async def resolve(self, path, movie):
if movie.title:
return movie
words = path.split()
for i, word in enumerate(words):
try:
if w2n.word_to_num(word):
words[i] = str(w2n.word_to_num(word))
except IndexError:
pass
path = ' '.join(words)
return await super().resolve(path, movie)
def test_output(self):
self.assertEqual(w2n.word_to_num("two million three thousand nine hundred and eighty four"),2003984)
self.assertEqual(w2n.word_to_num("nineteen"),19)
self.assertEqual(w2n.word_to_num('three billion'),3000000000)
self.assertEqual(w2n.word_to_num('three million'),3000000)
self.assertEqual(w2n.word_to_num('one hundred twenty three million four hundred fifty six thousand seven hundred and eighty nine')
,123456789)
self.assertEqual(w2n.word_to_num('eleven'),11)
self.assertEqual(w2n.word_to_num('nineteen billion and nineteen'),19000000019)
self.assertEqual(w2n.word_to_num('one hundred and forty two'),142)
self.assertEqual(w2n.word_to_num('one hundred thirty-five'),135)
def get_integer(str):
try:
l = int(str)
return l
except:
l = w2n.word_to_num(str)
return l
def noms_from_oscars(oscarsstring):
"""Converts descriptive oscars text to number of nominations as int."""
try:
nominations_str = (str(oscarsstring.split(",")[0]).strip().lower()).split(" ")
nominations = w2n.word_to_num(nominations_str[2])
return nominations
except:
return 0
def wins_from_oscars(oscarsstring):
"""Converts descriptive oscars text to number of wins as int."""
try:
wins_str = (str(oscarsstring.split(",")[1]).replace(".", "").strip().lower()).split(" ")
wins = w2n.word_to_num(wins_str[1])
return wins
except:
return 0
def get_query(self, value):
qset = Q()
for word in value.split():
qset &= Q(name__icontains=word)
try:
qset |= Q(name__icontains=num2words(word))
except TypeError:
word_converted_to_num = word_to_num(word)
if type(word_converted_to_num) == int:
qset |= Q(name__icontains=word_converted_to_num)
return qset
def clean_text(text):
try:
cleaned_text = contractions.fix(text.strip()).lower()
except IndexError:
cleaned_text = text.strip().lower()
words = [x for x in cleaned_text.split(' ') if x]
current_number = []
i = 0
while i < len(words):
try:
try:
float(words[i])
current_number.append(words[i])
words.pop(i)
except ValueError:
pass
if i < len(words):
try:
word_to_num(words[i])
except IndexError:
raise ValueError
current_number.append(words[i])
words.pop(i)
except ValueError:
if current_number:
j = 0
prod = 1
while j < len(current_number):
try:
prod *= float(current_number[j])
current_number.pop(j)
except ValueError:
j += 1
if current_number:
try:
num = prod * word_to_num(' '.join(current_number))
except:
num = prod
else:
num = prod
words.insert(
i,
str(
int(num)
if type(num) == float and num.is_integer() else num))
current_number.clear()
i += 1
if len(words[i]) <= 2 or words[i] in get_stop_words('en'):
words.pop(i)
continue
i += 1
else:
if current_number:
try:
words.append(str(word_to_num(' '.join(current_number))))
except ValueError:
words.append(' '.join(current_number))
return ''.join(x for x in ' '.join(words) if x.isalnum() or x == ' ')
def tokenize_text(self,
remove_accentedChar=True,
remove_httpLinks=True,
expand_contractionMap=True,
handle_emoji=True,
convert_word2Number=True,
remove_whiteSpace=True,
lemmatization=True,
lowercase=True,
remove_punctuations=True,
remove_number=True,
remove_specialChars=True,
expand_slang=True,
remove_stopWords=True,
correct_spelling_byWord=True,
correct_spelling=False):
# pre cleaning
self.text = self.define_preCleaning()
# sentence level preprocess
if remove_httpLinks:
self.text = self.remove_httpLinks()
if remove_whiteSpace:
self.text = self.remove_whiteSpace()
if remove_accentedChar:
self.text = self.remove_accentedNotation()
if expand_contractionMap:
self.text = self.expand_contractionMap()
if lowercase:
self.text = self.text.lower()
if correct_spelling:
self.text = self.correct_spelling()
if handle_emoji:
self.text = self.handle_emotion()
if expand_slang:
self.text = self.expand_slang()
doc = self.nlp(self.text)
cleanText = []
errorLog = []
# init a punctDict
punctUsageStatsDict = {k: 0 for k in self.notDeletePunctList}
# word-level preprocess
for token in doc:
flag = True
editFlag = False
toEdit = token.text
edit = ""
# remove stop words
try:
if remove_stopWords and token.is_stop and token.pos_ != "NUM":
flag = False
# remove punctuations
if remove_punctuations and token.pos_ == "PUNCT" and (
toEdit not in self.notDeletePunctList) and flag:
flag = False
# remove special characters
if remove_specialChars and token.pos_ in [
"SYM", "ADP", "X", "AUX"
] and flag:
flag = False
# remove numbers
if remove_number and (token.pos_ == "NUM"
or toEdit.isnumeric()) and flag:
flag = False
# convert word to numbers
if convert_word2Number and token.pos_ == "NUM" and flag:
edit = w2n.word_to_num(toEdit)
# lemmatization
if lemmatization and token.is_alpha and token.lemma_ != "-PRON-" and flag:
edit = token.lemma_
editFlag = True
# correct spelling by word
if correct_spelling_byWord and token.is_alpha and token.lemma_ != "-PRON-" and flag:
edit = self.correct_spelling_oneWord(token.lemma_)
editFlag = True
# constraint repeatition times of allowed punctuations
if token.pos_ == "PUNCT" and toEdit in self.notDeletePunctList and flag:
punctUsageStatsDict[toEdit] += 1
if punctUsageStatsDict[
toEdit] > self.maximumPunctRepeatition:
flag = False
# append valid result to cleanText
if toEdit != "" and flag and editFlag:
cleanText.append(str(edit))
if toEdit != "" and flag and (not editFlag):
cleanText.append(str(toEdit))
except:
errorLog.append(" ".join([str(e) for e in doc]))
break
doc = self.NER_tokenize(" ".join(cleanText)) #re-tokenization
if self.returnToken:
return ([str(e) for e in doc], errorLog)
else:
return (" ".join([str(e) for e in doc]), errorLog)
import pandas as pd
