def check_emojis(e1: EMOJI_CHECK,
e2: EMOJI_CHECK) -> tuple[bool, Optional[str]]:
"""Validate the emojis, the user put."""
if isinstance(e1, str) and emojis.count(e1) != 1:
return False, e1
if isinstance(e2, str) and emojis.count(e2) != 1:
return False, e2
return True, None
def check_emojis(e1: EMOJI_CHECK,
e2: EMOJI_CHECK) -> Tuple[bool, Optional[str]]:
""" Validates the emojis the user put.
:param e1: The first emoji.
:param e2: The second emoji. """
if isinstance(e1, str) and emojis.count(e1) != 1:
return False, e1
if isinstance(e2, str) and emojis.count(e2) != 1:
return False, e2
return True, None
def get_all_specific_features(Authors):
features = []
for author in Authors:
features_user = []
emoji = 0
first_person = 0
pronouns = 0
negations = 0
for tweet in author['tweets']:
current_preprocessed_tweet = clean_en.tokenize(tweet)
emoji = emoji + emojis.count(tweet)
'''emoji = emoji +
len(re.findall(u'[\U0001f600-\U0001f650]',
tweet))'''
first_person = first_person + \
current_preprocessed_tweet.count('<first_person>')
pronouns = pronouns + current_preprocessed_tweet.count('<pronoun>')
negations = negations + \
current_preprocessed_tweet.count('<negation>')
features_user.append(emoji)
features_user.append(first_person)
features_user.append(pronouns)
features_user.append(negations)
features.append(features_user)
return features
def emojiCount(file):
file = open(file, 'r', encoding='utf-8')
i = 0
for line in file:
if emojis.count(line) > 0:
i += 1
print(i)
def emoticon_number_sd(author):
emoticon_number_list = []
for tweet in author['tweets']:
# emoticon_number = len(re.findall(u'[\U0001f600-\U0001f650]', tweet))
emoticon_number_list.append(emojis.count(tweet))
sd = stats.stdev(emoticon_number_list)
return sd
def final_process(word):
if emojis.count(word) > 0:
return [word]
elif len(re.sub(r'\W+', '', word)) > 1:
return re.sub(r'\W+', ' ', word).split()
else:
return ''
def get_all_specific_features_label(Authors):
print("specific es")
features = []
i = 0
for author in Authors:
print(i)
i = i + 1
features_user = []
emoji = 0
first_person = 0
pronouns = 0
negations = 0
for tweet in author['tweets']:
current_preprocessed_tweet = clean_es.tokenize(tweet)
emoji = emoji + emojis.count(tweet)
first_person = first_person + \
current_preprocessed_tweet.count('<first_person>')
pronouns = pronouns + current_preprocessed_tweet.count('<pronoun>')
negations = negations + \
current_preprocessed_tweet.count('<negation>')
features_user.append(emoji)
# features_user.append(first_person)
# features_user.append(pronouns)
features_user.append(negations)
features.append(features_user)
return features
def count_unique_emojis(text: str, unique=True):
'''
https://emojis.readthedocs.io/en/latest/api.html#sample-code
import emojis
'''
emoj = emojis.count(text)
return emoj
def emoji_1(Authors):
for author in Authors:
emoji = 0
for tweet in author['tweets']:
emoji = emoji + emojis.count(tweet)
'''emoji = emoji +
len(re.findall(u'[\U0001f600-\U0001f650]',
tweet))'''
author['emoji'] = emoji
def emoticon_ratio(author):
tweet_having_emoticon = 0
emoticon_number = 0
for tweet in author['tweets']:
# emoticon_number = len(re.findall(u'[\U0001f600-\U0001f650]', tweet))
emoticon_number = emojis.count(tweet)
if emoticon_number > 0:
tweet_having_emoticon += 1
return tweet_having_emoticon/len(author['tweets'])
def emoji_2(Authors):
features = []
for author in Authors:
emoji = 0
for tweet in author['tweets']:
emoji = emoji + emojis.count(tweet)
'''emoji = emoji +
len(re.findall(u'[\U0001f600-\U0001f650]',
tweet))'''
features.append(emoji)
return features
def text(update, context):
text_received = update.message.text
# Hack to check if all graphemes are emojis
if emojis.count(text_received) == len(
list(grapheme.graphemes(text_received))):
print("emojis:", text_received)
update.message.reply_text(f'uwu {text_received}')
SSEFuckery.sse_broadcast("emojis", text_received)
return
print("scrolly-text:", text_received)
update.message.reply_text(f'auzi cica >{text_received}')
SSEFuckery.sse_broadcast("scrolly-text", text_received)
async def radio_mode_message(message):
"""Listens to messages and deletes them if they're not text-only in a radio channel"""
if message.author == OMEGA.user:
return
cursor.execute("SELECT EXISTS(SELECT 1 FROM radio WHERE channel_id=?);",
[message.channel.id])
if cursor.fetchall(
)[0][0] and (message.attachments or emojis.count(message.content) or any(
re.search(expression, message.content) for expression in
[
r"<:\w*:\d*>",
r"(([\w]+:)?//)?(([\d\w]|%[a-fA-f\d]{2})+(:([\d\w]|%[a-fA-f\d]{2})+)?@)"
r"?([\d\w][-\d\w]{0,253}[\d\w]\.)+[\w]{2,63}(:[\d]+)?"
