import codecs
import hashlib
import json
import pickle
import re
import sys

#32 or 64 bit platform?
if sys.maxsize > 2**32:
    HASH_FUNC = hashlib.blake2b
else:
    HASH_FUNC = hashlib.blake2s

def load_words(filename):
    with open(filename) as file:
        text = file.read()
    return set(map(str.lower,filter(bool,text.split('\n'))))

def _get_wordlist_hash(word_list_s):
    _hash = HASH_FUNC()
    for word in sorted(word_list_s):
        word_bytes = word.encode()
        _hash.update(word_bytes)
    return _hash.digest()

def hash_wordlist(word_list,raw = False):
    word_list = sorted(word_list)
    fhash = _get_wordlist_hash(word_list)
    if raw:
        return fhash
    illegal_hash = codecs.encode(fhash,'base64').decode()
    t_table = str.maketrans({'+':'-','/':'_'})
    return illegal_hash.translate(t_table)

def load_freq_cache(word_list):
    fname = hash_wordlist(word_list) + '.pkl'
    fname = os.path.join('__hangchache__',fname)
    if os.path.exists(fname):
        with open(fname,'rb') as file:
            return pickle.load(file)

def save_freq_cache(word_list,freq):
    if not os.path.exists('__hangchache__'):
        os.mkdir('__hangchache__')
    fname = hash_wordlist(word_list) + '.pkl'
    fname = os.path.join('__hangchache__',fname)
    with open(fname,'wb') as file:
        pickle.dump(file,freq)

def generate_letter_frequency(word_list):
    cached = load_freq_cache(word_list)
    if cached is not None:
        return cached
    ret = {}
    for word_num,word in enumerate(word_list):
        letter_counts = {}
        for i,letter in enumerate(word):
            try:
                ret[letter][0] += 1
            except KeyError:
                ret[letter] = [1,0]
            in_word = letter_counts.get(letter,0) + 1
            letter_counts[letter] = in_word
        for letter,count in letter_counts.items():
            word_portion = count/len(word)
            avg = (ret[letter][1] * word_num)  + word_portion
            avg /= word_num + 1
            ret[letter][1] = avg
    if cached is None:
        save_freq_cache(word_list,ret)
    return ret
PROMPT = "Enter word with '.' to represent missing letters: "
def iterate(word_list,let_freq):
    entered_word = input(PROMPT)
    entered_word = entered_word.replace(' ')
    entered_letters = set(letter for letter in entered_word.replace('.',''))
    remaining_letters = set(let_freq.keys()) - entered_letters
    regex = entered_word.replace('.','[A-Za-z]')
    remaining_possibilities = list(filter(lambda word: re.match(regex,word),word_list))
    print('Matches found:\n' + '\n'.join(remaining_possibilities[i] for i in range(min(30,len(remaining_possibilities)))))
    print( 'Good candidates by overall frequency:\n' + '\n'.join(sorted(remaining_letters,key = lambda letter: let_freq[letter][0],reverse = True)) )
    print('Good candidates by per-word frequency:\n' + '\n'.join(sorted(remaining_letters,key = lambda letter: let_freq[letter][1],reverse = True)) )
    return entered_word,remaining_possibilities

if __name__ == "__main__":
    words = load_words('words.txt')
    FREQ = generate_letter_frequency(words)
    while True:
        try:
            last,WORDS = iterate(words,FREQ)
        except KeyboardInterrupt:
            break