint_fast64_t, std::uintmax_t, std::uintptr_t respectively, equivalent to u for unsigned int (macro constant) |
SCNo8SCNo16SCNo32SCNo64SCNoLEAST8SCNoLEAST16SCNoLEAST32SCNoLEAST64SCNoFAST8SCNoFAST16SCNoFAST32SCNoFAST64SCNoMAXSCNoPTR |
format conversion specifier to input an unsigned octal integer value of type std::uint8_t, std::uint16_t, std::uint32_t, std::uint64_t, std::uint_least8_t, std::uint_least16_t, std::uint_least32_t, std::uint_least64_t, std::uint_fast8_t, std::uint_fast16_t, std::uint_fast32_t, std::uint_fast64_t, std::uintmax_t, std::uintptr_t respectively, equivalent to o for unsigned int (macro constant) |
SCNx8SCNx16SCNx32SCNx64SCNxLEAST8SCNxLEAST16SCNxLEAST32SCNxLEAST64SCNxFAST8SCNxFAST16SCNxFAST32SCNxFAST64SCNxMAXSCNxPTR |
format conversion specifier to input an unsigned hexadecimal integer value of type std::uint8_t, std::uint16_t, std::uint32_t, std::uint64_t, std::uint_least8_t, std::uint_least16_t, std::uint_least32_t, std::uint_least64_t, std::uint_fast8_t, std::uint_fast16_t, std::uint_fast32_t, std::uint_fast64_t, std::uintmax_t, std::uintptr_t respectively, equivalent to x for unsigned int (macro constant) |