r"(/([-+_~.\d\w]|%[a-fA-f\d]{2})*)"
r"*(\?(&?([-+_~.\d\w]|%[a-fA-f\d]{2})=?)*)?(#([-+_~.\d\w]|%[a-fA-f\d]{2})*)?",
])):
await message.delete()
def process_tweets(tweet):
'''
Takes in a single tweet (string of text, e.g. the 'tweet_text' key)
and returns a processed, tokenized version of the tweet (list)
'''
# both emoji encoders exclude different emojis, so figured why not use both for increased coverage
tweet = emoji.emojize(emojis.encode(tweet))
# remove @ tags (usernames)
tweet = re.sub('@[A-Za-z0-9_]+', '', tweet)
# remove commas from numbers (e.g. 1,000 -> 1000)
tweet = re.sub('([0-9],[0-9])', lambda x: str(x.group(0)).replace(',', ''),
tweet)
# remove strange unicode encodings and use NLTK TweetTokenizer to tokenize
tweet = [
word.encode('ascii', 'ignore').decode('ascii')
if emojis.count(word) == 0 else word for word in tt.tokenize(tweet)
]
# remove nonalphanumeric words except for emojis
tweet = [
final_process(word) for word in tweet if final_process(word) != ''
]
return [item for sublist in tweet for item in sublist]
def create_tweet_features(tweets_data, run_countvec=False, run_tfidf=False):
# clean_twitter_text should be run before this function for the best results
# Create feature for number of exclamation marks
tweets_data['exc_count'] = tweets_data['text'].map(lambda x: x.count("!"))
# Create a column for number of characters
tweets_data['characters_nb'] = tweets_data.text.apply(len)
# Add count for emojis
tweets_data['emoji_count'] = tweets_data['text'].map(
lambda x: emojis.count(x))
# Add text features
vocab_list = vocab.get_all_vocab()
if run_countvec:
tweets_data_vocab = apply_countvectorizer(tweets_data,
'text_dec',
vocab_def=vocab_list)
elif run_tfidf:
tweets_data_vocab = apply_tfidf(tweets_data,
'text_dec',
vocab_def=vocab_list)
return tweets_data_vocab
def count_emojis(text: str):
'''
import emojis
'''
emoj = emojis.count(text)
return emoj
id_list = []
for i in range(1):
if i == 0:
emoji_list = ":smile: :joy: :heart: :disappointed: :sob: :sparkles: :relieved: :wink: :cry: :angry:"
keywords = emojis.encode(emoji_list).split()
#keywords = ["the","be","to","of","and","a","in","that","have","I"]
else:
keywords = [
"good", "happy", "free", "great", "easy", "bad", "alone",
"upset", "sad", "lost"
]
#for tweet in tweets:
for tweet in Cursor(api.search, q=["trump"], lang="en").items(10):
if (emojis.count(tweet.text) != 0) and (tweet.id not in id_list):
#basic information of tweets
my_id = tweet.id
id_list.append(tweet.id)
date = tweet.created_at
author = tweet.author.screen_name
#many tweets don't have location information
if tweet.place != None:
location = tweet.place.name
elif tweet.author.location != None:
location = tweet.author.location
else:
location = tweet.user.location
likes = tweet.favorite_count
retweets = tweet.retweet_count
def has_permissions(user: discord.Member) -> bool:
return any(role in user.roles
for role in [Constants.PEPEHACK_ROLE, Constants.BARTENDER_ROLE])
# Выбирает из 2 вариантов male/female в зависимости от наличия соответственной роли в дискорде
def gender(user, male, female):
