# 下載小說文本庫(kù)
import urllib.request
url = 'https://cloud.tencent.com/developer/attachments/1632706'
response = urllib.request.urlopen(url)
text = response.read().decode('utf-8')
print(text)

# 對(duì)文本進(jìn)行處理
import string
from collections import Counter
import re
# 清理文本
text = text.lower()
text = re.sub(r'\W+', ' ', text)
text = text.split()

# 構(gòu)建LSTM模型
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import Dropout
from keras.layers import LSTM
from keras.callbacks import ModelCheckpoint
from keras.utils import np_utils
# 創(chuàng)建序列，將文本進(jìn)行X, y的拆分
sequence_length = 100
sequences = []
for i in range(sequence_length, len(text)):
    seq = text[i-sequence_length:i+1]
    sequences.append(' '.join(seq))
# 構(gòu)建字典
words = ' '.join(sequences).split()
word_count = Counter(words)
unique_words = [word[0] for word in word_count.most_common(2000)]
word_to_index = {}
index_to_word = {}
for i, word in enumerate(unique_words):
    word_to_index[word] = i
    index_to_word[i] = word
# 構(gòu)建x和y
X = []
y = []
for sequence in sequences:
    seq = sequence.split()
    X_sequence = [word_to_index[word] for word in seq[:-1]]
    X.append(X_sequence)
    y.append(word_to_index[seq[-1]])
# 將x和y轉(zhuǎn)換為Numpy數(shù)組
X = np.array(X)
y = np_utils.to_categorical(y, num_classes=len(unique_words))
# 創(chuàng)建LSTM模型
model = Sequential()
model.add(LSTM(512, input_shape=(X.shape[1], X.shape[2]), return_sequences=True))
model.add(Dropout(0.2))
model.add(LSTM(512))
model.add(Dropout(0.2))
model.add(Dense(y.shape[1], activation='softmax'))
# 編譯模型
model.compile(loss='categorical_crossentropy', optimizer='adam')

# 訓(xùn)練模型
filepath="model_weights_saved.hdf5"
checkpoint = ModelCheckpoint(filepath, monitor='loss', verbose=1, save_best_only=True, mode='min')
model.fit(X, y, epochs=50, batch_size=512, callbacks=[checkpoint])
# 加載已經(jīng)訓(xùn)練好的模型
filename = "model_weights_saved.hdf5"
model.load_weights(filename)
model.compile(loss='categorical_crossentropy', optimizer='adam')

# 生成文本
from random import randint
from pickle import dump
from keras.models import load_model
from keras.preprocessing.sequence import pad_sequences
# 檢索一到多個(gè)不同長(zhǎng)度序列中的隨機(jī)序列
def generate_seq(model, mapping, seq_length, seed_text, n_chars):
    in_text = seed_text
    # 結(jié)果文本
    results = []
    # 進(jìn)行文本測(cè)試，循環(huán)獲取字符
    for _ in range(n_chars):
        # 將input字符進(jìn)行文本處理，轉(zhuǎn)為索引號(hào)
        encoded = [mapping[word] for word in in_text.split()]
        # 進(jìn)行padding，確保到達(dá)公共長(zhǎng)度
        encoded = pad_sequences([encoded], maxlen=seq_length, truncating='pre')
        # 模型進(jìn)行測(cè)試，預(yù)測(cè)下一個(gè)字符
        yhat = model.predict_classes(encoded, verbose=0)
        # 將文本轉(zhuǎn)化為字符輸出
        out_word = ''
        for word, index in mapping.items():
            if index == yhat:
                out_word = word
                break
        # 添加結(jié)果
        in_text += ' ' + out_word
        results.append(out_word)
    return ' '.join(results)
# 選擇一個(gè)隨機(jī)序列
seed_text = sequences[randint(0,len(sequences))]
print(seed_text + '\n')
# 用模型生成新的文本
generated = generate_seq(model, word_to_index, sequence_length, seed_text, 50)
print(generated)