letter_compress.py

   1 #!/usr/bin/env python3
   2
   3 import bitstring
   4
   5 from collect.bidict import bidict
   6
   7 special_characters = bidict(
   8     {
   9         ' ': 27,
  10         '.': 28,
  11         ',': 29,
  12         "-": 30,
  13         '"': 31,
  14     }
  15 )
  16
  17
  18 def compress(uncompressed: str) -> bytes:
  19     """
  20     Compress a word sequence into a stream of bytes.  The compressed
  21     form will be 5/8th the size of the original.  Words can be lower
  22     case letters or special_characters (above).
  23
  24     >>> import binascii
  25     >>> binascii.hexlify(compress('this is a test'))
  26     b'a2133da67b0ee859d0'
  27
  28     >>> binascii.hexlify(compress('scot'))
  29     b'98df40'
  30
  31     >>> binascii.hexlify(compress('scott'))
  32     b'98df4a00'
  33
  34     """
  35     compressed = bitstring.BitArray()
  36     for (n, letter) in enumerate(uncompressed):
  37         if 'a' <= letter <= 'z':
  38             bits = ord(letter) - ord('a') + 1   # 1..26
  39         else:
  40             if letter not in special_characters:
  41                 raise Exception(f'"{uncompressed}" contains uncompressable char="{letter}"')
  42             bits = special_characters[letter]
  43         compressed.append(f"uint:5={bits}")
  44     while len(compressed) % 8 != 0:
  45         compressed.append("uint:1=0")
  46     return compressed.bytes
  47
  48
  49 def decompress(kompressed: bytes) -> str:
  50     """
  51     Decompress a previously compressed stream of bytes back into
  52     its original form.
  53
  54     >>> import binascii
  55     >>> decompress(binascii.unhexlify(b'a2133da67b0ee859d0'))
  56     'this is a test'
  57
  58     >>> decompress(binascii.unhexlify(b'98df4a00'))
  59     'scott'
  60
  61     """
  62     decompressed = ''
  63     compressed = bitstring.BitArray(kompressed)
  64     for chunk in compressed.cut(5):
  65         chunk = chunk.uint
  66         print(f'0x{chunk:x}')
  67         if chunk == 0:
  68             break
  69         elif 1 <= chunk <= 26:
  70             letter = chr(chunk - 1 + ord('a'))
  71         else:
  72             letter = special_characters.inverse[chunk][0]
  73         decompressed += letter
  74     return decompressed
  75
  76
  77 if __name__ == '__main__':
  78     import doctest
  79     doctest.testmod()