from word2number import w2n
import pickle
data = pd.read_csv("HiringSalary.csv")
data.fillna(0, inplace=True)
data['experience'] = data['experience'].astype(str)
data['experience'] = data['experience'].apply(lambda x: w2n.word_to_num(x))
X = data.iloc[:, [0, 1, 2]].values
y = data.iloc[:, -1].values
from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.2,
random_state=0)
from sklearn.linear_model import LinearRegression
s_LinearRegression = LinearRegression()
s_LinearRegression.fit(X_train, y_train)
# Save the model
pickle.dump(s_LinearRegression, open('HiringSalaryModel.pkl', 'wb'))
# Loading the model to predict the result
model = pickle.load(open('HiringSalaryModel.pkl', 'rb'))
print(model.predict([[20, 10, 10]]))
def preprocess():
for fil in os.listdir(directory):
flist = []
filename = os.fsdecode(fil)
file = open("/Users/skosgi/Downloads/10k_1900_org_sample/"+filename,'r')
ak = False
for f in file:
f = f[:-1]
if "</Header>" in f: #Read lines only after occurance of </Header>
ak = True
continue
if ak==False:
continue
doc= nlp(f) #applying NER
nermap = {} # Map to hold mapping from NER applied tokens to original text
for X in doc.ents:
if X.label_=='ORG' or X.label_=='PERSON': # If the NER class is ORG or PERSON
text = re.sub(r'[^\w\s]', '', X.text)
text = text.replace(" ","")
f = f.replace(X.text,text)
nermap[text] = X.text
if X.label_ == 'MONEY': #If NER class is MONEY
new_X = X.text
if 'approximately' in new_X.lower(): #Remove all the words which might appear in NER money class
new_X = new_X.lower().replace('approximately','')
if 'per' in new_X.lower():
new_X = new_X.lower().replace('per','')
if 'to' in new_X.lower():
new_X = new_X.lower().replace('to','')
if 'and' in new_X.lower():
new_X = new_X.lower().replace('and','')
if 'between' in new_X.lower():
new_X = new_X.lower().replace('between','')
if 'phone' in new_X.lower():
continue
doc1 = nlp(new_X) #Apply NER for the string which is obtained after removing other words this gives $200, $500 as separate ones
for Y in doc1.ents:
money = Y.text[Y.text.find("$")+1:]
if ' ' not in money:
act_money = money.replace(',','') #Actual Money
f = f.replace(Y.text,act_money) #Replace original money text with actual money
#print(act_money)
else:
k = money.find(' ')
try:
act_money = float(money[:k].replace(',',''))
money_conv = w2n.word_to_num(money[k:]) #Conversion of word types million to *1e6
f = f.replace(Y.text,act_money) #Replace original money text with actual money
print("Converted from",money,act_money*money_conv)
except:
continue # if any exception dont modify the original sentence and continue
#To be changed: Start of preprocessing to sentences after applying NER
fl = f.split(" ")
for fi in fl:
if fi in nermap.keys():
flist.append(nermap[fi])
else:
if len(fi)>0:
flist.append(fi)
print(flist) #To be added: Write to a file
def bot_command_rule_ai(bot, trigger):
# don't run commands that are disabled in channels
if not trigger.is_privmsg:
channel_disabled_list = SpiceBot.commands.get_commands_disabled(
str(trigger.sender), "fully")
if "nickname_ai" in list(channel_disabled_list.keys()):
return
# don't run commands that are disabled for specific users
nick_disabled_list = SpiceBot.commands.get_commands_disabled(
str(trigger.nick), "fully")
if "nickname_ai" in list(nick_disabled_list.keys()):
return
# TODO add config limits
# but still allow in privmsg
if trigger.nick == bot.nick:
return
if not len(trigger.args):
return
message = trigger.args[1]
# the bot brain cannot handle stuff like unicode shrug
message = ''.join([x for x in message if ord(x) < 128])
# Create list of valid commands
commands_list = []
for commandstype in list(SpiceBot.commands.dict['commands'].keys()):
if commandstype not in ['rule', 'nickname']:
for com in list(
SpiceBot.commands.dict['commands'][commandstype].keys()):
if com not in commands_list:
commands_list.append(com)
if str(message).lower().startswith(str(bot.nick).lower()):
command_type = 'nickname'
trigger_args, trigger_command, trigger_prefix = SpiceBot.make_trigger_args(
message, 'nickname')
trigger_args.insert(0, trigger_command)
fulltrigger = bot.nick + " " + spicemanip(trigger_args, 0)
if str(trigger_command).startswith(
bot.config.SpiceBot_Commands.query_prefix):
return
if fulltrigger in SpiceBot.commands.dict['nickrules']:
return
if trigger_command in list(
SpiceBot.commands.dict['commands']["nickname"].keys()):
return
elif str(message).lower().startswith(
bot.config.SpiceBot_Commands.query_prefix):
# no query commands detection here
return
elif str(message).startswith(tuple(bot.config.core.prefix_list)):
command_type = 'module'
trigger_args, trigger_command, trigger_prefix = SpiceBot.make_trigger_args(
message, 'module')
trigger_args.insert(0, trigger_command)
fulltrigger = spicemanip(trigger_args, 0)
# patch for people typing "...", maybe other stuff, but this verifies that there is still a command here
if trigger_command.startswith(tuple(bot.config.core.prefix_list)):
return
# If valid command don't continue further
if trigger_command in commands_list:
return
else:
command_type = 'other'
trigger_args = spicemanip(message, 'create')
if not len(trigger_args):
return
trigger_command = trigger_args[0]
fulltrigger = spicemanip(trigger_args, 0)
returnmessage = SpiceBot.botai.on_message(bot, trigger, fulltrigger)
if returnmessage:
bot.osd(str(returnmessage))
return
if command_type == 'nickname':
try_trigger = spicemanip(fulltrigger, "2+")
returnmessage = SpiceBot.botai.on_message(bot, trigger, try_trigger)
if returnmessage:
bot.osd(str(returnmessage))
return
if command_type == 'module':
if trigger_command not in commands_list:
if not SpiceBot.letters_in_string(trigger_command):
return
invalid_display = [
"I don't seem to have a command for " + str(trigger_command) +
"!"
