def execute(self):
numberOfPages = 1
numberOfTweetsPerPage = 200
counter = 0
cloud = ""
txt = ""
if self.numberOfTweets > 200:
numberOfPages = ceil(self.numberOfTweets / 200)
else:
numberOfTweetsPerPage = self.numberOfTweets
for i in range(numberOfPages):
tweets = self.api.user_timeline(screen_name=self.username,
count=numberOfTweetsPerPage,
page=i)
for each in tweets:
cloud = each.text
cloud = re.sub(r'[A-Za-z@_]*', '', cloud)
counter += 1
txt = txt + ' ' + each.text
txt = re.sub(r'[A-Za-z@]*', '', txt)
twitter_mask = np.array(
Image.open(path.join(self.d, "templates/cloud/twitter-logo.jpg")))
stop = [
'می', 'من', 'که', 'به', 'رو', 'از', 'ولی', 'با', 'یه', 'این',
'نمی', 'هم', 'شد', 'ها', 'اما', 'تو', 'واقعا', 'در', 'نه', 'دارم',
'باید', 'آره', 'برای', 'تا', 'چه', 'کنم', 'بود', 'همه', 'دیگه',
'ای', 'اون', 'تی', 'حالا', 'بی', 'د', 'چرا', 'بابا', 'منم', 'کیه',
'توی', 'نیست', 'چی', 'باشه', 'که', 'بودم', 'می کنم', 'که', 'اینه',
'بهتر', 'داره', 'اینه', 'که', 'کردن', 'می', 'کن', 'بعد', 'دیگه',
'', '', '', ''
]
wc = WordCloudFa(
# font_path='IranNastaliq.ttf',
persian_normalize=True,
max_words=1000,
margin=0,
width=3000,
height=2500,
min_font_size=1,
max_font_size=1000,
background_color=self.backGround,
mask=twitter_mask,
include_numbers=False,
collocations=False)
wc.add_stop_words(stop)
wc.generate(txt)
directory = 'static/images/' + self.username + '.png'
directory = path.join(self.d, directory)
image = wc.to_image()
image.save(directory)
def save_word_cloud(user_name: str, api):
raw_tweets = []
for tweet in tweepy.Cursor(api.user_timeline, id=user_name).items():
raw_tweets.append(tweet.text)
# Normalize words
tokenizer = WordTokenizer()
lemmatizer = Lemmatizer()
normalizer = Normalizer()
stopwords = set(
list(
map(lambda w: w.strip(),
codecs.open(STOPWORDS_PATH, encoding='utf8'))))
words = []
for raw_tweet in raw_tweets:
raw_tweet = re.sub(r"[,.;:?!،()]+", " ", raw_tweet)
raw_tweet = re.sub('[^\u0600-\u06FF]+', " ", raw_tweet)
raw_tweet = re.sub(r'[\u200c\s]*\s[\s\u200c]*', " ", raw_tweet)
raw_tweet = re.sub(r'[\u200c]+', " ", raw_tweet)
raw_tweet = re.sub(r'[\n]+', " ", raw_tweet)
raw_tweet = re.sub(r'[\t]+', " ", raw_tweet)
raw_tweet = normalizer.normalize(raw_tweet)
raw_tweet = normalizer.character_refinement(raw_tweet)
tweet_words = tokenizer.tokenize(raw_tweet)
tweet_words = [
lemmatizer.lemmatize(tweet_word).split('#', 1)[0]
for tweet_word in tweet_words
]
tweet_words = list(filter(lambda x: x not in stopwords, tweet_words))
words.extend(tweet_words)
if len(words) == 0:
return
# Build word_cloud
mask = np.array(Image.open(MASK_PATH))
clean_string = ' '.join([str(elem) for elem in words])
clean_string = arabic_reshaper.reshape(clean_string)
clean_string = get_display(clean_string)
word_cloud = WordCloudFa(persian_normalize=False,
mask=mask,
colormap=random.sample(cmaps, 1)[0],
background_color=BACKGROUND_COLOR,
include_numbers=False,
font_path=FONT_PATH,
no_reshape=True,
max_words=1000,
min_font_size=2)
wc = word_cloud.generate(clean_string)
image = wc.to_image()
image.save(word_cloud_address)
def show_chat_word_cloud(directory):
with codecs.open(os.path.join(directory, 'chats.txt'),
'r',
encoding='utf8') as file:
print("Start putting words in picture")
mask_array = np.array(Image.open("telegram.png"))
wordcloud = WordCloudFa(persian_normalize=True,
mask=mask_array,
collocations=False)
stop_words = []
with codecs.open("stop_words.txt", 'r', encoding='utf8') as words:
for word in words:
stop_words.append(word[:-2])
wordcloud.add_stop_words(stop_words)
text = delete_extra_characters(file.read())
wc = wordcloud.generate(text)
image = wc.to_image()
image.show()
image.save(os.path.join(directory, 'wordcloud.png'))
def main():
database = DB("ganjoor.s3db")
database.connect()
rows = database.select("""select p.cat_id,v.text
from poem as p join verse as v on p.id=v.poem_id
where p.cat_id=24
""")
f = open("verses.txt", "w")
for r in rows:
f.write(r[1])
f.write('\n')
f.close()
wc = WordCloudFa(width=1200, height=800)
with open('verses.txt', 'r') as file:
text = file.read()
word_cloud = wc.generate(text)
image = word_cloud.to_image()
image.show()
image.save('hafez.png')
from wordcloud_fa import WordCloudFa
import numpy as np
from PIL import Image
mask = np.array(Image.open("mask.png"))
# Passing `no_reshape` parameter for you may cause problem in showing Farsi texts. If your output from the example
# is not true, you can remove that parameter
wodcloud = WordCloudFa(persian_normalize=True, include_numbers=False, background_color="white", mask=mask, no_reshape=True)
# Adding extra stop words:
wodcloud.add_stop_words(['the', 'and', 'with', 'by', 'in', 'to', 'to the', 'of', 'it', 'is', 'th', 'its', 'for', '[ ]', '. [', '] ['])
text = ""
with open('mixed-example.txt', 'r') as file:
text = file.read()
wc = wodcloud.generate(text)
image = wc.to_image()
image.show()
image.save('masked-example.png')
exit()
text = ""
print("cleaning")
text = " ".join([clean_word(word) for word in raw_text.split()])
#################################
print_stats(text)
print("generating cloud")
mask_array = np_array(Image.open(MASK))
wc_instance = WordCloudFa(
width=900,
height=900,
background_color=BG_COLOR,
font_path=FONT,
mask=mask_array,
persian_normalize=True,
include_numbers=False,
stopwords=load_stop_words(),
)
word_cloud = wc_instance.generate(text)
result_image = word_cloud.to_image()
result_image.save(RESULT_FILE_ADD)
result_image.show()
#print( "\n----\n".join(tweets_simple) )
to_print = "\n\n".join(tweets_simple)
f = open("out/cleaned.txt", "w")
f.write(to_print)
f.close()
#######################################
mask_array = np.array(Image.open("masks/tw.png"))
with open('out/cleaned.txt', 'r') as file:
text = file.read()
wc = WordCloudFa(
width=900,
height=900,
background_color="white",
font_path="fonts/font2.ttf",
mask=mask_array,
persian_normalize=True,
include_numbers=False,
)
word_cloud = wc.generate(text)
image = word_cloud.to_image()
image.save(f"out/{idish}.png")
image.show()
text = get_tweets_from_user(
username) # to get tweets of a specific user by its username
break
else:
print("you should enter f or u!")
text = get_tweets(text)
text = remove_bad_tweets(text)
text = "\n".join(text)
text = get_words(text)
print(len(text))
text = remove_bad_words(text)
print(len(text))
text1 = "\n".join(text)
text1 = removeWeirdChars(text1)
mask_array = np.array(Image.open(mask_path))
my_wc = WordCloudFa(width=1200,
height=1200,
background_color=background_color,
mask=mask_array,
persian_normalize=True,
repeat=False,
collocations=True)
my_wc.add_stop_words_from_file("../stop_words_kian.txt")
open("edited_tweets.txt", "w").write(text1)
my_wc.generate(text1)
image = my_wc.to_image()
image.show()
filename = datetime.now().strftime("%Y-%m-%d-%H-%M")
image.save('Images/{time}_photo.png'.format(time=filename))