def getEmployeeName(self, num):
''' This method get a employee name
@param num: int - is a current employee
'''
objEmployee = self.objJSON("Items")[num]
return "{0} {1}".format(uni(objEmployee[u"FirstName"]),
uni(objEmployee[u"LastName"]))
def parse_figma(json, frames=False, selection=''):
# Succeed placed here due to startup time of fzf
spinner.succeed()
for canvas in json['document']['children']:
canvasString = "{} -- {}".format(uni(canvas['name']),
uni(canvas['id']))
canvasString = canvasString.replace('\n', '')
yield canvasString
for canvas in json['document']['children']:
for frame in canvas['children']:
frameString = "{}.{} -- {}".format(uni(canvas['name']),
uni(frame['name']),
uni(frame['id']))
frameString = frameString.replace('\n', '')
yield frameString
def say(self):
words = []
dbase = self.learn()
put = uni(raw_input('> ').decode('utf-8')).upper()
put = re.sub('[.,?!]', '', put)
if put in self.questioned:
print 'Creuza: de novo?'
return
self.questioned.append(put)
if len(put.split()) < 2:
if put.find('OI') != -1 or put.find('OLA') != -1:
print 'Creuza: Oi, tudo bem?'
if put.find('PRAZER') != -1:
print 'Creuza: o prazer é todo seu..'
if put.find('PORQUE') != -1:
print 'Creuza: não enche..'
return
for question in dbase:
words = []
q = question[0].upper()
q = re.sub('[.,?!]', ' ', q)
answer = question[1]
# text = re.split('\W+', q)
text = q.split()
for w in text:
# if len(w) > 0:
words.append(w)
confirm = set(put.split()) & set(words)
# print put.split()
# if len(confirm):
# print confirm
if len(confirm) == len(put.split()):
print 'Creuza:', choice(answer)
return
def unidecode(string) -> str:
"""Attempt to convert a Unicode object into an ASCII string.
Args:
string (str): The string to be decoded.
:Design:
This function is a light wrapper around the
``unidecode.unidecode`` function.
**Per the** ``unicode`` **docstring:**
"Transliterate an Unicode object into an ASCII string.
>>> unidecode(u"北亰")
"Bei Jing "
This function first tries to convert the string using ASCII
codec. If it fails (because of non-ASCII characters), it falls
back to transliteration using the character tables.
This is approx. five times faster if the string only contains
ASCII characters, but slightly slower than using unidecode
directly if non-ASCII chars are present."
:Dependencies:
* unidecode
:Example:
::
import utils3.utils as u
s = u.unidecode(string)
Returns:
If the passed ``string`` value is a str data type, the decoded
string is returned, otherwise the original value is returned.
"""
# EXTERNAL IMPORTS
# RENAME unidecode TO AVOID CONFLICT WITH utils.unidecode FUNCTION
from unidecode import unidecode as uni
# INITIALISE VARIABLE
decoded = None
# TEST PASSED VALUE AS BEING A STRING >> STORE DECODED (OR ORIGINAL) VALUE
decoded = uni(string) if isinstance(string, str) else string
# RETURN VALUE
return decoded
def get_tokens(inp):
tokens = []
inp_target = inp[target_key]
if target == 'title':
inp_target = [inp_target]
for sent in inp_target:
# convert all non-ascii to nearest ascii
sent = uni(sent).lower()
for token in word_tokenize(sent):
# if token is the beginning or end of a quotation
# drop the quotation. however, 's is OK.
if token != "'s":
if len(token) > 1 and token[0] == "'":
token = token[1:]
if len(token) > 1 and token[-1] == "'":
token = token[:-1]
tokens.append(token)
return tokens
def say(self):
words = []
dbase = self.learn()
put = uni(raw_input('> ').decode('utf-8')).upper()
put = re.sub('[.,?!]', '', put)
# if put in self.questioned:
# print 'Creuza: de novo?'
# return
self.questioned.append(put)
if len(put.split()) < 2:
if put.find('#ADD') != -1:
self.add()
print
if put.find('KKK') != -1 or put.find('HAHA') != -1:
print 'Creuza: ha ha ha'
if put.find('OI') != -1 or put.find('OLA') != -1:
print 'Creuza: Oi, tudo bem?'
if put.find('OBRIGAD') != -1:
print 'Creuza: por nada :)'
if put.find('PRAZER') != -1:
print 'Creuza: o prazer é todo seu..'
if put.find('TCHAU') != -1:
print 'Creuza: vaza!!'
if put.find('PORQUE') != -1:
print 'Creuza: não enche..'
return
for question in dbase:
words = []
q = question[0].upper()
q = re.sub('[.,?!]', ' ', q)
answer = question[1]
# text = re.split('\W+', q)
text = q.split()
for w in text:
words.append(w)
confirm = set(put.split()) & set(words)
if len(confirm) == len(put.split()):
time.sleep(1)
print 'Creuza:', choice(answer).rstrip()
return
with open('index.txt', 'r') as f: # Ler o sumário
index = f.read()
with open('all_text.txt', 'r') as f: # Ler as tabelas em texto
all_text = f.read()
# Regex que identifica o título de todas as tabelas, sendo do tipo 'Registro' ou não
TABLE_TITLE_REGEX = r'\n((Registros do \n?evento )?S\n-\n(\d+) \n-\n \n([\w,. /]+\n))'
titles = re.findall(TABLE_TITLE_REGEX, all_text, re.MULTILINE|re.UNICODE)
# As tabelas são o conteúdo entre dois títulos consecutivos
tables = re.sub(TABLE_TITLE_REGEX, '$', all_text).split('$')[1:]
# Dicionário que contém como chave o número da tabela, e como valor o nome da tabela
# e seus dados, como uma list(list(string))
table_numbers = {
re.sub(r'\n', '', number): {
'name': re.sub(r'[\-/\.,\s]+', '_', uni(desc[:-3].lower().replace('\n', '')))
}
for number, desc in re.findall(
# Procura no índice o número e nome da tabela
r' \nS\n-\n(\d+(?:\n\d+)?) \n-\n \n((?:(?:\w[\w,\. /]+| |-)\n)+)\.{2,}',
index, re.IGNORECASE)
}
# Um grande problema é que o único separador presente nas tabelas é a quebra de linha ('\n \n').
# Entretanto, em muitos textos do campo 'Condição' há quebras de linha, o que impede diferenciar
# se uma quebra representa uma nova célula da tabela ou uma quebra de linha dentro da mesma célula.
# Isso será explicado adiante.
for title, table in list(zip(titles, tables))[1::2]:
# Transforma o campo 'Ocorr.' para uma única linha
tables_cells = re.sub(r'(\d+)\n-\n(\d+)', r'\g<1>-\g<2>', table)