def funcao_limpa_tudo(artigo): #limpa a palavra, limita a palavra em uma
lista_nova = []
#logger.info('Setting it to 0, do not use it in your scoring function.')
#print(docs)
artigo = stem_text(artigo)
artigo = split_alphanum(artigo)
artigo = tokenizer.tokenize(
artigo) #artigo vira uma lista com todas as palavras
list_artigo = list(artigo) #pesquisar o que é o list
try:
for palavra in list_artigo:
palavra = palavra.encode('utf-8')
if re.match('^\d+$', palavra):
#artigo.remove(palavra)
pass
elif palavra in stop_words:
#artigo.remove(palavra)
pass
elif len(palavra) < 3:
#artigo.remove(palavra)
pass
else:
lista_nova.append(palavra) #recebe palavra util
except Exception as erro: #evita travar o codigo e continua,
numero_palavra = artigo.index(palavra)
artigo.pop(numero_palavra) #achou lixo
del list_artigo #garbage do python
#log.info("Lista de Tokens: %s", docs)
if len(lista_nova) < 5:
return None
else:
return lista_nova
def transformText(text):
text = split_alphanum(text)
# Convert text to lower
text = text.lower()
text = replace_word(text)
text = replace_numbers(text)
text = reduce_lengthening(text)
text = remove_alphanumerics(text)
# Removing non ASCII chars
text = re.sub(r'[^\x00-\x7f]', r' ', text)
# Removing all the stopwords
filtered_words = [word for word in text.split() if word not in STOP_WORDS]
# Preprocessed text after stop words removal
text = " ".join(filtered_words)
# Remove the punctuation
text = gensim.parsing.preprocessing.strip_punctuation2(text)
# remove html markup
text = re.sub("(<.*?>)", "", text)
# Correct words
spell = SpellChecker()
misspelled = text.split()
wordnet_lemmatizer = WordNetLemmatizer()
for i in range(len(misspelled)):
# Get the one `most likely` answer
word = spell.correction(misspelled[i])
misspelled[i] = word
misspelled[i] = wordnet_lemmatizer.lemmatize(misspelled[i], pos="v")
misspelled[i] = wordnet_lemmatizer.lemmatize(misspelled[i], pos="n")
text = " ".join(misspelled)
# Strip multiple whitespaces
text = gensim.corpora.textcorpus.strip_multiple_whitespaces(text)
# Strip all the numerics
#text = gensim.parsing.preprocessing.strip_numeric(text)
return text
def raw_text_preprocess(d):
d = re.sub(r"http\S+", "", d)
d = strip_non_alphanum(d).lower().strip()
d = split_alphanum(d)
d = strip_short(d, minsize=2)
d = strip_numeric(d)
d = ViTokenizer.tokenize(d)
return d
def raw_text_preprocess(raw):
raw = re.sub(r"http\S+", "", raw)
raw = strip_non_alphanum(raw).lower().strip()
raw = split_alphanum(raw)
raw = strip_short(raw, minsize=2)
raw = strip_numeric(raw)
raw = ViTokenizer.tokenize(raw)
return raw
def preprocess_mail(mail):
mail = re.sub("https\S+", "", mail) # Loại bỏ các đường dẫn
mail = strip_non_alphanum(mail).lower().strip(
) # Loại bỏ các kí tự không phải là chữ cái, chuyển tất cả kí tự thành chữ thường
mail = split_alphanum(mail) # Tách văn bản thành các từ
mail = strip_short(
mail, minsize=2
) # Lấy các từ có độ dài >= 2 kí tự, loại bỏ các từ có 1 chữ cái
mail = strip_numeric(mail) #
mail = ViTokenizer.tokenize(mail)
return mail
def convert_num_words(cls, text):
"""
This method will numbers to text which help preprocessing much easier.
