Types

This page contains a canonical list of type symbols recognized by attach_function and Struct#layout

Types

• :char – 8-bit character
• :uchar – 8-bit unsigned character
• :int8 – 8-bit signed integer
• :uint8 – 8-bit unsigned integer
• :short – 16-bit signed integer
• :ushort – 16-bit unsigned integer
• :int16 – 16-bit signed integer
• :uint16 – 16-bit unsigned integer
• :int – 32-bit signed integer
• :uint – 32-bit unsigned integer
• :int32 – 32-bit signed integer
• :uint32 – 32-bit unsigned integer
• :long – long int – platform-specific size ¹
• :ulong – unsigned long int – platform-specific size
• :int64 – 64-bit signed integer
• :uint64 – 64-bit unsigned integer
• :long_long – 64-bit signed integer
• :ulong_long – 64-bit unsigned integer
• :float – 32-bit floating point
• :double – 64-bit floating point (double-precision)
• :pointer – pointer – platform-specific size
• :string – C-style (NULL-terminated) character string. NOTE: If the function modifies the string buffer use :pointer instead. That is, :string should be considered to be const char *. (In fact, MRI will let you modify a :string in place, but jruby will not, so this usage is not portable)
• :bool – (?? 1 byte in C++)
• :size_t – size_t from stddef.h – platform-dependent size
• :in_addr_t – 32-bit address in network byte order (FFI >= 1.0.8)
• :in_port_t – 16-bit port in network byte order (FFI >= 1.0.8)
• Enums – You can use an enum group (or its symbol) as a type

For function return type only:

• :void – For functions that return nothing (return type void).

For function argument type only:

• :buffer_in – Similar to :pointer, but optimized for Buffers that the function can only read (not write).
• :buffer_out – Similar to :pointer, but optimized for Buffers that the function can only write (not read).
• :buffer_inout – Similar to :pointer, but may be optimized for Buffers.
• :varargs – see examples for an example

Apparently there are most types available as well. For a full list see lib/ffi/platform/<your platform>/types.conf (FFI >= 1.0.8)

¹ this thread explains some of the different sizes on 64 versus 32-bit, but you shouldn’t have to worry about it.