My code currently calculates the single-precision float manually based on the IEEE 754 specs. However, in languages like C, this entire operation is effectively a nop. Is there a better (faster) way to do this conversion?
func bin2float(pkt:PoolByteArray) -> float:
#Convert packet to an array of bools
var arr:Array=[]
for byte in pkt:
for _i in range(0,8):
arr.append(byte%2)
byte/=2
#Extract sign
var r_sign:int
if arr[31]==1:
r_sign=-1
else:
r_sign=1
#Extract exponent
var r_exp:int=0
for i in range(30,22,-1):
r_exp*=2
r_exp+=arr[i]
r_exp-=127
#Extract mantissa
var r_mant:float=0
for i in range(0,23):
r_mant+=arr[i]
r_mant/=2
r_mant+=1
return r_sign*pow(2,r_exp)*r_mant
I tried using some of the Data Packet classes, but they seem to require an entire protocol setup to use.
The StreamPeerBuffer can be used to do the conversion, by writing the bytes into it, an then retrieving a float.
func bin2float(pkt:PoolByteArray) -> float:
var buff = StreamPeerBuffer.new()
buff.data_array = pkt
return buff.get_float()
Source: https://godotengine.org/qa/81109