neural_fuzz_xml_1.py

"""


"""

from __future__ import print_function
from keras.models import Sequential
from keras.layers import Dense, Activation
from keras.layers import LSTM
from keras.optimizers import RMSprop
from keras.utils.data_utils import get_file
import numpy as np
import random
import sys
import io

import datetime
from keras.utils import plot_model

# read training data from file (data set)

text = open('trainset/larg.xml', encoding="utf8").read()
print('corpus length:', len(text))

chars = sorted(list(set(text)))
print('total chars:', len(chars))
print(chars)

char_indices = dict((c, i) for i, c in enumerate(chars))
print(char_indices)
indices_char = dict((i, c) for i, c in enumerate(chars))
print(indices_char)


# cut the text in semi-redundant sequences of maxlen characters
maxlen = 40
step = 3
sentences = [] # list of all sentence
next_chars = [] # list of all next chars
for i in range(0, len(text) - maxlen, step): # arg2 why this?
    sentences.append(text[i: i + maxlen])
    #print(sentences)
    next_chars.append(text[i + maxlen])
    #print(next_chars)
    

print('semi sequences:', len(sentences))

print('One-Hot vectorization...')
x = np.zeros((len(sentences), maxlen, len(chars)), dtype=np.bool) #input x
y = np.zeros((len(sentences), len(chars)), dtype=np.bool) #output label y
for i, sentence in enumerate(sentences):
    for t, char in enumerate(sentence):
        x[i, t, char_indices[char]] = 1
    y[i, char_indices[next_chars[i]]] = 1


# build the model: a single LSTM layer # we need to deep it
print('Build model...')
model = Sequential()
model.add(LSTM(128, input_shape=(maxlen, len(chars))))
model.add(Dense(len(chars)))
model.add(Activation('softmax'))

optimizer = RMSprop(lr=0.01)
model.compile(loss='categorical_crossentropy', optimizer=optimizer)

#sys.exit()

# plot the model
dt = datetime.datetime.now().strftime('%Y%m%d_%H%M%S')
plot_model(model, to_file = 'lstm_text_generation_edited1_' + dt + '.png', show_shapes=True, show_layer_names=True)

# lower temperature will cause the model to make more likely,
# but also more boring and conservative predictions.
def sample(preds, temperature=1.0):
    # helper function to sample an index from a probability array
    preds = np.asarray(preds).astype('float64')
    preds = np.log(preds) / temperature
    exp_preds = np.exp(preds)
    preds = exp_preds / np.sum(exp_preds)
    probas = np.random.multinomial(1, preds, 1)
    return np.argmax(probas)

# train the model, output generated text after each iteration
for iteration in range(1, 50):
    print()
    print('-' * 50)
    print('- Iteration', iteration)
    model.fit(x, y, batch_size=128, epochs=3) # why epochs=?

     #choose random start index for seed sentence from trian set
    start_index = random.randint(0, len(text) - maxlen - 1)
    for diversity in [0.2, 0.5, 1.0, 1.2, 1.5]:
        print()
        print('-- Diversity:', diversity)

        #generated = ''
        sentence = text[start_index: start_index + maxlen] # len(sentence) equals 40 here
        generated = sentence
        print('-- Generating text with seed: "' + sentence + '"')
        sys.stdout.write(generated)

        for i in range(500):
            x_pred = np.zeros((1, maxlen, len(chars)))
            for t, char in enumerate(sentence):
                x_pred[0, t, char_indices[char]] = 1.

            preds = model.predict(x_pred, verbose=0)[0]
            next_index = sample(preds, diversity)
            next_char = indices_char[next_index]

            generated += next_char
            sentence = sentence[1:] + next_char

            sys.stdout.write(next_char)
            sys.stdout.flush()
        #save generated text to file inside program
        file_name = 'XMLs/' + 'step' + repr(step) + '_itr' + repr(iteration) + '_div' + repr(diversity) + '.txt'
        with open(file_name, 'w') as generated_text_file:
             generated_text_file.write(generated)
        print()
        

# Save model (requires HDF5 and h5py)
dt = datetime.datetime.now().strftime('%Y%m%d_%H%M%S')
model.save('lstm_text_generation_edited1_'+ dt + '.h5')