mirror of
https://github.com/appgurueu/modlib.git
synced 2024-11-29 10:43:43 +01:00
127 lines
3.1 KiB
Lua
127 lines
3.1 KiB
Lua
-- Localize globals
|
|
local assert, math_huge, math_frexp, math_floor
|
|
= assert, math.huge, math.frexp, math.floor
|
|
|
|
local positive_nan, negative_nan = modlib.math.positive_nan, modlib.math.negative_nan
|
|
|
|
-- Set environment
|
|
local _ENV = {}
|
|
setfenv(1, _ENV)
|
|
|
|
-- All little endian
|
|
|
|
--+ Reads doubles (f64) or floats (f32)
|
|
--: double reads an f64 if true, f32 otherwise
|
|
function read_float(read_byte, double)
|
|
-- First read the mantissa
|
|
local mantissa = 0
|
|
for _ = 1, double and 6 or 2 do
|
|
mantissa = (mantissa + read_byte()) / 0x100
|
|
end
|
|
-- Second and first byte in big endian: last bit of exponent + 7 bits of mantissa, sign bit + 7 bits of exponent
|
|
local byte_2, byte_1 = read_byte(), read_byte()
|
|
local sign = 1
|
|
if byte_1 >= 0x80 then
|
|
sign = -1
|
|
byte_1 = byte_1 - 0x80
|
|
end
|
|
local exponent = byte_1 * 2
|
|
if byte_2 >= 0x80 then
|
|
exponent = exponent + 1
|
|
byte_2 = byte_2 - 0x80
|
|
end
|
|
mantissa = (mantissa + byte_2) / 0x80
|
|
if exponent == 0xFF then
|
|
if mantissa == 0 then
|
|
return sign * math_huge
|
|
end
|
|
-- Differentiating quiet and signalling nan is not possible in Lua, hence we don't have to do it
|
|
return sign == 1 and positive_nan or negative_nan
|
|
end
|
|
assert(mantissa < 1)
|
|
if exponent == 0 then
|
|
-- subnormal value
|
|
return sign * 2^-126 * mantissa
|
|
end
|
|
return sign * 2 ^ (exponent - 127) * (1 + mantissa)
|
|
end
|
|
|
|
--+ Reads a single floating point number (f32)
|
|
function read_single(read_byte)
|
|
return read_float(read_byte)
|
|
end
|
|
|
|
--+ Reads a double (f64)
|
|
function read_double(read_byte)
|
|
return read_float(read_byte, true)
|
|
end
|
|
|
|
function read_uint(read_byte, bytes)
|
|
local factor = 1
|
|
local uint = 0
|
|
for _ = 1, bytes do
|
|
uint = uint + read_byte() * factor
|
|
factor = factor * 0x100
|
|
end
|
|
return uint
|
|
end
|
|
|
|
function write_uint(write_byte, uint, bytes)
|
|
for _ = 1, bytes do
|
|
write_byte(uint % 0x100)
|
|
uint = math_floor(uint / 0x100)
|
|
end
|
|
assert(uint == 0)
|
|
end
|
|
|
|
--: on_write function(double)
|
|
--: double set to true to force f64, false for f32, nil for auto
|
|
function write_float(write_byte, number, on_write, double)
|
|
local sign = 0
|
|
if number < 0 then
|
|
number = -number
|
|
sign = 0x80
|
|
end
|
|
local mantissa, exponent = math_frexp(number)
|
|
exponent = exponent + 127
|
|
if exponent > 1 then
|
|
-- TODO ensure this deals properly with subnormal numbers
|
|
mantissa = mantissa * 2 - 1
|
|
exponent = exponent - 1
|
|
end
|
|
local sign_byte = sign + math_floor(exponent / 2)
|
|
mantissa = mantissa * 0x80
|
|
local exponent_byte = (exponent % 2) * 0x80 + math_floor(mantissa)
|
|
mantissa = mantissa % 1
|
|
local mantissa_bytes = {}
|
|
-- TODO ensure this check is proper
|
|
if double == nil then
|
|
double = mantissa % 2^-23 > 0
|
|
end
|
|
if on_write then
|
|
on_write(double)
|
|
end
|
|
local len = double and 6 or 2
|
|
for index = len, 1, -1 do
|
|
mantissa = mantissa * 0x100
|
|
mantissa_bytes[index] = math_floor(mantissa)
|
|
mantissa = mantissa % 1
|
|
end
|
|
assert(mantissa == 0)
|
|
for index = 1, len do
|
|
write_byte(mantissa_bytes[index])
|
|
end
|
|
write_byte(exponent_byte)
|
|
write_byte(sign_byte)
|
|
end
|
|
|
|
function write_single(write_byte, number)
|
|
return write_float(write_byte, number, nil, false)
|
|
end
|
|
|
|
function write_double(write_byte, number)
|
|
return write_float(write_byte, number, nil, true)
|
|
end
|
|
|
|
-- Export environment
|
|
return _ENV |