def pull_tweet(input_file_name):
CONSUMER_KEY = "9At2u3Y2DraTHLSg3D9w6LhE9"
CONSUMER_KEY_SECRET = "DRFCbI2t0gMhfV2KnEub6cljowW9zRwmkeMJ0GT9MlMkrkzspM"
ACCESS_TOKEN = "1259913765614751745-LwtSI48si3sYekzvxW86syIFsRgirl"
ACCESS_TOKEN_SECRET = "e0gpJdT0IXOSxFrhplKMl8FlP0dVnuLg1vwBHzt5Fc9J9"
t = Twarc(CONSUMER_KEY, CONSUMER_KEY_SECRET, ACCESS_TOKEN,
ACCESS_TOKEN_SECRET)
inputF = open(input_file_name, "r")
line = inputF.readline()
data = []
i = 0
while line != "" and i < 10:
try:
tweet = t.tweet(line.strip())
if tweet["lang"] == "en":
if 'retweeted_status' in tweet.keys():
data.append(tweet['retweeted_status']['full_text'].replace(
'\n', ' '))
else:
data.append(data, tweet['full_text'].replace('\n', ' '))
i += 1
line = inputF.readline()
except Exception as e:
line = inputF.readline()
return data
def readIdFile(input_file_name, output_file_name):
try:
t = Twarc(CONSUMER_KEY, CONSUMER_KEY_SECRET, ACCESS_TOKEN,
ACCESS_TOKEN_SECRET)
output_file = open(output_file_name, "w")
inputF = open(input_file_name, "r")
line = inputF.readline()
i = 1
while line != "":
tweet = t.tweet(line.strip())
if tweet["lang"] == "en":
output_file.write(json.dumps(tweet) + "\n")
print(i)
i += 1
line = inputF.readline()
except Exception as e:
print(e)
return 1
def pull_tweet(input_file_name):
"""
Args:
input_file_name: A txt file containing tweet IDs
Returns: the corresponding tweet text
"""
# Required Keys
CONSUMER_KEY = "9At2u3Y2DraTHLSg3D9w6LhE9"
CONSUMER_KEY_SECRET = "DRFCbI2t0gMhfV2KnEub6cljowW9zRwmkeMJ0GT9MlMkrkzspM"
ACCESS_TOKEN = "1259913765614751745-LwtSI48si3sYekzvxW86syIFsRgirl"
ACCESS_TOKEN_SECRET = "e0gpJdT0IXOSxFrhplKMl8FlP0dVnuLg1vwBHzt5Fc9J9"
# Initializing twarc module
t = Twarc(CONSUMER_KEY, CONSUMER_KEY_SECRET, ACCESS_TOKEN,
ACCESS_TOKEN_SECRET)
inputF = open(input_file_name, "r")
line = inputF.readline()
data = []
i = 0
# Read each line of the input file
while line != "" and i < 10:
try:
tweet = t.tweet(line.strip())
if tweet["lang"] == "en":
if 'retweeted_status' in tweet.keys():
data.append(tweet['retweeted_status']['full_text'].replace(
'\n', ' '))
else:
data.append(data, tweet['full_text'].replace('\n', ' '))
i += 1
if i % 10 == 0:
print('Pulled {} tweets'.format(i))
line = inputF.readline()
# Skip line if the tweet no longer exists
except Exception as e:
line = inputF.readline()
return data
def f(file):
print file
today = datetime.date.today()
margin = datetime.timedelta(days=7)
analyzer = SentimentIntensityAnalyzer()
acess_changer = [c8, c9, c10, c7, c1, c2, c3, c4]
# base_dir='/home/stealthuser/Perosnal/Sentimental/12 august/Data/Hamid output csvs/'
base_dir = c_p.input_csv_path
if os.path.exists(base_dir + 'reply/'):
output_dir = base_dir + 'reply/'
pass
else:
os.mkdir(base_dir + 'reply/')
output_dir = base_dir + 'reply/'
data = pd.read_csv(base_dir + file,
low_memory=False,
error_bad_lines=False)
data['reply'] = ''
reply = []
acess_changer_counter = 0
max_tweet = 35
config_key = 0
for index, tweet in enumerate(data['text']):
print file + " " + str(index)
t_id = data.ix[index, 'permalink'].split('/')[-1:][0]
tweet_date = data.ix[index, 'date'].split()[0]
if (acess_changer_counter % max_tweet == 0):
access_point = acess_changer[config_key % len(acess_changer)]
config_key += 1
t = Twarc(access_point.consumer_key, access_point.consumer_secret,
access_point.access_token,
access_point.access_token_secret)
acess_changer_counter += 1
print access_point
tweet_r = t.tweet(t_id)
if (len(tweet_r) > 0):
reply_tweets = []
req_format_date = tuple(
map(lambda x: int(x), tweet_date.split('-')))
if (today - margin <=
datetime.date(req_format_date[0], req_format_date[1],
req_format_date[2])):
for reply_tweet in t.replies(tweet_r):
reply.append(reply_tweet['text'].encode('utf-8'))
# print reply_tweet['text'].encode('utf-8')
reply_tweets.append(reply_tweet['text'].encode('utf-8'))
# delimiter for replies is '==<>=='
data.ix[index, 'reply'] = '==<>=='.join(reply_tweets)
df_reply_comment = pd.DataFrame({'replies': reply})
for index_comment, reply in enumerate(df_reply_comment['replies']):
score = analyzer.polarity_scores(str(reply))
df_reply_comment.ix[index_comment, 'positive'] = score['pos']
df_reply_comment.ix[index_comment, 'negative'] = score['neg']
df_reply_comment.ix[index_comment, 'neutral'] = score['neu']
df_reply_comment.ix[index_comment, 'compound'] = score['compound']
# df_reply_comment.to_csv(output_dir+file[:-4]+'_replies_sentiment.csv')
data.to_csv(output_dir + file[:-4] + '_comments.csv')
status, paths_reply_label = classification_tweet(df_reply_comment,
output_dir, file[:-4])
# Check that tweet_id was provided
if len(sys.argv) != 2:
raise RuntimeError(
"Program should be called like: `python main.py <tweet_id>`")
tweet_id = sys.argv[1]
# Main
t = Twarc(
consumer_key=consumer_key,
consumer_secret=consumer_secret,
access_token=access_token,
access_token_secret=access_token_secret,
)
tweet = t.tweet(tweet_id)
if not tweet:
raise RuntimeError(f"tweet with id {tweet_id} does not exist")
# replies is a generator object
replies = t.replies(tweet, True)
# List to hold dict of relevant photo data from each of the replies
photo_data = []
for reply in replies:
# Photos will be in a list stored at reply['extended_entities']['media']
print("Processing next reply")
ee = reply.get("extended_entities")
if ee is None:
continue