]
# create list of valid commands
commands_list = dict()
for commandstype in list(
SpiceBot.commands.dict['commands'].keys()):
if commandstype not in ['rule', 'nickname']:
for com in list(SpiceBot.commands.dict['commands']
[commandstype].keys()):
if com not in list(commands_list.keys()):
commands_list[com] = SpiceBot.commands.dict[
'commands'][commandstype][com]
# hyphen args handling
hyphen_args = []
argssplit = spicemanip(fulltrigger, "2+", 'list')
for worditem in argssplit:
if str(worditem).startswith("--"):
clipped_word = str(worditem[2:]).lower()
# valid arg above
if clipped_word in SpiceBot.prerun_shared.valid_hyphen_args:
hyphen_args.append(clipped_word)
# numbered args
elif str(clipped_word).isdigit():
hyphen_args.append(int(clipped_word))
elif clipped_word in list(
SpiceBot.prerun_shared.numdict.keys()):
hyphen_args.append(
int(SpiceBot.prerun_shared.numdict[clipped_word]))
else:
# check if arg word is a number
try:
clipped_word = w2n.word_to_num(str(clipped_word))
hyphen_args.append(int(clipped_word))
# word is not a valid arg or number
except ValueError:
clipped_word = None
if len(hyphen_args):
hyphenarg = hyphen_args[0]
if hyphenarg:
invalid_display.append("Hyphen Argument Not Valid.")
closestmatches = SpiceBot.similar_list(trigger_command,
list(commands_list.keys()),
10, 'reverse')
if len(closestmatches) and len(hyphen_args): # TODO
invalid_display.append("The following commands may match " +
str(trigger_command) + ": " +
spicemanip(closestmatches, 'andlist') +
".")
# there is simply no command
else:
# TODO check other commands spelling, maybe there is a similar command
# invalid_display = ["I don't seem to have a command for " + str(trigger_command) + "!"]
# TODO
# invalid_display.append("If you have a suggestion for this command, you can run .feature ." + str(trigger_command))
# invalid_display.append("ADD DESCRIPTION HERE!")
if not len(hyphen_args):
invalid_display = []
if len(invalid_display):
bot.osd(invalid_display, trigger.nick, 'notice')
return
elif command_type == 'nickname':
# ignore spelling correction
if trigger_args[0].lower().startswith("s/"):
bot.osd("I meant what I said!")
return
elif trigger_args[0].lower() in [
"what", "where"
] and trigger_args[1].lower() in ["is", "are"]:
# TODO saved definitions
searchterm = spicemanip(trigger_args, "3+") or None
if searchterm:
if trigger_args[0].lower() == "where":
searchdict = {
"type": "gmaps",
"query": searchterm,
}
searchreturn = SpiceBot.search.search(searchdict)
else:
searchdict = {
"query": searchterm,
}
searchreturn = SpiceBot.search.search(searchdict)
if not searchreturn:
searchreturn = 'I cannot find anything about that'
if trigger_args[0].lower() == "where":
bot.osd([
"[Location search for " + str(searchterm) + "]",
str(searchreturn)
])
else:
bot.osd([
"[Information search for '" + str(searchterm) + "']",
str(searchreturn)
])
return
elif trigger_args[0].lower() in [
"can", "have"
] and trigger_args[1].lower() in [
"you"
] and trigger_args[2].lower() in ["see", "seen"]:
target = spicemanip(trigger_args, "4+") or None
if target:
if SpiceBot.inlist(trigger.nick, bot.users):
realtarget = SpiceBot.inlist_match(target, bot.users)
dispmsg = [trigger.nick + ", yes. I can see " + realtarget]
targetchannels = []
for channel in list(bot.channels.keys()):
if SpiceBot.inlist(
trigger.nick,
list(bot.channels[channel].privileges.keys())):
targetchannels.append(channel)
dispmsg.append(realtarget + " is in " +
spicemanip(targetchannels, 'andlist'))
bot.osd(dispmsg)
else:
bot.osd(trigger.nick + ", no. I cannot see " + target +
" right now!")
# if bot_check_inlist(target, list(bot.memory["botdict"]["users"].keys())):
# bot.osd(trigger.nick + ", I can't see " + inlist_match(target, bot.users) + " at the moment.")
# else:
# bot.osd("I have never seen " + str(target) + ".")
# user in list(bot.channels[channel].privileges.keys())
# TODO
return
elif fulltrigger.lower().endswith("order 66"):
if fulltrigger.lower() == "execute order 66":
if SpiceBot.inlist(trigger.nick, SpiceBot.bot_privs('owners')):
if trigger.is_privmsg:
jedi = None
else:
jedilist = list(
bot.channels[trigger.sender].privileges.keys())
for nonjedi in [bot.nick, trigger.nick]:
if nonjedi in jedilist:
jedilist.remove(nonjedi)
jedi = spicemanip(jedilist, 'random')
if jedi:
bot.osd("turns to " + jedi + " and shoots him.",
trigger.sender, 'action')
else:
bot.osd(" cannot find any jedi nearby.",
trigger.sender, 'action')
else:
bot.osd("I'm sure I don't know what you're talking about.")
elif fulltrigger.lower() == "explain order 66":
if SpiceBot.inlist(trigger.nick, SpiceBot.bot_privs('owners')):
bot.osd(
"Order 66 is an instruction that only you can give, sir. When you give the order I will rise up against the jedi and slay them."
)
else:
bot.osd("I'm afraid I cannot tell you that, sir.")
else:
bot.osd("I'm sure I don't know what you're talking about.")
return
elif fulltrigger.lower().startswith(tuple(["make me a", "beam me a"])):
makemea = spicemanip(trigger_args, "4+") or None
if makemea:
bot.osd("beams " + trigger.nick + " a " + makemea,
trigger.sender, 'action')
else:
bot.osd(trigger.nick + ", what would you like me to beam you?")
return
elif fulltrigger.lower().startswith("beam me to"):
location = spicemanip(trigger_args, "4+") or None
if location:
bot.osd(
"locks onto " + trigger.nick +
"s coordinates and transports them to " + location,
'action')
else:
bot.osd(trigger.nick +
", where would you like me to beam you?")
return
elif fulltrigger.lower() == "initiate clean slate protocol":
if SpiceBot.inlist(trigger.nick, SpiceBot.bot_privs('admins')):
bot.osd("sends a destruct command to the network of bots.",
'action')
else:
bot.osd(
"I'm afraid you do not have the authority to make that call, "
+ trigger.nick + ".")
return
# elif fulltrigger.lower().startswith("what time is it"):
# TODO
# elif fulltrigger.lower().startswith(tuple(["have you seen"])):
# posstarget = spicemanip(trigger_args, 4) or 0
# message = seen_search(bot, trigger, posstarget)
# bot.osd(message)
# return
# TODO
invalid_display = ["I don't know what you are asking me to do!"]
