Write string of escaped byte characters, ie '\x00\x20\…', as if they were bytes

Question

I am playing around with .wav file generation and have functions to convert base 10 integer into hexadecimal. I need to write these hex numbers as if they were bytes, not simply encoding them to be written as bytes. I can hard code using escape sequences but I need to use strings provided by my functions.

My temporary work around makes use of eval as follows.

eval("b'" + endian(bytestring(tohex(16, 32))) + "'")

What is a more preferred method to accomplish this task without eval?

Here is some code to better understand the situation.

with open('song.wav', 'wb') as f:
    f.write(ChunkID)
    f.write(ChunkSize)
    f.write(Format)
    f.write(Subchunk1ID)
    f.write(Subchunk1Size)
    f.write(AudioFormat)
    f.write(NumChannels)
    f.write(SampleRate)
    f.write(ByteRate)
    f.write(BlockAlign)
    f.write(BitsPerSample)
    f.write(Subchunk2ID)
    f.write(Subchunk2Size)
    f.write(Data)

Using hard coded values works fine.

ChunkID = b'RIFF'               #big        1
ChunkSize = b'\x1a\xea\xc4\x00'     #little     2
Format = b'WAVE'                #big        3
Subchunk1ID = b'fmt '               #big        4
Subchunk1Size = b'\x10\x00\x00\x00'         #little     5
AudioFormat = b'\x01\x00'           #little     6
NumChannels = b'\x02\x00'           #little     7
SampleRate = b'\x22\x56\x00\x00'        #little     8
ByteRate = b'\x88\x58\x01\x00'          #little     9
BlockAlign = b'\x04\x00'            #little     10
BitsPerSample = b'\x10\x00'         #little     11
Subchunk2ID = b'data'               #big        12
Subchunk2Size = b'\x10\xb1\x02\x00'     #little     13
Data = b''                  #little     14

I need more flexibility and need similar results using functions.

def tohex(val, nbits):
    a = hex((val + (1 << nbits)) % (1 << nbits))
    pad = nbits/4 - (len(a) - 2)
    for i in range(0, pad):
        a = a[:2] + '0' + a[2:]
    return a
def bytestring(hexstring):
    hexstring = hexstring[2:]
    bstring = ''
    for i in range(0, len(hexstring)/2):
        bstring += '\\x'
        bstring += hexstring[2*i: 2*i+2]
    return bstring
def endian(bstring):
    return '\\'+'\\'.join(reversed(bstring.split('\\'))).rstrip('\\')

This is how I call the functions

sample_rate = 44100
channels = 2
sample_bits = 16
sample_count = 176400

ChunkID = b'RIFF'
Format = b'WAVE'
Subchunk1ID = b'fmt '
Subchunk1Size = eval("b'" + endian(bytestring(tohex(16, 32))) + "'")    #n bit samples : tohex(n,m)
AudioFormat = eval("b'" + endian(bytestring(tohex(1, 16))) + "'")       #uncompressed when n=1
NumChannels = eval("b'" + endian(bytestring(tohex(channels, 16))) + "'")        #n channels
SampleRate = eval("b'" + endian(bytestring(tohex(sample_rate, 32))) + "'")
BitsPerSample = eval("b'" + endian(bytestring(tohex(sample_bits, 16))) + "'")
ByteRate = eval("b'" +  endian(bytestring(tohex(sample_rate*channels*sample_bits/8, 32))) + "'")
BlockAlign = eval("b'" + endian(bytestring(tohex(channels*sample_bits/8, 16))) + "'")
Subchunk2ID = b'data'
Subchunk2Size = eval("b'" + endian(bytestring(tohex(sample_count*channels*sample_bits/8, 32))) + "'")
ChunkSize = eval("b'" + endian(bytestring(tohex(36 + sample_count*channels*sample_bits/8, 32))) + "'")
Data = b''

for i in range(0, 176400): #Just test samples
    if i % 32 < 16:
        Data += b'\x10\xe7\x10\xe7'
    else:
        Data += b'\xf0\xa8\xf0\xa8'

Update: I added a new function to handle previous formatting and return a form I can write directly to file. The solution I marked makes use of bytes but in python 2.7 I had to use bytearray.

def barray(bstring):
    hexmap = {'0':0, '1':1, '2':2, '3':3, '4':4, '5':5, '6':6, '7':7,'8':8, '9':9, 'a':10, 'b':11, 'c':12, 'd':13, 'e':14, 'f':15}
    bstring = bstring.lstrip('\\x')
    bstring = bstring.split('\\x')
    bstring = map(lambda x: 16*hexmap[x[0]]+hexmap[x[1]], bstring)
    return bytearray(bstring)

Now after I call the functions the data is ready to write to file without using eval().

variable = barray(endian(bytestring(tohex(n, m))))

Answer 1

Did you try using bytes ?

data = '\x00\x20\x10'
newdata = [int(i) for i in data.encode()]
print(newdata)
>>> [0, 32, 16]
print(bytes(newdata))
>>> b'\x00 \x10'  # it only looks like the middle value is missing, it isn't
print([i for i in b'\x00 \x10'])
>>> [0, 32, 16]  # yep, all still there

Edited based on a better understanding of the question after it was clarified.