Python tokenize.WordPunctTokenizer类代码示例

    def number_of_different_words(self):
        # TODO: Stemming, then move to language specific classes
        tokenizer = WordPunctTokenizer()
        words = tokenizer.tokenize(self.text.strip())
        only_textual_words = filter(unicode.isalpha, words)

        return len(set(only_textual_words))

def message_to_wordlist(message, lemmas_bool, remove_stopwords=False):
    # Function to convert a document to a sequence of words,
    # optionally removing stop words.  Returns a list of words.
    #
    # 1. Remove HTML
    #review_text = BeautifulSoup(review).get_text()
    #
    # 2. Remove messages numbers
    message_text = re.sub(">>\d+","", message)
    message_text = message_text.lower()
    message_text = re.sub(u"ё", 'e', message_text, re.UNICODE)
    message_text = clean_str(message_text)
    tokenizer = WordPunctTokenizer()
    # 3. Convert words to lower case and split them
    words = tokenizer.tokenize(message_text)
    lemmas = []
    # 4. Optionally remove stop words (false by default)
    if remove_stopwords:
        stops = set(stopwords.words("english"))
        words = [w for w in words if not w in stops]
    if lemmas_bool == 'l':
        for word in words:
            word_parsed = morph.parse(word)
            if len(word_parsed) > 0:
                lemmas.append(word_parsed[0].normal_form)
    elif lemmas_bool == 's':
        for word in words:
            word = stemmer.stem(word)
            if len(word) > 0:
                lemmas.append(word)
    else:
        lemmas = words
    # 5. Return a list of words
    return(lemmas)

def tfIdf():
	TFIDF_MIN_SCORE = 100
	import nltk
	from nltk.tokenize import WordPunctTokenizer
	tokenizer = WordPunctTokenizer()		
	collection = initialize_collection('documents')

	docs = collection.find()
	tfidf = []
	idfMap = create_idf_map()
	docs = collection.find()
	for d in docs:
		tfMap = {}
		for word in set(tokenizer.tokenize(d['content'].lower())):
		 	if word not in tfMap:
		 		tfMap[word] = 1
		 	else:
		 		tfMap[word] += 1
		tfIdfValues = []
		for word in set(tokenizer.tokenize(d['content'].lower())):
			if (tfMap[word] * 1000 / idfMap[word]) > TFIDF_MIN_SCORE:
				tfIdfValues.append((word, tfMap[word] * 1000 / idfMap[word]))
		tfIdfValues = sorted(tfIdfValues, key = lambda x : x[1], reverse = True)
		d['tfidf'] = tfIdfValues
		tfidf.append({'d' : d,
					  'tfidf' : tfIdfValues})
		collection.save(d)


	genFreq = generaral_frequency(idfMap)
	return render_template("tfidf.html", documents = tfidf)

def extract_nl_text(ms):
    """
    Extracts and tokenizes text from malware sample object

    :param ms: MalwareSample object
    :return: list of tokenized strings found in malware sample object's internal strings list
    """
    wpt = WordPunctTokenizer()
    all_tokenized_strings_in_ms = []
    inside_xml_privileges = False
    for s in ms.strings:
        if 'requestedPrivileges' in s or 'This program cannot be run in DOS mode' in s:
            continue
        elif inside_xml_privileges:
            continue
        elif '<assembly xmlns' in s:
            inside_xml_privileges = True
            continue
        elif '</assembly>' in s:
            inside_xml_privileges = False
            continue

        tokenized_string = []
        tokens = wpt.tokenize(s)
        if tokens:
            for t in tokens:
                if wordnet.synsets(t) and len(t) > 3:  # had to use length to eliminate false positives
                    tokenized_string.extend(tokens)
                    break
        if tokenized_string:
            all_tokenized_strings_in_ms.append(tokenized_string)
    return all_tokenized_strings_in_ms

    def tokens(self):
        """Tokenize the text.
        """
        tokenizer = WordPunctTokenizer()

        # Get token character spans.
        spans = list(tokenizer.span_tokenize(self.text))

        # Materialize the token stream.
        tokens = [self.text[c1:c2] for c1, c2 in spans]

        tags = pos_tag(tokens)

        return [

            Token(
                token=token.lower(),
                char1=c1,
                char2=c2,
                pos=pos,
            )

            for (c1, c2), token, (_, pos) in
            zip(spans, tokens, tags)

        ]

def tokenize_words(sentence):
    """
    :param sentence:
    :return: list of words in sentence
    """
    tokenizer = WordPunctTokenizer()
    return tokenizer.tokenize(sentence)

def class1():
	import nltk
	from nltk.tokenize import WordPunctTokenizer
	docId = request.args.get('d')
	tokenizer = WordPunctTokenizer()		
	collection = initialize_collection('documents')

	featuresets = []
	tagSet = set()
	for d in collection.find():	
		bagOfWords = bag_of_words(tokenizer.tokenize(d['content']))
		if 'tags' not in d: continue
		for tag in d['tags']:
			featuresets.append((bagOfWords, tag))
			tagSet.add(tag)
	classifier = nltk.NaiveBayesClassifier.train(featuresets)

	d = collection.find_one({'_id' : ObjectId(docId)})

	#classifier.show_most_informative_features(100)
	cl = classifier.prob_classify(bag_of_words(tokenizer.tokenize(d['content'])))
	probs = []
	for tag in tagSet:
		probs.append((tag, round(cl.prob(tag)*100) ))
	classifier.show_most_informative_features(n=20)
	probs = sorted(probs, key = lambda x : x[1],  reverse = True)
	return render_template('class1.html', probs = probs, d=d)

def tokenize(text):
    """Tokenize a raw text.