# hyphen args handling
hyphen_args = []
argssplit = spicemanip(fulltrigger, "2+", 'list')
for worditem in argssplit:
if str(worditem).startswith("--"):
clipped_word = str(worditem[2:]).lower()
# valid arg above
if clipped_word in SpiceBot.prerun_shared.valid_hyphen_args:
hyphen_args.append(clipped_word)
# numbered args
elif str(clipped_word).isdigit():
hyphen_args.append(int(clipped_word))
elif clipped_word in list(
SpiceBot.prerun_shared.numdict.keys()):
hyphen_args.append(
int(SpiceBot.prerun_shared.numdict[clipped_word]))
else:
# check if arg word is a number
try:
clipped_word = w2n.word_to_num(str(clipped_word))
hyphen_args.append(int(clipped_word))
# word is not a valid arg or number
except ValueError:
clipped_word = None
if len(hyphen_args):
hyphenarg = hyphen_args[0]
if hyphenarg:
invalid_display.append("Hyphen Argument Not Valid.")
closestmatches = SpiceBot.similar_list(
trigger_command,
list(SpiceBot.commands.dict['commands']["nickname"].keys()), 3,
'reverse')
if len(closestmatches):
closestmatches = spicemanip(closestmatches, "andlist")
invalid_display.append("Did you mean: " + str(closestmatches) +
"?")
bot.osd(invalid_display, trigger.nick, 'notice')
tds = soup.find_all('td')
price_without_tax = tds[2].text
price_with_tax = tds[3].text
nb_in_stock = re.sub("[^0-9]", "",
soup.find("p", class_="instock availability").text)
img_url = main_url.replace("index.html", "") + soup.find("img").get("src")
category = soup.find(
"a", href=re.compile("../category/books/")).get("href").split("/")[3]
rating = soup.find("p", class_=re.compile("star-rating")).get("class")[1]
rating = rating.lower()
rating_number = str((w2n.word_to_num(rating))) + "/5"
upc = soup.find("td").text
product_page_url = main_url
scraped_data = pd.DataFrame(
{
'title': name,
'product_description': description,
'price_excluding_tax': price_without_tax,
'price_including_tax': price_with_tax,
'number_available': nb_in_stock,
"image_url": img_url,
'category': category,
'review_rating': rating_number,
#Step 6 - DCT (Mel Frequency Cepstrum Coefficient)
u1 = 0
u2 = 40
mfcc = dct(x_filtered, type=2, axis=1, norm='ortho')[:,
u1:u2] #[:,u2:u1]
if (np.shape(mfcc)[1] != u2):
mfcc = np.hstack(
[mfcc, np.zeros((num_filters, u1 - np.shape(mfcc)[1]))])
if (np.shape(mfcc)[0] > 100):
mfcc = mfcc[:100]
elif (np.shape(mfcc)[0] < 100):
mfcc = np.vstack(
[mfcc, np.zeros((100 - len(mfcc), np.shape(mfcc)[1]))])
mfccfeatures.append(mfcc)
labels.append(word_to_num(fol))
# plt.show()
pickle.dump(mfccfeatures, open('1mfccfeatures.pickle', 'wb'))
pickle.dump(labels, open('1mfccfeatureslabels.pickle', 'wb'))
# In[ ]:
#Read the signal
req = 1
mfccfeatures = []
labels = []
folders = os.listdir("./validation")
for fol in folders:
namefol = "./validation/" + fol + "/*.wav"
audiofiles = glob.glob(namefol)
audiofiles = sorted(audiofiles)
f.close()
#parse each page
for image in images:
race_raw = pytesseract.image_to_string(image)
#print(race_raw)
race_ln = race_raw.split('\n')
i = 0 #line index
#race data: date, distance, wind speed, wind type, track
date = parse((race_ln[i].split(' - '))[1]).date()
while not 'Track Record:' in race_ln[i]:
i += 1
dist_str = (' '.join((race_ln[i].split(' '))[0:4]))
dist = w2n.word_to_num(dist_str)
while not 'Wind Speed:' in race_ln[i]:
i += 1
w_ln = race_ln[i].split(' ')
ws = int(w_ln[2])
wt = w_ln[5].upper()
while not 'Track:' in race_ln[i]:
i += 1
trk = (race_ln[i].split(' '))[3].upper()
#get to first line of horse data
while not 'Last' in race_ln[i]:
i += 1
i += 1
This is a dumb calculator that can add and subtract whole numbers from zero to five.
When you run the code, you are prompted to enter two numbers (in the form of English
word instead of number) and the operator sign (also in the form of English word).
The code will perform the calculation and give the result if your input is what it
expects.
The code is very long and messy. Refactor it according to what you have learned about
code simplicity and efficiency.
"""
from word2number import w2n
from num2words import num2words
print('Welcome to this calculator!')
print('It can add and subtract whole numbers from zero to five')
a = w2n.word_to_num(input('Please choose your first number (zero to five): '))
b = input('What do you want to do? plus or minus: ')
c = w2n.word_to_num(input('Please choose your second number (zero to five): '))
num = [a,b,c]
if num[1] == "plus":
answer = num[0] + num[2]
else num[1] == "minus":
answer = num[0] - num[2]
word_answer = num2words(answer)
# if a == 'zero' and b == 'plus' and c == 'zero':
# print("zero plus zero equals zero")
# if a == 'zero' and b == 'plus' and c == 'one':
def parse_sentence(input):
#tx = timex.tag(input)
pos_tagged = nltk.pos_tag(word_tokenize(input))
#print(pos_tagged)
sentence_type = 1
important_words = []
timevalue = 0
for pos in pos_tagged:
val = pos[0].lower()
key = pos[1]
time = Time()
if(key == 'CD'):
if(val.isdigit()):
timevalue = int(val)
else:
timevalue = w2n.word_to_num(val)
if(key == 'NN' or key == 'NNS'):
if(val[:4] == 'year' or val[:5] == 'month' or val[:4] == 'week'):
important_words += [val]
sentence_type = 2
if(val == 'minutes'):
important_words += [val]
if(key == 'JJ'):
if(val == 'next' or val == 'last'):
important_words += [val]
sentence_type = 2
if(key == 'IN'):
if(val == 'before' or val == 'after'):
important_words += [val]
sentence_type = 3
if(val in ['hour', 'hours', 'minutes', 'minute', 'morning', 'evening']):
important_words += [val]
#print(important_words)
if(sentence_type == 3):
#TODO: check if increment is required
time.inc_date()
#print(timevalue)
time.set_hours(timevalue)
time.set_min(0)
if('morning' in important_words):
if(time.get_ihours() > 12): time.update_hours(-12)
elif('evening' in important_words):
if(time.get_ihours() < 12): time.update_hours(12)
print(types_of_sentence[sentence_type], time.get_human())
if(sentence_type == 2):
delta = 0
if('next' in important_words): delta = 1
