lz78/lz_78_rle.py

   1 import sys
   2 import struct
   3
   4 ########################################################################################
   5 # Apenas comprimir
   6 #
   7 # python lz_78_rle.py <caminho_para_o_ficheiro_original>
   8 #
   9 #
  10 # Os ficheiros gerados ficarão na mesma pasta que o script
  11 ########################################################################################
  12
  13 def rle_transform(bitstream):
  14     sequencia = []
  15     pixel_atual = '0' # Começa sempre com o pixel 0 conforme enunciado
  16     contador = 0
  17
  18     for bit in bitstream:
  19         if bit == pixel_atual:
  20             if contador == 255: # Limite de 8 bits
  21                 sequencia.append(255)
  22                 sequencia.append(0) # Troço de 0 do outro pixel para continuar no mesmo [cite: 41]
  23                 contador = 1
  24             else:
  25                 contador += 1
  26         else:
  27             sequencia.append(contador)
  28             pixel_atual = bit
  29             contador = 1
  30     sequencia.append(contador)
  31     return sequencia
  32
  33 def lz78_compression(image_data):
  34     dictionary = {0: []}
  35     output = []
  36     symbol = []
  37
  38     for i in image_data:
  39         temp = symbol + [i]
  40         dict_values = list(dictionary.values())
  41         if temp in dict_values:
  42             symbol = temp
  43         else:
  44             if not symbol:
  45                 output.append([0, i])
  46                 dictionary[len(dictionary)] = [i]
  47             else:
  48                 idx = dict_values.index(symbol)
  49                 output.append([idx, i])
  50                 dictionary[len(dictionary)] = symbol + [i]
  51             symbol = []
  52
  53     if symbol:
  54         idx = list(dictionary.values()).index(symbol)
  55         output.append([idx, -1])
  56     return output
  57
  58 def clean_image_data(image_data):
  59
  60     clean_text = ""
  61
  62     for line in image_data:
  63         line = line.strip()
  64         line = line.replace(" ", "")
  65         if not line or line.startswith('#'):
  66             continue
  67         clean_text += line
  68
  69     return clean_text
  70
  71 def chunkstring(string, length):
  72     return [string[i : length+i] for i in range(0, len(string), length)]
  73
  74
  75 with open(sys.argv[1], "r") as image:
  76     lines = image.readlines()
  77     image_size = lines[1].split()
  78
  79     if lines[0].strip() != 'P1':
  80         raise ValueError("This is not a pbm file")
  81
  82     clean_text = clean_image_data(lines[2:]) #Limpa os espaços em branco para comprimir melhor
  83
  84     clean_text = rle_transform(clean_text)
  85
  86     output = lz78_compression(clean_text)
  87
  88     print(f"Output: {output}\n")
  89
  90     #Compressão em texto (human-readable)
  91     with open(sys.argv[1] + "_compressed", "w") as save_file:
  92         save_file.write(f"{output} \n{image_size[0]} {image_size[1]}")
  93
  94     #Compressão em binário (compressão "a sério")
  95     with open(sys.argv[1] + ".bin", "wb") as save_file:
  96         largura = int(image_size[0])
  97         altura = int(image_size[1])
  98         save_file.write(struct.pack('II', largura, altura))
  99
 100         for indice, val in output:
 101             simbolo_byte = val if val != -1 else 255
 102             save_file.write(struct.pack('HB', indice, simbolo_byte))