"""
 feb2.py

 The huffman part of the Feb 2 homework, 
 using the routines in utilities.py and huffman.py

 
 
 
 Jim Mahoney | cs.marlboro.edu | Feb 2017 | MIT License
"""
from utilities import file_to_substrings, probabilities, entropy, mean
from huffman import Huffman

def analyze(filename, symbolsize, verbose=False):
    """ Return entropy and compression ratio for
        a file treated as symbols of a given length.
        Optionally print out lots of details. """

    def _print(message):
        if verbose:
            print(message)
    
    _print(" -- file {} with symbolsize {} --".format(filename, symbolsize))
    symbols = file_to_substrings(filename, symbolsize)
    p = probabilities(symbols)
    pvals = p.values()
    (mean_p, max_p, min_p) = (mean(pvals), max(pvals), min(pvals))
    _print(" symbol count = {}; unique symbols = {}".format(
        len(symbols), len(p)))
    _print(" probability max,mean,min = {:.4f}, {:.4f}, {:.4f}".format(
          max_p, mean_p, min_p))
    h = entropy(p)
    _print(" entropy = {:.4f} per symbol , = {:.4f} per bit".format
          (h, h/symbolsize))
    huff = Huffman(p)
    huff_mean = huff.mean_code_length()
    huff_max = huff.max_code_length()
    huff_min = huff.min_code_length()
    _print(" huffman code has max, mean, min = {}, {}, {}".format(
        huff_max, huff_mean, huff_min))
    compression = huff_mean/symbolsize
    _print(" compression factor is (huff_mean/symbolsize) = {:.4f}".format(
        compression))
    _print('')
    return (h/symbolsize, compression)

def entropies_and_compressions(verbose=False):
    """ Return a dictionary of entropy approximations
        and huffman compression ratios for several
        files and symbol lengths.
        Optionally print out lots (and lots) of details.
    """
    n_sizes = 8
    filenames = ('stream1.txt', 'stream2.txt')
    entropy = {file : [None]*n_sizes for file in filenames}
    compression = {file : [None]*n_sizes for file in filenames}
    for filename in ('stream1.txt', 'stream2.txt'):
        for symbolsize in range(1, n_sizes+1):
            (this_entropy, this_compression) = analyze(filename,
                                                       symbolsize,
                                                       verbose)
            entropy[filename][symbolsize-1] = this_entropy
            compression[filename][symbolsize-1] = this_compression
    return (entropy, compression)

if __name__=='__main__':
    entropies_and_compressions(verbose=True)