"""
text = split_alphanum(text)
temp_str = text.split()
new_string = [
INFLECT_ENGINE.number_to_words(word) if word.isdigit() else word
for word in temp_str
]
temp_str = " ".join(new_string)
return temp_str
def preprocess_text(text):
text = parse_html_v2(text)
text = text.lower()
text = remove_links_content(text)
text = remove_emails(text)
text = remove_special_tags(text) # remove content between {}
text = remove_punctuation(text) # remove all puntuations
text = split_alphanum(text) # add space between word and numeric
text = strip_numeric(text) # remove digits
text = strip_non_alphanum(text) # remove non-alphabetic characters
text = strip_short(text, minsize=2) # remove word with length < minsize
text = remove_multiple_space(text).strip() # remove space and strip
text = ViTokenizer.tokenize(text)
return text
def gensim_clean_string(textIn, _strip_tags=True, _split_alphanumeric=True, _strip_nonalphanumeric=True,
_strip_muliple_whitespace=True, _strip_short=True, _short_charcount_min=3,
_strip_punctuation=False, _convert_to_lower = False):
cleaner = textIn
if _strip_tags:
cleaner = strip_tags(textIn)
if _strip_nonalphanumeric:
cleaner = strip_non_alphanum(cleaner)
if _strip_muliple_whitespace:
cleaner = strip_multiple_whitespaces(cleaner)
if _split_alphanumeric:
cleaner = split_alphanum(cleaner)
if _strip_short:
cleaner = strip_short(cleaner, minsize=_short_charcount_min)
if _convert_to_lower:
cleaner = cleaner.lower()
return cleaner
def transformText(text):
text = split_alphanum(text)
# Convert text to lower
text = text.lower()
text = replace_word(text)
text = normaliser_word(text)
#stops = set(stopwords.words("english"))
stops={'at', 'only', 'your', 'yourself', 'a', 'i', 'during', 'off', 'myself', 'so', 'o', 'after', 'under',
'there', 'against', 'over', 'ourselves', 'they', 'me', 'its', 'then', 'above', 'theirs', 'this', 'into',
'from', 'very', 'on', 'yours', 'yourselves', 'herself', 'themselves', 'between', 'if', 'below', 'own',
'and', 'you', 'itself', 'him', 'while', 's', 'who', 'we', 'what', 'by', 'ma', 'further', 'such', 'until',
'through', 'too', 'until', 'through', 't', 'too', 'where', 'up', 'my', 'm', 'out', 'down', 're', 'to',
'she', 'd', 'those', 'when', 'it', 'because', 'he', 'in', 'other','each', 'both', 'her', 'but', 'as', 'all',
'his', 'again', 'with', 'once', 'am', 'just', 'should', 'why', 'than', 'any', 'should', 'why', 'than',
'more', 'most', 'that', 've', 'will', 'ours', 'our', 'll', 'the', 'y', 'which', 'whom', 'hers', 'an', 'here',
'how', 'before', 'about', 'for', 'them', 'these', 'their', 'for', 'them', 'these', 'their', 'or', 'must',
'shall', 'would', 'could' , 'need', 'might'}
# Removing non ASCII chars
text = re.sub(r'[^\x00-\x7f]',r' ',text)
# Removing all the stopwords
filtered_words = [word for word in text.split() if word not in stops]
# Preprocessed text after stop words removal
text = " ".join(filtered_words)
# Remove the punctuation
text = gensim.parsing.preprocessing.strip_punctuation2(text)
# Correct words
spell = SpellChecker()
misspelled = text.split()
wordnet_lemmatizer = WordNetLemmatizer()
for i in range(len(misspelled)):
# Get the one `most likely` answer
word = spell.correction(misspelled[i])
misspelled[i]=word
misspelled[i] = wordnet_lemmatizer.lemmatize(misspelled[i], pos="v")
misspelled[i] = wordnet_lemmatizer.lemmatize(misspelled[i], pos="n")
text = " ".join(misspelled)
# Removing all the stopwords
filtered_words = [word for word in text.split() if word not in stops]
text = " ".join(filtered_words)
# Strip multiple whitespaces
text = gensim.corpora.textcorpus.strip_multiple_whitespaces(text)
return text
def text_preprocess(bodyItem): # bodyItem: string (of one mail) => return: list of words (of one mail)
# Remove http, https
bodyItem = re.sub(r'^https?:\/\/.*[\r\n]*', '', bodyItem, flags=re.MULTILINE)
bodyItem = re.sub(r'^http?:\/\/.*[\r\n]*', '', bodyItem, flags=re.MULTILINE)
bodyItem = re.sub(r'''(?i)\b((?:https?://|www\d{0,3}[.]|[a-z0-9.\-]+[.][a-z]{2,4}/)(?:[^\s()<>]+|\(([^\s()<>]+|(\([^\s()<>]+\)))*\))+(?:\(([^\s()<>]+|(\([^\s()<>]+\)))*\)|[^\s`!()\[\]{};:'".,<>?«»“”‘’]))''', " ", bodyItem)
# Decode some bodyItems which are not decoded
bodyItem = bodyItem.replace("=", "%")
bodyItem = urllib.parse.unquote(bodyItem)
# Remove a word which has numbers and alphabets
bodyItem = strip_number_alphabets(bodyItem)
# Remove meaningless words, convert to lower words and split meaningful words
bodyItem = strip_non_alphanum(bodyItem).lower().strip()
bodyItem = split_alphanum(bodyItem)
# Join two words which have meaning in Vietnamese. Ex: hội thảo -> hội_thảo
bodyItem = ViTokenizer.tokenize(bodyItem)
# Remove a word which has one letter
bodyItem = strip_short(bodyItem, minsize=2)
# Remove stopwords
words = [word for word in bodyItem.split() if word not in stopwordsVN_ENG.getStopwordsVN_ENG()]
return words
def funcao_limpa_tudo(artigo):
tokenizer = RegexpTokenizer(r'\w+')
lista_nova = []
#logger.info('Setting it to 0, do not use it in your scoring function.')
