|
| 1 | +import struct |
| 2 | +from typing import List |
| 3 | + |
| 4 | +def lzw_compress(uncompressed: str, max_table_size: int = 4096) -> List[int]: |
| 5 | + """ |
| 6 | + Compress a string to a list of output symbols using the LZW algorithm. |
| 7 | +
|
| 8 | + :param uncompressed: The input string to compress. |
| 9 | + :param max_table_size: Maximum size for the dictionary. |
| 10 | + :return: A list of integer codes representing the compressed data. |
| 11 | + """ |
| 12 | + if not uncompressed: |
| 13 | + return [] |
| 14 | + |
| 15 | + # Initialize the dictionary with single-character mappings. |
| 16 | + dict_size = 256 |
| 17 | + dictionary = {chr(i): i for i in range(dict_size)} |
| 18 | + |
| 19 | + w = "" |
| 20 | + result = [] |
| 21 | + for c in uncompressed: |
| 22 | + wc = w + c |
| 23 | + if wc in dictionary: |
| 24 | + w = wc |
| 25 | + else: |
| 26 | + # Output the code for w. |
| 27 | + result.append(dictionary[w]) |
| 28 | + # Only add new sequences if we haven't reached the maximum table size. |
| 29 | + if dict_size < max_table_size: |
| 30 | + dictionary[wc] = dict_size |
| 31 | + dict_size += 1 |
| 32 | + w = c |
| 33 | + |
| 34 | + # Output the code for w if it's not empty. |
| 35 | + if w: |
| 36 | + result.append(dictionary[w]) |
| 37 | + |
| 38 | + return result |
| 39 | + |
| 40 | + |
| 41 | +def lzw_decompress(compressed: List[int], max_table_size: int = 4096) -> str: |
| 42 | + """ |
| 43 | + Decompress a list of output symbols to a string using the LZW algorithm. |
| 44 | +
|
| 45 | + :param compressed: The list of integer codes to decompress. |
| 46 | + :param max_table_size: Maximum size for the dictionary. |
| 47 | + :return: The original uncompressed string. |
| 48 | + """ |
| 49 | + if not compressed: |
| 50 | + return "" |
| 51 | + |
| 52 | + # Initialize the dictionary with single-character mappings. |
| 53 | + dict_size = 256 |
| 54 | + dictionary = [chr(i) for i in range(dict_size)] |
| 55 | + |
| 56 | + # Start with the first code. |
| 57 | + result = [] |
| 58 | + w = dictionary[compressed[0]] |
| 59 | + result.append(w) |
| 60 | + |
| 61 | + # Process the rest of the codes. |
| 62 | + for k in compressed[1:]: |
| 63 | + if k < len(dictionary): |
| 64 | + entry = dictionary[k] |
| 65 | + elif k == dict_size: |
| 66 | + # Special case: entry = w + w[0] |
| 67 | + entry = w + w[0] |
| 68 | + else: |
| 69 | + raise ValueError(f"Bad compressed code: {k}") |
| 70 | + |
| 71 | + result.append(entry) |
| 72 | + # Add new entry to the dictionary if the maximum size hasn't been reached. |
| 73 | + if dict_size < max_table_size: |
| 74 | + dictionary.append(w + entry[0]) |
| 75 | + dict_size += 1 |
| 76 | + w = entry |
| 77 | + |
| 78 | + return "".join(result) |
| 79 | + |
| 80 | + |
| 81 | +def lzw_compress_bytes(uncompressed: str, max_table_size: int = 4096) -> bytes: |
| 82 | + """ |
| 83 | + Compress a string and return a bytes object. |
| 84 | + |
| 85 | + This function wraps lzw_compress() by converting its list-of-int output into a bytes |
| 86 | + object, where each code is stored as an unsigned 16-bit integer. |
| 87 | + |
| 88 | + :param uncompressed: The input string to compress. |
| 89 | + :param max_table_size: Maximum size for the dictionary. |
| 90 | + :return: A bytes object containing the packed compressed data. |
| 91 | + """ |
| 92 | + # Get the list of integer codes from the standard LZW compression. |
| 93 | + codes = lzw_compress(uncompressed, max_table_size) |
| 94 | + |
| 95 | + # Pack each integer as an unsigned short (16 bits) in big-endian order. |
| 96 | + # Note: With max_table_size=4096 the codes fit in 12 bits and are safe to pack in 16 bits. |
| 97 | + return struct.pack('>' + 'H' * len(codes), *codes) |
| 98 | + |
| 99 | + |
| 100 | +def lzw_decompress_bytes(compressed_bytes: bytes, max_table_size: int = 4096) -> str: |
| 101 | + """ |
| 102 | + Decompress a bytes object (produced by lzw_compress_bytes) back to a string. |
| 103 | + |
| 104 | + This function unpacks the bytes object into a list of integers (codes) and then uses |
| 105 | + lzw_decompress() to reconstruct the original string. |
| 106 | + |
| 107 | + :param compressed_bytes: The bytes object containing the packed compressed data. |
| 108 | + :param max_table_size: Maximum size for the dictionary. |
| 109 | + :return: The original uncompressed string. |
| 110 | + """ |
| 111 | + # Each code is stored as 2 bytes. Compute the number of codes. |
| 112 | + num_codes = len(compressed_bytes) // 2 |
| 113 | + |
| 114 | + # Unpack the bytes object to a tuple of integers. |
| 115 | + codes = list(struct.unpack('>' + 'H' * num_codes, compressed_bytes)) |
| 116 | + |
| 117 | + # Decompress using the standard LZW decompression. |
| 118 | + return lzw_decompress(codes, max_table_size) |
| 119 | + |
| 120 | + |
| 121 | +# Example usage: |
| 122 | +if __name__ == "__main__": |
| 123 | + sample_text = "TOBEORNOTTOBEORTOBEORNOT" |
| 124 | + print("Original Text: ", sample_text) |
| 125 | + |
| 126 | + # Compress to a list of integer codes. |
| 127 | + codes = lzw_compress(sample_text) |
| 128 | + print("Compressed Codes:", codes) |
| 129 | + |
| 130 | + # Compress to bytes. |
| 131 | + compressed_bytes = lzw_compress_bytes(sample_text) |
| 132 | + print("Compressed Bytes:", compressed_bytes) |
| 133 | + |
| 134 | + # Decompress from list of integer codes. |
| 135 | + decompressed_text = lzw_decompress(codes) |
| 136 | + print("Decompressed Text (from codes):", decompressed_text) |
| 137 | + |
| 138 | + # Decompress from bytes. |
| 139 | + decompressed_text_bytes = lzw_decompress_bytes(compressed_bytes) |
| 140 | + print("Decompressed Text (from bytes):", decompressed_text_bytes) |
0 commit comments