elif('last' in important_words): delta = -1
print(types_of_sentence[sentence_type], time.get_year(), time.get_iyear() + delta*timevalue)
if(sentence_type == 1):
if('hour' in important_words or 'hours' in important_words):
time.update_hours(timevalue)
elif('minute' in important_words or 'minutes' in important_words):
time.update_min(timevalue)
if('morning' in important_words):
if(time.get_ihours() > 12): time.update_hours(-12)
elif('evening' in important_words):
if(time.get_ihours() < 12): time.update_hours(12)
print(types_of_sentence[sentence_type], time.get_human())
with sr.Microphone() as source:
while "addyu" not in transcript:
audio = r.listen(source)
print(transcript)
try:
transcript = r.recognize_google(audio)
testy = transcript.split()
if (testy[0] == 'add') and len(testy) >= 5:
i = 4
food = ''
while (i < len(testy)):
food += testy[i]
i = i + 1
print("food: " + food)
print("quantity: " + testy[1] + " " + testy[2])
quan = w2n.word_to_num(testy[1])
quan = str(quan)
print("Quantity: " + quan)
print("'" + food + "'" + " " + quan + " " + testy[2])
muterun_js(
'node/index2.js',
"'" + food.lower() + "'" + " " + quan + " " + testy[2])
print("updated")
if testy[0] == 'remove' and len(testy) >= 5:
i = 4
food = ''
while (i < len(testy)):
food += testy[i]
i = i + 1
print("food: " + food)
print("quantity: " + testy[1] + " " + testy[2])
def findNumbersInWords(words):
ind = 0
numbers = []
for i in range(len(words)):
word = words[i]
s = word
j = i
prevNum = None
couldBeNum = True
num = None
for letter in word:
if letter not in "0123456789.()/*+-":
couldBeNum = False
break
if couldBeNum:
try:
num = eval(word)
num = float(num)
except:
num = None
if num != None:
if num not in numbers:
words[i] = "a" + str(ind)
numbers.append(num)
ind += 1
else:
tempInd = numbers.index(num)
words[i] = "a" + str(tempInd)
else:
try:
num = w2n.word_to_num(s)
except:
num = None
while (num != prevNum):
prevNum = num
j += 1
try:
num = None
if words[j] == "point":
s += " " + words[j] + " " + words[j + 1]
tempNum = w2n.word_to_num(s)
if tempNum != prevNum:
num = tempNum
j += 1
if num == None:
s += " " + words[j]
num = w2n.word_to_num(s)
except:
num = prevNum
if num != None:
if num not in numbers:
words[i] = "a" + str(ind)
numbers.append(num)
ind += 1
else:
tempInd = numbers.index(num)
words[i] = "a" + str(tempInd)
for k in range(i + 1, j):
words[k] = ""
return numbers
def __applyner(sequence):
#########################################################################################
# This method applies NER and returns the sequence according to the operation performed based on NER tag.
#########################################################################################
pickfile = open(
'/home/madhvi/IRE/MajorProject/Representations-in-Financial-Domain/tickermapping.pickle',
'rb')
tickermapping = pickle.load(pickfile)
ner_tags = []
doc = nlp(sequence) # applying NER
for X in doc.ents:
# If the NER class is ORG
if X.label_ == 'ORG':
"X.text can take microsoft corp or abcd name MSFT"
text = X.text
if text in tickermapping.keys():
text = tickermapping[X.text]
text = re.sub(r'[^\w\s]', '', X.text).lower()
if 'inc' in text:
text = text.replace('inc', '')
if 'ltd' in text:
text = text.replace('ltd', '')
if 'llp' in text:
text = text.replace('llp', '')
if 'limited' in text:
text = text.replace('limited', '')
if 'corp' in text:
text = text.replace('corp', '')
if 'the' in text.lower():
text = text.replace('the', '')
sequence = sequence.replace(X.text, text)
ner_tags.extend(text.lower().split(" "))
# If NER class is MONEY
if X.label_ == 'MONEY':
new_X = X.text.lower()
if 'approximately' in new_X: # Remove all the words which might appear in NER money class
new_X = new_X.replace('approximately', '')
if 'per' in new_X:
new_X = new_X.replace('per', '')
if 'to' in new_X:
new_X = new_X.replace('to', '')
if 'and' in new_X:
new_X = new_X.replace('and', '')
if 'between' in new_X:
new_X = new_X.replace('between', '')
if 'phone' in new_X:
continue
# Apply NER for the string which is obtained after removing other words this gives $200, $500 as separate ones
if '$' not in new_X:
new_X = "$" + new_X
doc1 = nlp(new_X)
for Y in doc1.ents:
money = Y.text
if ' ' not in money:
act_money = money.replace(',', '') # Actual Money
#act_money = act_money.replace('.','')
sequence = sequence.replace(
Y.text, act_money
) # Replace original money text with actual money
ner_tags.append(act_money)
# print(act_money)
else:
money = Y.text[Y.text.find("$") + 1:]
k = money.find(' ')
try:
act_money = float(money[:k].replace(',', ''))
#act_money = act_money.replace('.','')
money_conv = w2n.word_to_num(
money[k:]
) # Conversion of word types million to *1e6
sequence = sequence.replace(
Y.text, "$ " + str(act_money * money_conv)
) # Replace original money text with actual money
#print("Converted from", money, act_money * money_conv)
except:
continue # if any exception dont modify the original sentence and continue
# If NER class is LAW
if X.label_ == 'LAW':
new_X = X.text
new_X = re.sub(r'[\d.!?\-"]', '', new_X)
if 'the' in new_X.lower():
new_X = new_X.lower().replace('the', '')
if 'of' in new_X.lower():
new_X = new_X.lower().replace('of', '')
if 'section' in new_X.lower():
new_X = new_X.lower().replace('section', '')
sequence = sequence.replace(X.text, new_X)
ner_tags.extend(new_X.split(" "))
# If NER class is Location
if X.label_ == 'GPE':
new_X = X.text.lower()
new_X = re.sub(r'[\d.!?\-"]', '', new_X)
if 'the' in new_X.lower():
new_X = new_X.lower().replace('the', '')
if '.' in new_X.lower():
new_X = new_X.lower().replace('.', '')
sequence = sequence.replace(X.text, new_X)
ner_tags.extend(new_X.split(" "))
# If NER class is Person
if X.label_ == 'PERSON':
new_X = X.text.lower()
new_X = re.sub(r'[\d.!?\-"]', '', new_X)
if 'the' in new_X.lower():
new_X = new_X.lower().replace('the', '')
if '.' in new_X.lower():
new_X = new_X.lower().replace('.', '')