return female if Constants.FEMALE_ROLE in user.roles else male
# Выдаёт форматированный смайл с названием emote_name в формате строки для сообщения в дискорде
# Если смайла с данным именем нет на сервере, возвращаяется пустая строка
def emote(emote_name: str) -> str:
if (emote := discord.utils.get(Constants.GUILD.emojis, name=emote_name)):
return str(emote)
elif emojis.count(emojis.encode(f':{emote_name}:')) > 0:
return emojis.encode(f':{emote_name}:')
else:
return ''
# Возвращает id юзера типа int при заданном пинге юзера (отметка через @)
# В случае ошибки поднимает исключение ValueError
def get_id(user_mention: str) -> int:
user_mention = user_mention.replace("<", "")
user_mention = user_mention.replace(">", "")
user_mention = user_mention.replace("@", "")
user_mention = user_mention.replace("!", "")
user_mention = user_mention.replace("&", "")
return int(user_mention)
def is_emoji(s):
try:
return emojis.count(s)
except:
return 0
else:
location = tweet.user.location
likes = tweet.favorite_count
retweets = tweet.retweet_count
#use TextBlob to do sentiment analysis
clean_tweet = ' '.join(re.sub("(@[A-Za-z0-9]+)|([^0-9A-Za-z \t])|(\w+:\/\/\S+)", " ", tweet.text).split())
analysis = TextBlob(clean_tweet)
polarity = analysis.polarity
subjectivity = analysis.subjectivity
#use emojis to do emoji process
emoji_all = emojis.get(tweet.text)
emoji_see = emojis.decode(" ".join(emoji_all))
emoji_num = emojis.count(tweet.text)
#previous tweet text data, use unicode to encode, also can .decode("unicode_escape")
tweet = tweet.text.encode("unicode_escape")
s = pd.Series({'id':my_id,'date':date,'author':author,'location':location,'likes':likes,'retweets':retweets,'polarity':polarity,'subjectivity':subjectivity,'emoji_all':emoji_all,'emoji_see':emoji_see,'emoji_num':emoji_num,'tweet':tweet})
df = df.append(s, ignore_index=True)
print(df.shape[0])
print(i)
time.sleep(3)
df.to_excel('TweetData_emoji.xlsx', sheet_name='Sheet1') # Excel is better, beecause we can see the emoji
# 测试Emoji表情符号包Emojis的功能
import emojis
print(emojis.count('Python is fun 👍'))
# 读取搜狗输入法的表情示例，验证搜狗输入法的表情代码，的确为Unicode，如U+0001F601
from collections import defaultdict
file = open('data\sogouw.txt', 'w', encoding='utf-8-sig')
with open('data\sogou.txt', 'r', encoding='utf-8-sig') as f:
i = 0
for line in f:
line = line.replace(' ', '\n')
while i < len(line):
file.write(line[i])
i += 1
file.close()
frequencies = defaultdict(int)
with open('data\sogouw.txt', 'r', encoding='utf-8-sig') as f:
for line in f:
line = line.replace('\n', '')
a = line.encode('unicode-escape').decode('utf-8')
print(line, a)
a = a.replace('\\u', '').replace('\\U', '').replace('feff', '')
# 计算Sogou Unicode的十进制值以划定范围
try:
b = int(a, 16)
except:
print(a)
def EmojiCount(tweet):
tweet = emojis.encode(tweet)
return emojis.count(tweet, unique=True)
def emoticon_number_avg(author):
emoticon_number = 0
for tweet in author['tweets']:
# emoticon_number = len(re.findall(u'[\U0001f600-\U0001f650]', tweet))
emoticon_number += emojis.count(tweet)
return emoticon_number/len(author['tweets'])
