def main():
results = []
sents = open('../d2_data/all_sents_spellchecked.txt',
'r').read().split('\n')
caller = pipeline_caller.PipelineCaller()
tool_name = "morphanalyzer"
api_token = "sQj6zxcVt7JzWXHNTdRu3QRzc6i8KZz7"
start_i = 346
i = 0
for s in sents[start_i:]:
try:
print(s)
curr_result = []
for w in s.split():
r = caller.call(tool_name, w, api_token)
r = ' '.join(r.split('\n'))
curr_result.append(r)
curr_result = '\n'.join(curr_result)
results.append(
'<S> <S>+BSTag\n{0}\n</S> </S>+ESTag'.format(curr_result))
i += 1
print(curr_result)
except ConnectionResetError:
True
if i % 1000 == 0 or i == len(sents[start_i:]):
with open('parsed_sents_{0}.txt'.format(i), 'w') as f:
f.write('\n'.join(results))
def main():
caller = pipeline_caller.PipelineCaller()
tool_name = "spellcheck"
api_token = "sQj6zxcVt7JzWXHNTdRu3QRzc6i8KZz7"
result = ''
text = open('../d2_data/all_verbs.txt', 'r').read()
result += caller.call(tool_name, text, api_token)
with open('../d2_data/all_verbs_spellchecked.txt', 'w') as f:
f.write(result)
def module_pipelineNoisy_whole_test(self):
try:
caller = pipeline_caller.PipelineCaller(
'pipelineNoisy', KATANA, os.environ['pipeline_token'], 'whole')
r = re.compile(r'(\d+)(\t.+?){7,}', re.MULTILINE)
response = caller.call()
print(response)
assert len(re.findall(r, response)) == 33
except:
self.fail('Exception thrown')
def module_Vowelizer_word_test(self):
try:
caller = pipeline_caller.PipelineCaller(
'Vowelizer', KELIME, os.environ['pipeline_token'], 'word')
r = re.compile(r'(.+?\n)', re.MULTILINE)
response = caller.call()
print(response)
assert len(re.findall(r, response)) == 4
except:
self.fail('Exception thrown')
def module_pipelineNoisy_sentence_test(self):
try:
caller = pipeline_caller.PipelineCaller(
'pipelineNoisy', UCDORT, os.environ['pipeline_token'],
'sentence')
r1 = re.compile(r'(1)(\t.+?){7,}', re.MULTILINE)
r2 = re.compile(r'(5)(\t.+?){7,}', re.MULTILINE)
response = caller.call()
print(response)
assert len(re.findall(r1, response)) == 2 and len(
re.findall(r2, response)) == 1
except:
self.fail('Exception thrown')
def main(argv):
table_name = ''
token = 'LQiWv0FTmQEJRVbun8Rqld6WZCIrGUyO'
tool = 'normalize'
try:
opts, args = getopt.getopt(argv, "ut:", ["table="])
except getopt.GetoptError:
print('normalization.py -t <table_name>')
sys.exit(2)
for opt, arg in opts:
if opt == '-u':
print('normalization.py -t <table_name>')
sys.exit()
elif opt in ("-t", "--table"):
table_name = arg
db = PathyDB()
table_name=table_name.lower()
table_name=table_name.replace('ç','c')
table_name=table_name.replace('Ç','c')
table_name=table_name.replace('ü','u')
table_name=table_name.replace('Ü','u')
table_name=table_name.replace('Ö','o')
table_name=table_name.replace('Ğ','g')
table_name=table_name.replace('ü','u')
table_name=table_name.replace('ğ','g')
table_name=table_name.replace('Ş','s')
table_name=table_name.replace('ş','s')
table_name=table_name.replace('ı','i')
table_name=table_name.replace('İ','i')
tweets = db.get_all_not_normalized_tweets(table_name)
for tweet in tweets:
caller = pipeline_caller.PipelineCaller(tool, unidecode(tweet['tweet_text']), token)
normalized_text = caller.call()
print(normalized_text)
tweet['normalized_text'] = normalized_text
db.update_tweet(table_name, tweet)
def main():
caller = pipeline_caller.PipelineCaller()
tool_name = "spellcheck"
api_token = "sQj6zxcVt7JzWXHNTdRu3QRzc6i8KZz7"
result = ''
data = open('../d2_data/query_results_all_joined_sents.csv')
reader = csv.reader(data)
sents = []
indices =[]
for r in reader:
sents.append(r[0])
indices.append(r[2])
with open('../d2_data/target_indices.txt', 'w') as f:
f.write('\n'.join(indices))
for i in range(0, 8):
text = '\n'.join(sents[i*10000:(i+1)*10000])
result += caller.call(tool_name, text, api_token)
with open('../d2_data/all_sents_spellchecked.txt', 'w') as f:
f.write(result)
"""
Send list of word windows to ITU for spellchecking
Heikal Badrulhisham <*****@*****.**>, 2019
"""
import pipeline_caller
import csv
from collections import defaultdict
import pickle
import os
# For calling ITU pipeline
caller = pipeline_caller.PipelineCaller()
tool_name = "spellcheck"
api_token = "sQj6zxcVt7JzWXHNTdRu3QRzc6i8KZz7"
# Dictionary of past spellcheck results
if os.path.isfile('spellcheck_history.pkl'):
spellcheck_history = pickle.load(open('spellcheck_history.pkl', 'rb'))
else:
spellcheck_history = defaultdict(str)
def spellcheck(word):
if word in spellcheck_history:
return spellcheck_history[word]
else:
sc = caller.call(tool_name, word, api_token).replace('\r\n', '')
spellcheck_history[word] = sc
return sc
def nlpPipeline(text):
caller = pipeline_caller.PipelineCaller('normalize', text, '')
result = caller.call(
) # call function takes tool name, text, and API access token
return result
def module_exception_test(self):
try:
caller = pipeline_caller.PipelineCaller()
caller.call("pipelineNoisy", "test sentence", "random token")
except:
self.fail("Exception thrown")
def nlpPipeline(text):
text = text.encode('utf8')
caller = pipeline_caller.PipelineCaller(
'normalize', text, 'MKHVuqqLiARKHNFq7eEOuOJr54Mncxir', 'whole')
result = caller.call()
return result
