The stdint library provides signed and unsigned integer types of various fixed widths: 8, 16, 24, 32, 40, 48, 56, 64 and 128 bit.
This interface is similar to Int32 and Int64 from the base library but provides more functions and constants like arithmetic and bit-wise operations, constants like maximum and minimum values, infix operators conversion to and from every other integer type (including int, float and nativeint), parsing from and conversion to readable strings (binary, octal, decimal, hexademical), conversion to and from buffers in both big endian and little endian byte order.
Published: 19 Oct 2022
This OCaml library provides integer types having specified widths.
It provides the following unsigned integer types:
and the following list of signed integer types:
int32 (identical to int32 from the base library)
int64 (identical to int64 from the base library)
There is a module for every integer type that implements the Int interface. This interface is similar to Int32 and Int64 from the base library but provides more functions and constants like
arithmetic and bit-wise operations
constants like maximum and minimum value
conversion to and from every other integer type (including int, float and nativeint)
parsing from and conversion to readable strings (binary, octal, decimal, hexademical)
conversion to and from buffers in both big endian and little endian byte order
The library also comes with C header files that allow easy access to the integer values in C bindings. The functions are modeled on base of the int32 and int64 access functions provided by the base library. The semantics of all operations is identical to the Int32, Int64 modules, C, C++, Java etc.: Conversion will silently truncate larger values as will other operations leading to overflowing integer values.
To use the integer types, we recommend to
open Stdint but not the individual modules:
open Stdint let _ = let a = Uint8.of_int 21 in let x = Uint8.(a * (one + one)) in print_endline (Uint8.to_string x)
The 128 bit integer types are currently only available on 64 bit platforms; the compatibility layer for 32 bit platforms is not yet fully implemented and will raise
Failure for several functions.
The representation of integers depends on their size:
Signed integers smaller than the standard integer type are stored in a standard
int. They are left-aligned so that most arithmetic operations are just the same as the ones on normal integers. The standard OCaml integer type is 31 bit on 32 bit machines and 63 bit on 64 bit machines. Operations like addition and division require an extra shift; others like xor require an additional mask to keep the unused bits (at the right) at
Unsigned integers smaller than the standard integer types are stored in the standard
int, too. They are right-aligned making most arithmetic operations compatible to standard integer operations. Operations like addition require an additional mask operation to keep the unused bits (at the left) at
uint64have their custom in-memory representation implemented in C
Signed integers larger than the standard integer but smaller than
int64are stored in the latter. The requirements are otherwise identical to small signed integers store in the standard integer.
Unsigned integers larger than the standard integer but smaller then
uint64are stored in the latter. The requirements are otherwise identical to small unsigned integers store in the standard integer.
128 Bit integers have a custom in-memory representation implemented in C. On 64 Bit platforms they use the specialized arithmetic operations provided by the C compiler that are much faster than manual operations, to which a fallback solution exists on 32 Bit platforms.
The source-code of stdint is available under the MIT license.
Used by (25)
>= "4.0-6" & < "5.2.1"