def get_info_from_EndNote(file_url, return_source=False):
"""Populate the Publication with information from its profile"""
result = True
EndNode_file = None
while result and EndNode_file is None:
EndNode_file = utils.get_text_data(file_url)
if EndNode_file is None:
result = None
# while result is None:
# result = input('Do not load EndNote file from scholar. Try again? [Y/N]').lower()
# if result == "y": result = True
# elif result == "n": result = False
if EndNode_file is None:
logger.debug("Download empty EndNote file.")
return None
EndNode_file = EndNode_file.replace("\r", "")
logger.debug("EndNote file:\n%s" % EndNode_file)
EndNote_info = EndNote_parsing(EndNode_file)
if not EndNote_info:
return None
if "pages" in EndNote_info:
try:
pages = EndNote_info["pages"].split("-")
if len(pages) == 2:
start_page = pages[0].strip()
end_page = pages[1].strip()
re_st_page = re.search("[0-9]+", start_page)
re_end_page = re.search("[0-9]+", end_page)
if re_st_page:
EndNote_info["start_page"] = int(re_st_page.group(0))
if re_end_page:
EndNote_info["end_page"] = int(re_end_page.group(0))
if re_st_page and re_end_page:
EndNote_info["pages"] = abs(EndNote_info["end_page"] -
EndNote_info["start_page"] + 1)
else:
re_st_page = re.search("[0-9]+", EndNote_info["pages"])
EndNote_info["pages"] = int(re_st_page.group(0))
except Exception as error:
logger.warn("Can't eval count of pages for paper.")
try:
EndNote_info["pages"] = int(EndNote_info["pages"])
except BaseException:
EndNote_info["pages"] = None
if return_source:
EndNote_info.update({"EndNote": EndNode_file})
return EndNote_info
def train_text():
x_train, x_valid, x_test, y_train, y_valid, y_test = utils.get_text_data()
embedding = embedding_utils.Embedding()
rnn_units = 128
dropout = 0.2
recurrent_dropout = 0.2
num_classes = len(utils.text_label_list)
models = model.Text_classification_models(
units=rnn_units,
dropout=dropout,
recurrent_dropout=recurrent_dropout,
num_classes=num_classes)
rnn_model = models.bi_rnn()
rnn_model.fit(x_train, y_train, epochs=hm_epoch, batch_size=hm_batch)
rnn_model.evaluate(x_test, y_test, batch_size=hm_batch, verbose=0)
print('Saving model...')
rnn_model.save(model_path + '/rnn_model.h5')
tf.flags.DEFINE_integer("num_checkpoints", 5,
"Number of checkpoints to store (default: 5)")
# Misc Parameters
tf.flags.DEFINE_boolean("allow_soft_placement", True,
"Allow device soft device placement")
tf.flags.DEFINE_boolean("log_device_placement", False,
"Log placement of ops on devices")
FLAGS = tf.flags.FLAGS
FLAGS._parse_flags()
print("\nParameters:")
for attr, value in sorted(FLAGS.__flags.items()):
print("{}={}".format(attr.upper(), value))
print("")
X_train, Y_train, X_test, Y_true, vocab_processor = utils.get_text_data()
# print("fitting/saving")
# clf = Pipeline([ ("word2vec vectorizer", TfidfEmbeddingVectorizer(w2v)),
# ("logistic regression", linear_model.LogisticRegression())])
with tf.Graph().as_default():
session_conf = tf.ConfigProto(
allow_soft_placement=FLAGS.allow_soft_placement,
log_device_placement=FLAGS.log_device_placement)
sess = tf.Session(config=session_conf)
with sess.as_default():
cnn = TextCNN(sequence_length=X_train.shape[1],
num_classes=2,
vocab_size=len(vocab_processor.vocabulary_),
embedding_size=FLAGS.embedding_dim,
s = ['But lets face it: At the core of this line of thinking isnt safety -- its sex',
'--These are parts of their cars.']
# test_len = 10
max_generated_len = 10
x = tokenizer_en.tokenize(s)
print(x)
print(tokenizer_vi.merge([i for i in model.predict(x, max_len=max_generated_len)]))
optimizer = torch.optim.AdamW(model.parameters(), lr=0.001,)
scheduler = torch.optim.lr_scheduler.StepLR(optimizer, step_size=1, gamma=0.82)
train_en_text, train_vi_text, valid_en_text, valid_vi_text, test_en_text, test_vi_text = get_text_data()
train_en, valid_en, test_en = [tokenizer_en.tokenize(i) for i in [train_en_text, valid_en_text, test_en_text]]
train_vi, valid_vi, test_vi = [tokenizer_vi.tokenize(i) for i in [train_vi_text, valid_vi_text, test_vi_text]]
def eval(valid_en=valid_en, valid_vi=valid_vi, full_detail=False, confusion=False):
y_true = []
y_pred = []
total_loss = []
batch_size = config.batch_size
s = ['But lets face it: At the core of this line of thinking isnt safety -- its sex',
'Process finished with exit code 0']
max_generated_len = 20
x = tokenizer_en.tokenize(s)
#
# Main script to actually run cipher decryption
#
from utils import get_args, get_text_data
from substitution import Substitution
from vigenere import Vigenere
from caesar import Caesar
args = get_args()
file_data = get_text_data(args['FILENAME'])
text = file_data
if args['CIPHER'] == 'SUBSTITUTION':
substitution = Substitution()
if args['SHOULD_ENCRYPT']:
key = args['ENCRYPTION_KEY']
encrypted = substitution.encrypt(text, key)
print(encrypted)
elif args['SHOULD_DECRYPT']:
decrypted = substitution.decrypt(text)
print(decrypted)
elif args['CIPHER'] == 'VIGENERE':
vigenere = Vigenere()
if args['SHOULD_ENCRYPT']:
key = args['ENCRYPTION_KEY']
encrypted = vigenere.encrypt(text, key)