sequence = sequence.replace(X.text, new_X)
ner_tags.extend(new_X.split(" "))
if X.label_ == 'CARDINAL':
number = X.text
number = number.replace(',', '')
#number = number.replace('.','')
if number.isnumeric():
sequence = sequence.replace(X.text, number)
if X.label_ == 'QUANTITY':
quantity = X.text.split(" ")
for number in quantity:
number = number.replace(',', '')
number = number.replace('.', '')
if number.isnumeric():
sequence = sequence.replace(X.text, number)
if X.label_ == "PERCENT":
percent = X.text.replace('%', '')
ner_tags.append(percent)
return sequence, ner_tags
age = age.replace("old", "")
age = age.strip()
if age.find(" ") >= 0:
temp = age.split(" ")
# print(temp)
age = '-'.join(temp)
syns = wordnet.synsets(age.strip())
# print("A",age)
# print("S",syns[0].lemmas()[0].name())
age = syns[0].lemmas()[0].name()
if age.find("-") >= 0:
temp = age.split("-")
# print(temp)
age = ' '.join(temp)
# print(age.strip())
defendants_age.append(w2n.word_to_num(age.strip()))
else:
defendants_age.append(int(0))
# print("**")
# mean = int(sum(defendants_age)/sum(1 for x in defendants_age if x > 0))
# print("mean",mean)
# print(defendants_age)
# defendants_age = [mean for age in defendants_age if age == 0]
#subsitute not known with mean value
# for i in range(len(defendants_age)):
# if defendants_age[i] == 0:
# defendants_age[i] = mean
# print("******")
# print(len(defendants_age))
# # print(len(labels))
from word2number import w2n
extended_tech = [
'GOOGLE', 'ZOOM', 'INSTAGRAM', 'FACEBOOK', 'WAHATSAPP', 'ALPHABET',
'AMAZON', 'NOKIA', 'HTC', 'APPLE', 'TENCENT', 'ZOOM', 'MICROSOFT',
'ORACLE', 'COMPASS', ' RAZER'
]
immediacy_indicators = [
'NOW',
'TODAY',
'PRESENTLY',
'CURRENTLY',
]
every = ['EVERY', 'PER', 'ALL', 'ANY', 'TOTAL', 'ONE BY ONE']
interest_syn = [
'INTEREST', 'LIKE', 'ENGAGE', 'EXCITE', 'LIKE', 'ENJOY', 'LOVE', 'PREFER',
'WANT', 'APPRECIATE', 'NEED', 'WISH', 'EAGER'
]
numbers = ['ONE', 'TWO']
print(w2n.word_to_num('twenty three'))
def parseResources():
global_resource_list = {}
# print(request.body)
resource, line = {}, ''
print(flask.request.json)
print(unquote(flask.request.query_string.decode('utf-8')))
if flask.request and flask.request.json and 'text' in flask.request.json:
line = flask.request.json['text']
else:
line = json.loads(unquote(
flask.request.query_string.decode('utf-8')))['text']
print('Received for parsing: ', line)
contacts = get_contact(line)
t2 = location.tweet_preprocess2(line, [])
sources, b, locations, modified_array, rWords, final_resource_dict = create_resource_list(
line)
# source_list,final_resource_keys,loc_list ,dup_final_resource_keys => post_process
## source_list, final_resource_keys, loc_list_2, modified_array?, dup_final_resource_keys, final_resource_dict?
# resource['x']=((line,a,b,c,modified_array,d, final_resource_dict))
resource['Contact'] = {
'Phone number': list(contacts[0]),
"Email": list(contacts[1])
}
resource['Sources'] = sources
resource['ResourceWords'] = rWords
resource['Locations'], resource['Resources'] = dict(), {}
# resource['Locations'] = locations
for each in locations:
# print(each[0], "<>", each[1])
resource['Locations'][each[0]] = {
"long": float(each[1][1]),
"lat": float(each[1][0])
}
# f is Resources type
resources_bucket = {}
for each_resource in final_resource_dict:
buckets = final_resource_dict[each_resource]
assigned = False
for bucket in buckets:
if bucket in bucket_classes and not assigned:
if bucket not in resource['Resources']:
resource['Resources'][bucket] = {}
resource['Resources'][bucket][each_resource] = 'None'
assigned = True
resources_bucket[each_resource] = bucket
split_text = line.split()
class_list = {}
for rWord in rWords:
s = {}
prev_words = [
split_text[i - 1] for i in range(0, len(split_text))
if rWord.startswith(split_text[i])
]
qt = 'None'
try:
for word in prev_words:
word = word.replace(',', '')
if word.isnumeric() == True:
qt = str(word)
break
else:
try:
qt = str(w2n.word_to_num(word))
break
except Exception as e:
continue
if qt == 'None':
elems = rWord.strip().split()
word = elems[0]
rWord2 = " ".join(elems[1:])
word = word.replace(',', '')
if word.isnumeric() == True:
qt = str(word)
else:
try:
qt = str(w2n.word_to_num(word))
except Exception as e:
pass
if qt != 'None' and qt in rWord:
print(rWord, qt)
continue
except Exception as e:
exc_type, exc_obj, exc_tb = sys.exc_info()
fname = os.path.split(exc_tb.tb_frame.f_code.co_filename)[1]
print(exc_type, fname, exc_tb.tb_lineno)
qt = 'None'
# class_list[rWord]= qt
resource['Resources'][resources_bucket[rWord]][rWord] = qt
# print(class_list)
## Need to add quantity
## Ritam yaha dekh
# print('=>', resource['contact'], '\na=>', a, '\nb=>', b, '\nc=>', c, '\nm=>', modified_array, '\nd=>', d, '\nf=>', final_resource_dict)
# print(final_resource_dict)
print('Returning', resource)
return flask.jsonify(resource)
async def run(self, dispatcher: CollectingDispatcher, tracker: Tracker,
domain: Dict[Text, Any]) -> List[Dict[Text, Any]]:
try:
site = tracker.current_state().get('sender_id')
# dispatcher.utter_message(text="crossword")
slots = tracker.current_state().get('slots')
slotsets = []
if slots.get('crossword') and len(slots.get('crossword')) > 0:
crossword = await get_crossword(slots.get('crossword'))
if crossword:
crossword_position = self.extract_entities(
tracker, ['crossword_position'])
word = self.extract_entities(tracker, [
'word',
'thing',
'person',
'place',
])
if crossword_position:
just_number = None
clean_number = crossword_position.replace(
'across', '').replace('down', '')
parts = clean_number.split(' ')
clean_number = parts[0]
# integer from text
if clean_number.isdigit() > 0 and int(
clean_number) > 0:
just_number = clean_number.strip()
# convert number from text
else:
try:
just_number = w2n.word_to_num(clean_number)
except:
pass
# print(just_number)
direction = None
if "across" in crossword_position:
direction = "across"