    Args:
        text (str)

    Returns: list of {token, char1, char2, pos}
    """
    tokenizer = WordPunctTokenizer()

    # Get token character spans.
    spans = list(tokenizer.span_tokenize(text))

    # Materialize the token stream.
    tokens = [text[c1:c2] for c1, c2 in spans]

    # Tag parts-of-speech.
    tags = pos_tag(tokens)

    return [

        dict(
            token=token.lower(),
            char1=c1,
            char2=c2,
            pos=pos,
        )

        for (c1, c2), token, (_, pos) in
        zip(spans, tokens, tags)

    ]

def TextProcessor(src, tgt, low=True, num=True):

    print "processing "+src
    if low==True:
        print "lowercasing.."
    if num==True:
        print "removing numeric.."

    srcfile = codecs.open(src,"r","utf-8")
    tgtfile = codecs.open(tgt,"w","utf-8")

    word_punct_tokenizer = WordPunctTokenizer()

    linecount=0
    for line in srcfile:
        linecount+=1
        line = word_punct_tokenizer.tokenize(line)
        if low==True:
            for i in range(0,len(line)):
                line[i] = line[i].lower()
        if num==True:
            for i in range(0,len(line)):
                if line[i].isnumeric()==True:
                    line[i] = "<number>"

        tgtfile.write(listtostring(line))

    srcfile.close()
    tgtfile.close()
    print "done processing "+str(linecount)+" lines!!"

def clean_data(input_file_name, output_file_name):
    def clean_word(word):
        word = word.encode('ascii', 'ignore')
        word = word.lower()
        word = re.sub(r'(\S)\1+', r'\1\1', word)  # normalize repeated characters to two
        word = re.sub(r'(\S\S)\1+', r'\1\1', word)

        if re.search(r'((([A-Za-z]{3,9}:(?:\/\/)?)(?:[-;:&=\+\$,\w][email protected])?[A-Za-z0-9.-]+|(?:www.|[-;:&=\+\$,\w][email protected])[A-Za-z0-9.-]+)((?:\/[\+~%\/.\w-]*)?\??(?:[-\+=&;%@.\w]*)#?(?:[\w]*))?)',word) is not None:
            word = 'GENERIC_HTTP'

        return word

    tokenizer = WordPunctTokenizer()
    data = []
    with open(input_file_name) as input_file:
        for sentences, label in json.load(input_file):
            cleaned_sentences = []
            for sentence in sentences:
                cleaned_sentence = " ".join(map(clean_word, sentence.split()))
                cleaned_sentence = tokenizer.tokenize(cleaned_sentence)
                cleaned_sentences.append(cleaned_sentence)

            data.append([cleaned_sentences, label])

    with codecs.open(output_file_name, 'w', encoding='utf-8') as output_file:
        json.dump(data, output_file)

def clean_data(input_file_name, output_file_name):
    def clean_word(word):
        word = word.lower()
        word = word.replace('&amp;','&').replace('&lt;','<').replace('&gt;','>').replace('&quot;','"').replace('&#39;',"'")
        word = re.sub(r'(\S)\1+', r'\1\1', word)  # normalize repeated characters to two
        word = re.sub(r'(\S\S)\1+', r'\1\1', word)

        word = word.encode('ascii', 'ignore')

        if re.search(r'((([A-Za-z]{3,9}:(?:\/\/)?)(?:[-;:&=\+\$,\w][email protected])?[A-Za-z0-9.-]+|(?:www.|[-;:&=\+\$,\w][email protected])[A-Za-z0-9.-]+)((?:\/[\+~%\/.\w-]*)?\??(?:[-\+=&;%@.\w]*)#?(?:[\w]*))?)',word) is not None:
            word = 'GENERIC_HTTP'

        return word.encode('ascii', 'ignore')

    tokenizer = WordPunctTokenizer()

    with gzip.open(input_file_name) as input_file:
        with gzip.open(output_file_name, 'w') as output_file:
            for line in input_file:
                sentences, score = json.loads(line)
                cleaned_sentences = []
                for sentence in sentences:
                    cleaned_sentence = " ".join(map(clean_word, sentence.split()))
                    cleaned_sentences.append(tokenizer.tokenize(cleaned_sentence))

                json.dump([cleaned_sentences, score], output_file)
                output_file.write("\n")

 def get_words_without_stopwords(self, text):
     stopwords = nltk.corpus.stopwords.words('english')
     stopwords.extend(string.punctuation)
     stopwords.append('')
     tokenizer = WordPunctTokenizer()
     tokens = [token.lower().strip(string.punctuation) for token in tokenizer.tokenize(text) \
               if token.lower().strip(string.punctuation) not in stopwords]
     return tokens

def extract_words(text):
    stemmer = PorterStemmer()

    tokenizer = WordPunctTokenizer()
    tokens = tokenizer.tokenize(text)

    result =  [stemmer.stem(x.lower()) for x in tokens if x not in stopwords.words('english') and len(x) > 1]
    return result

def get_tokens(sentence):
    """
    Tokenizes a list of sentences
    :param sentence: list of sentences
    :return: list of tokenized sentences
    """

    tokenizer = WordPunctTokenizer()
    return tokenizer.tokenize(sentence)

def getBigram(haystack):
    tokenizer = WordPunctTokenizer()
    words = tokenizer.tokenize(haystack)
    bcf = BigramCollocationFinder.from_words(words)
    stopset = set(stopwords.words('english'))
    filter_stops = lambda w: len(w) < 3 or w in stopset
    bcf.apply_word_filter(filter_stops)

    return bcf.nbest(BigramAssocMeasures.likelihood_ratio, 4)