Information Theory - HW2 (Lasha)

# LZcompress
from os.path import getsize as sz
import sys
from math import ceil, log
# Args
src, dst = open(sys.argv[1], 'rb'), open(sys.argv[2], 'wb+')
buff_text = src.read()
buff_binary = left = ''
lem = {"0": 0, "1": 1}
cur_index = 2
# Gamma
original_size = sz(sys.argv[1])
arr = ''.join(["{0:{f}8b}".format(byte, f='0') for byte in buff_text])
buff_binary += '0'*int(log(original_size, 2)) + \
    "{0:{f}8b}".format(original_size, f='0')
# Write to buffer
for bit in arr:
    if (left+bit) not in lem:
        inbin = "{0:b}".format(lem[left])
        buff_binary += (ceil(log(cur_index, 2)) - len(inbin)) * '0' + inbin
        lem[left + '0'], lem[left + '1'] = lem[left], cur_index
        cur_index += 1
        left = bit
    else:
        left += bit
# Padding
while left not in lem:
    left += '0'
inbin = "{0:b}".format(lem[left])
buff_binary += (ceil(log(cur_index, 2)) - len(inbin)) * '0' + inbin
buff_binary += '1{}'.format('0' * (7 - len(buff_binary) % 8))
# Write to file
for i in range(int(len(buff_binary) / 8)):
    num = int(buff_binary[i * 8:(i + 1) * 8], 2)
    dst.write(num.to_bytes(1, byteorder='big'))


# LZDecompress
from math import ceil, log
import os
import sys

sf, df = open(sys.argv[1], 'rb'), open(sys.argv[2], 'wb+')
lem = {0: "0", 1: "1"}
ind = 2
# read Gamma code
buf = sf.read(4096)
BUFTOREAD = ''.join(["{0:{f}8b}".format(byte, f='0') for byte in buf])
index = BUFTOREAD.index('1')
file_size = int('1{}'.format(''.join(BUFTOREAD[index + 1:2 * index + 1])), 2)
# Read Lempel-Ziv coding
BUFTOWRITE, BUFTOREAD = '', BUFTOREAD[2 * index + 1:]
while True:
    while ceil(log(len(lem), 2)) <= len(BUFTOREAD):
        i = int(BUFTOREAD[:ceil(log(len(lem), 2))], 2)
        BUFTOWRITE += lem[i]
        BUFTOREAD = BUFTOREAD[ceil(log(len(lem), 2)):]
        lem[ind], lem[i] = lem[i] + '1', lem[i] + '0'
        ind += 1
    buf = sf.read(4096)
    if not buf:
        break
    BUFTOREAD += ''.join(["{0:{f}8b}".format(byte, f='0') for byte in buf])
# Write to file
for i in range(int(len(BUFTOWRITE) / 8)):
    if file_size == 0:
        break
    num = int(BUFTOWRITE[i * 8:(i + 1) * 8], 2)
    df.write(num.to_bytes(1, byteorder='big'))
    file_size -= 1