#print(docs)
artigo = stem_text(artigo)
artigo = split_alphanum(artigo)
artigo = tokenizer.tokenize(artigo)
list_artigo = list(artigo)
try:
for palavra in list_artigo:
palavra = palavra.encode('utf-8')
if re.match('^\d+$', palavra):
#artigo.remove(palavra)
pass
elif palavra in pt_stop:
#artigo.remove(palavra)
pass
elif len(palavra) < 3:
#artigo.remove(palavra)
pass
else:
lista_nova.append(palavra)
except Exception as erro:
print(erro)
#numero_palavra = artigo.index(palavra)
#artigo.pop(numero_palavra) #achou lixo
pass
del list_artigo
del artigo
#log.info("Lista de Tokens: %s", docs)
if len(lista_nova) < 5:
return None
else:
return lista_nova
def test_split_alphanum(self):
self.assertEqual(split_alphanum("toto diet1 titi"), "toto diet 1 titi")
self.assertEqual(split_alphanum("toto 1diet titi"), "toto 1 diet titi")
def testSplitAlphanum(self):
self.assertEqual(split_alphanum("toto diet1 titi"), "toto diet 1 titi")
self.assertEqual(split_alphanum("toto 1diet titi"), "toto 1 diet titi")
def gensim_clean_string(textIn):
cleaner = strip_tags(textIn)
cleaner = split_alphanum(cleaner)
cleaner = strip_multiple_whitespaces(cleaner)
cleaner = strip_short(cleaner, minsize=3)
return cleaner
def transformText(text):
#pour séparer les numéros et les phrases
text = split_alphanum(text)
# Convert text to lower
text = text.lower()
text = replace_word(text)
text = replace_numbers(text)
text = reduce_lengthening(text)
text = remove_alphanumerics(text)
#stops = set(stopwords.words("english"))
stopword = {
'only', 'being', 'yourselves', 'out', 'nor', 'few', 'now', 'd',
'before', 'their', 'and', 'ourselves', 'doing', 'its', "you've",
'yourself', 'off', 'do', 'm', 'but', 'after', 'during', 'has', 'for',
'above', 'y', 'on', 've', 'itself', 'been', 'until', 'we', 'shan',
're', 'll', 'o', 'did', 'there', 'you', 'further', 'some', 'where',
'through', 'doesn', 'a', 'it', 'does', 'who', 'most', 'she', 'each',
'am', 'if', 'hadn', 'him', 'the', 'those', 'will', "you'll", 'any',
'why', 'weren', 'ours', 'that', 'other', 'own', 'once', 'was', 'from',
'this', 'at', 'such', 'than', 'between', 'because', 'while', 'when',
'so', 'himself', 'them', 'didn', 's', 'ma', 'shouldn', 'my', 'only',
'then', 'herself', 'under', 'theirs', 'about', 'here', 'his', 'were',
'be', 'with', "you'd", 'are', 'both', 'haven', 'of', 'over', 'won',
'below', 'down', 'ain', 'or', 'to', 'too', 'again', 'an', 'needn',
'what', 'aren', "should've", 'me', 'hers', "that'll", 'just', 'same',
'which', "you're", 'having', "it's", 'can', 'these', 'more', 'i', 'up',
'yours', 'your', 'themselves', 'by', 'don', 'wasn', 'how', 'mightn',
'as', 'all', 'mustn', 'into', 'in', 'is', 't', 'he', 'had', 'her',
'our', 'they', 'have', 'against', 'whom', 'isn', "she's", 'myself'
}
# Convert text to lower
text = text.lower()
# Removing non ASCII chars
text = re.sub(r'[^\x00-\x7f]', r' ', text)
# Strip multiple whitespaces
text = gensim.corpora.textcorpus.strip_multiple_whitespaces(text)
# Removing all the stopwords
filtered_words = [word for word in text.split() if word not in stopword]
#retourner les noms pluriels au singulier et les verbes à l'infinitif:limatisation
input_str = word_tokenize(text)
for i in range(len(input_str)):
input_str[i] = lemmatizer.lemmatize(input_str[i], pos='v')
input_str[i] = lemmatizer.lemmatize(input_str[i], pos='n')
#correction d'orthographe
spell = SpellChecker()
misspelled = text.split()
for i in range(len(misspelled)):
# Get the one `most likely` answer
word = spell.correction(misspelled[i])
misspelled[i] = word
text = " ".join(misspelled)
# Removing all the tokens with lesser than 3 characters
filtered_words = gensim.corpora.textcorpus.remove_short(filtered_words,
minsize=3)
# Preprocessed text after stop words removal
text = " ".join(filtered_words)
# Remove the punctuation
text = gensim.parsing.preprocessing.strip_punctuation2(text)
# Strip all the numerics
text = gensim.parsing.preprocessing.strip_numeric(text)
# Strip multiple whitespaces
text = gensim.corpora.textcorpus.strip_multiple_whitespaces(text)
# Stemming
return gensim.parsing.preprocessing.stem_text(text)
