def __init__(self, tagger_backup):
t0 = time()
print 'Initialisation ...'
self.extracter = Extracter()
self.tokenizer = Tokenizer()
self.tagger = Tagger(name=tagger_backup)
self.replacer = AntonymReplacer()
self.lemmatizer = Lemmatizer()
print 'Initialisation done in %0.3fs.' % (time() - t0)
sw = word_tokenize('Sup! Awesome bday? brb!')
sw2 = ""
for word in sw:
result = replacer.replace(word)
sw2 += result + " "
print(sw2)
# Tutorial 17 : https://www.youtube.com/watch?v=88yOCX4ZyoA&list=PLcTXcpndN-Sl9eYrKM6jtcOTgC52EJnqH&index=17
from replacer import AntonymReplacer
rep = AntonymReplacer()
antn = rep.replace('cowardice')
print(antn)
antn = rep.replace('heavy')
print(antn)
antn = rep.replace('weak')
print(antn)
antn = rep.replace('blue')
print(antn)
sent = rep.negreplace('this man is not salty')
print(sent)
class Preprocesser(object):
def __init__(self, tagger_backup):
t0 = time()
print 'Initialisation ...'
self.extracter = Extracter()
self.tokenizer = Tokenizer()
self.tagger = Tagger(name=tagger_backup)
self.replacer = AntonymReplacer()
self.lemmatizer = Lemmatizer()
print 'Initialisation done in %0.3fs.' % (time() - t0)
def preprocess(self, text):
t0 = time()
preprocessed_jokes = []
for joke in text:
# Jokes tokenization...
tokenized_joke = self.tokenizer.tokenize(joke, remove_stopwords=False)
# Jokes tokenization done!
# Jokes POS Tagging...
tagged_sentences = self.tagger.tag_sentences(tokenized_joke)
tagged_joke = self.tagger.map_tag(tagged_sentences)
# Jokes POS Tagging done!
# Jokes negation replacement...
tagged_joke2 = []
for sentence in tagged_joke:
tagged_joke2.append(self.replacer.replace_negations(sentence))
# Jokes negation replacement done!
# Jokes lemmatization...
lemmatized_joke = self.lemmatizer.lemmatize_sentences(tagged_joke2)
# Jokes lemmatization done!
# Jokes removing stopwords and punctuation...
normalized_joke = self.extracter.remove_stopwords(lemmatized_joke)
# Jokes removing stopwords and punctuation!
# Jokes transformation from list of tokens into concatenation of tokens...
preprocessed_joke = ''
for normalized_sentence in normalized_joke:
preprocessed_joke = preprocessed_joke +' '+' '.join(normalized_sentence)
preprocessed_jokes.append(preprocessed_joke.strip())
# Jokes transformation from list of tokens into concatenation of tokens!
print 'Preprocessing done in %0.3fs.' % (time() - t0)
return preprocessed_jokes
# def preprocess(self, path='data/jokes/'):
# #path = 'data/jokes/'
# html_jokes = self.retriever.read_from_directory(path)
# # Jokes extraction from HTML documents
# jokes = []
# for html_joke in html_jokes:
# jokes.append(extracter.extract_joke(html_joke))
# #jokes.append(extracter.replace(extracter.extract_text(html_joke)))
# # Jokes tokenization
# tokenized_jokes = []
# for joke in jokes:
# #without_punc = extracter.remove_punctuation(joke)
# tokenized_jokes.append(tokenizer.tokenize(joke, remove_stopwords=False))
# # Jokes POS Tagging
# tagged_jokes = []
# for sentences in tokenized_jokes:
# tagged_sentences = tagger.tag_sentences(sentences)
# tagged_jokes.append(tagger.map_tag(tagged_sentences))
# # Jokes negation replacement
# tagged_jokes_2 = []
# for sentences in tagged_jokes:
# tg_sentences = []
# for sentence in sentences:
# tg_sentences.append(replacer.replace_negations(sentence))
# tagged_jokes_2.append(tg_sentences)
# # Jokes lemmatization
# lemmatized_jokes = []
# for sentences in tagged_jokes_2:
# lemmatized_jokes.append(lemmatizer.lemmatize_sentences(sentences))
# # Jokes removing stopwords and punctuation
# normalized_jokes = []
# for lemmatized_joke in lemmatized_jokes:
# normalized_jokes.append(extracter.remove_stopwords(lemmatized_joke))
# # Jokes transformation from list of tokens into concatenation of tokens
# preprocessed_jokes = []
# for normalized_joke in normalized_jokes:
# preprocessed_joke = ''
# for normalized_sentence in normalized_joke:
# preprocessed_joke = preprocessed_joke +' '+' '.join(normalized_sentence)
# preprocessed_jokes.append(preprocessed_joke.strip())
#
# return preprocessed_jokes