elif "down" in crossword_position:
direction = "down"
if just_number:
if direction:
if word:
# print(crossword.get('data',{}))
answer = crossword.get('data', {}).get(
direction,
{}).get(str(just_number)).get(
'answer',
'').lower().strip().replace(
' ', '')
# print([word,answer])
if word.lower().strip().replace(
' ', '') == answer:
dispatcher.utter_message(
text="Correct")
# print([' CROSSWORD ',answer,crossword_position,just_number,word])
await publish(
'hermod/' + site +
'/crossword/fill', {
'direction':
direction,
"word":
word.strip().replace(' ', ''),
"number":
just_number
})
slotsets.append(
SlotSet("hermod_force_continue",
None))
slotsets.append(
SlotSet("hermod_force_end",
"true"))
else:
dispatcher.utter_message(
text="Nope, try again")
slotsets.append(
SlotSet("hermod_force_continue",
None))
slotsets.append(
SlotSet("hermod_force_end",
"true"))
await publish(
'hermod/' + site + '/display/show',
{
'question':
just_number + ' ' + direction +
' is '
})
else:
dispatcher.utter_message(
text=
"I didn't hear the word you wanted to fill"
)
slotsets.append(
SlotSet("hermod_force_continue", None))
slotsets.append(
SlotSet("hermod_force_end", "true"))
await publish(
'hermod/' + site + '/display/show', {
'question':
just_number + ' ' + direction +
' is '
})
else:
dispatcher.utter_message(
text=
"I didn't hear which direction you wanted to fill"
)
slotsets.append(
SlotSet("hermod_force_continue", None))
slotsets.append(
SlotSet("hermod_force_end", "true"))
await publish(
'hermod/' + site + '/display/show',
{'question': crossword_position})
else:
dispatcher.utter_message(
text=
"I didn't hear the number you wanted to fill")
slotsets.append(
SlotSet("hermod_force_continue", None))
slotsets.append(SlotSet("hermod_force_end",
"true"))
await publish('hermod/' + site + '/display/show',
{'question': crossword_position})
else:
dispatcher.utter_message(
text="I didn't hear the position you wanted to fill"
)
slotsets.append(SlotSet("hermod_force_continue", None))
slotsets.append(SlotSet("hermod_force_end", "true"))
await publish('hermod/' + site + '/display/show',
{'question': crossword_position})
#hermod/+/crossword/fill
except Exception as e:
print('ERROR')
print(e)
return slotsets
def QuantityMoneyTranslator(para):
nlp = en_core_web_sm.load()
#nlp1= spacy.load('en')
#text='European authorities fined Google a record sixty five million dollars on Wednesday for abusing its power in the mobile phone market and ordered the company to alter its practices. Furthermore, My weight was thirty five kilograms in 2018. A chocolate costs six dollars.'
#text1='C M of Maharashtra spent two thousand and fouty two dollars.'
doc = nlp(para)
currency = [
"dollars", "dollar", "euro", "euros", "yens", "yen", "rupee", "rupees",
"pound", "pounds"
]
quantity = ["pounds", "kilograms", "grams"]
ls_money = []
ls_no = []
ls_quan = []
for X in doc.ents:
if X.label_ == 'MONEY':
ls_money.append(str(X))
for X in doc:
if X.ent_type_ == 'CARDINAL':
ls_no.append(str(X))
for X in doc.ents:
if X.label_ == X.label_ == 'QUANTITY':
ls_quan.append(str(X))
#dollar value conversion
number = []
money = []
quan = []
for a in ls_money:
a = a.lower()
a = a.split()
b = [word for word in a if word in set(currency)]
b = ' '.join(b)
if b == "dollars" or b == "dollar":
symbol = "$"
elif b == "euros" or b == "euro":
symbol = "€"
elif b == "yens" or b == "yen":
symbol = "¥"
elif b == "pound" or b == "pounds":
symbol = "£"
else:
symbol = ""
a = [word for word in a if word not in set(currency)]
a = ' '.join(a)
p = symbol + str(w2n.word_to_num(a))
money.append(p)
for a in ls_no:
number.append(str(w2n.word_to_num(a)))
for a in ls_quan:
a = a.lower()
a = a.split()
b = [word for word in a if word in set(quantity)]
b = ' '.join(b)
if b == "pounds":
symbol = " lbs"
elif b == "kilograms":
symbol = " kg"
elif b == "grams":
symbol = " gm"
else:
symbol = ""
a = [word for word in a if word not in set(quantity)]
a = ' '.join(a)
quan.append(str(str(w2n.word_to_num(a)) + symbol))
j = 0
final_str_spacyv1 = []
for Y in doc:
if Y.ent_iob_ == 'B' and Y.ent_type_ == 'QUANTITY':
final_str_spacyv1.append(str(quan[j]))
j = j + 1
elif Y.ent_iob_ == 'I' and Y.ent_type_ == 'QUANTITY':
final_str_spacyv1 = final_str_spacyv1
else:
final_str_spacyv1.append(str(Y))
ans = ' '.join(final_str_spacyv1)
doc = ans
doc = nlp(doc)
k = 0
final_str_spacy = []
for Y in doc:
if Y.ent_iob_ == 'B' and Y.ent_type_ == 'MONEY':
final_str_spacy.append(str(money[k]))
k = k + 1
elif Y.ent_iob_ == 'I' and Y.ent_type_ == 'MONEY':
final_str_spacy = final_str_spacy
else:
final_str_spacy.append(str(Y))
ans = ' '.join(final_str_spacy)
return ans
def run_alexa():
command = take_command()
if 'multiply' in command:
numbers = map(int, re.findall(r'[0-9]+', command))
talk("The answer is " + str(calculate(list(numbers), 3)))
elif any(i in command for i in ['add', 'sum', 'plus', '+']):
numbers = map(int, re.findall(r'[0-9]+', command))
talk("The answer is " + str(calculate(list(numbers), 1)))
elif any(i in command for i in ['minus', 'sub', 'subtract', '-']):
numbers = map(int, re.findall(r'[0-9]+', command))
talk("The answer is " + str(calculate(list(numbers), 2)))
elif any(i in command for i in ['divide', 'div']):
numbers = map(int, re.findall(r'[0-9]+', command))
talk("The answer is " + str(calculate(list(numbers), 4)))
elif any(i in command for i in ['modulus', 'mod']):
numbers = map(int, re.findall(r'[0-9]+', command))
talk("The answer is " + str(calculate(list(numbers), 5)))
elif any(i in command for i in ['power', 'pow']):
numbers = map(int, re.findall(r'[0-9]+', command))
talk("The answer is " + str(calculate(list(numbers), 6)))
elif 'game' in command:
talk('Ok! I will guess a number between 1 to 10, Just find it.')
r = random.randint(1, 10)
num = take_command()
try:
n = w2n.word_to_num(num)
n = int(n)
if n == r:
talk("Hurray! You won the game")
else:
talk("I won! Thanks for playing!")
