# Markov chain generator.
#
# Run with:
#
#  $ python markov-simple.py < cnus-clean.txt
#
# The accompanying cnus-clean.txt is the complete works of Sherlock Holmes
# cleaned up from special characters, excessive line breaks and lowercased.
#
# Requires Python 3.6+
#
# Eli Bendersky [http://eli.thegreenplace.net]
# This code is in the public domain.

from collections import defaultdict, Counter
import random
import sys

# This is the length of the "state" the current character is predicted from.
# For Markov chains with memory, this is the "order" of the chain. For n-grams,
# n is STATE_LEN+1 since it includes the predicted character as well.

STATE_LEN = 6

data = sys.stdin.read().lower()                        #occa type(data) is 'str'
print('Testo lungo {0} caratteri'.format(len(data)))   #occa

model = defaultdict(Counter)
print('Learning model...')
for i in range(len(data) - STATE_LEN):
    state = data[i:i + STATE_LEN]        #occa  state e' str per def dinamica
    next = data[i + STATE_LEN]
    model[state][next] += 1

print('Inizio testo: ',data[:110])
print('STATE_LEN =',STATE_LEN, '.  Model has {0} states'.format(len(model)),'\n')
j = 0
for k, v in model.items():
    print(k, v)
    if j > 9:
        break
    j += 1

print('\nSampling...')
state = random.choice(list(model))
print('stato iniziale scelto a caso:',state)                      #occa
out = list(state)  #occa trasforma la stringa in lista per elaboraz seguente
for i in range(400):
    out.extend(random.choices(list(model[state]), model[state].values()))
#    out.extend(random.choices(model[state].keys(), model[state].values()))  # err
    state = state[1:] + out[-1]
print(''.join(out))