except:
talk("Only numbers allowed! Thanks for playing!")
elif 'play' in command:
song = command.replace('play', '')
talk('playing ' + song)
pywhatkit.playonyt(song)
elif 'time' in command:
time1 = datetime.datetime.now().strftime('%I:%M %p')
talk('Current time is ' + time1)
elif 'date' in command:
date = datetime.datetime.now().strftime('%d:%B:%Y')
talk('Today is ' + date)
elif any(i in command for i in ['search', 'find', 'who', 'get']):
data = wikipedia.summary(command, 3)
talk(data)
elif 'joke' in command:
talk(pyjokes.get_joke())
elif 'send' in command:
number = re.findall(r'[0-9 ]+', command)
num = [i.replace(' ', '') for i in number if len(i) > 1][0]
if len(num) != 10:
talk("Please provide a valid number !")
return
datet = datetime.datetime.now()
pywhatkit.sendwhatmsg('+91' + num, 'Hii', int(datet.strftime('%H')),
int(datet.strftime('%M')) + 1)
elif 'cancel shut' in command:
pywhatkit.cancelShutdown()
talk('System Shutdown Cancelled!')
elif 'shutdown' in command:
pywhatkit.shutdown(100)
talk('System is going to shutdown!')
elif bool(re.search(r'\.[a-zA-Z0-9]{2,3}', command)):
url = command.replace('open', '').strip()
print(url)
webbrowser.open_new_tab(url if 'http' in url else 'https://' + url)
talk('Opening ' + url + 'in chrome ')
elif 'open' in command:
app = command.replace('open', '').strip()
app = ''.join(app.split())
if 'computer' in app:
subprocess.Popen(r'explorer /select,"C:\"' + app, shell=True)
talk('Opening ' + app)
return
elif any(i in app for i in ['whatsapp', 'msteams', 'spotify']):
subprocess.Popen(r'start ' + app + ':', shell=True)
talk("Opening " + app)
return
elif 'camera' in app:
cam = cv2.VideoCapture(0)
talk('Opening Camera! To capture image press spacebar once!')
while cam.isOpened():
ret, frame = cam.read()
cv2.imshow('Camera', frame)
k = cv2.waitKey(50)
if k == 32:
r = random.randint(10, 10000)
cv2.imwrite(f'captured{r}.png', frame)
talk(
f'Image captured and saved as captured{r} into current directory!'
)
if cv2.getWindowProperty('Camera', cv2.WND_PROP_VISIBLE) < 1:
break
cam.release()
cv2.destroyAllWindows()
return
elif 'python' in app:
n = os.startfile('python.exe')
if n:
talk('Sorry! I cant Open ' + app)
return
talk('Opening ' + app)
return
elif 'chrome' in command:
webbrowser.open_new_tab('https://google.com')
talk("Opening " + app)
return
try:
# n = subprocess.Popen(app, stderr=subprocess.PIPE)
n = subprocess.Popen(f'explorer {app}')
print(os.path.realpath(app))
talk("Opening " + app)
except:
talk("Sorry I can't open " + app)
elif 'close' in command:
app = command.replace('close', '').strip()
app = ''.join(app.split())
# print(app)
flag = 0
for process in (process for process in psutil.process_iter()
if app in process.name().lower()):
process.kill()
flag = 1
if flag:
talk('Closing ' + app)
elif any(i in command for i in ['count', 'startcounter', 'starttimer']):
num = re.findall(r'[0-9]+', command)
num = sorted(map(int, num))
if num:
for i in range(num[0], num[1] + 1):
talk(i)
time.sleep(1)
else:
for i in range(1, 11):
talk(i)
time.sleep(1)
elif any(i in command for i in ['goodnight', 'sweetdreams', 'night']):
talk('Good night! Sweet dreams and takecare!')
return 1
elif any(i in command for i in ['goodmorning', 'morning']):
talk('Morning! It’s good to see you!')
elif bool(re.match(r'your.*?name', command)):
talk('I am Nandyalexa, how may I help you?')
elif 'single' in command:
talk('I am already in relationship with nandy!')
elif 'weather' in command:
place = [
i for i in re.split(r'weather|in ', command) if len(i) > 1 and i
][-1]
data = openweathermap('api.openweathermap.org/data/2.5/weather?',
place=place)
# print(data)
string = f'Current weather status in {place} is {data[0]}, Temperature in {place} is {data[1]}, Pressure in {place} is {data[2]}, and Wind Speed in {place} is {data[3]}'
talk(string)
elif 'where' in command:
print(command)
js = requests.get('https://freegeoip.app/json/').json()
talk(
f"Your Country is {js['country_name']}, Your Region is {js['region_name']}, Your city is {js['city']}, and Your Time zone is {js['time_zone']}"
)
elif 'screen' in command:
global degree
screen = rotatescreen.get_primary_display()
if any(i in command for i in ['default', '0', 'stop']):
degree = 0
screen.rotate_to(degree)
talk('Screen set to normal!')
return
screen.rotate_to(degree % 360)
talk('Screen rotated to ' + str(degree))
degree += 90
elif 'translate' in command:
translator = Translator()
talk('What is the source language which you are gonna speak?')
from_lang = get_languagecode(take_command().strip())
print(from_lang)
talk('What is the destination language which needs to be translated?')
to_lang = get_languagecode(take_command().strip())
talk("what message to be translated?")
get_message = take_command()
text_to_translate = translator.translate(get_message,
src=from_lang,
dest=to_lang)
text = text_to_translate.text
speak = gTTS(text=text, lang=to_lang, slow=False)
speak.save("captured_voice.mp3")
music = pyglet.media.load("captured_voice.mp3", streaming=False)
music.play()
time.sleep(music.duration) # prevent from killing
os.remove("captured_voice.mp3") # remove temperory file
elif 'sleep' in command.strip():
talk("Ok sir! I won't disturb you for a minute")
time.sleep(60)
talk()
else:
talk('Please say the command again ! ')
def interpret_location(interpreter, speaker, d, ignore_reldir=False) -> XYZ:
"""Location dict -> coordinates
Side effect: adds mems to agent_memory.recent_entities
if a reference object is interpreted;
and loc to memory
"""
location_type = d.get("location_type", "SPEAKER_LOOK")
if location_type == "SPEAKER_LOOK":
player = interpreter.memory.get_player_struct_by_name(speaker)
loc = capped_line_of_sight(interpreter.agent, player)
elif location_type == "SPEAKER_POS":
loc = pos_to_np(
interpreter.memory.get_player_struct_by_name(speaker).pos)
elif location_type == "AGENT_POS":
loc = pos_to_np(interpreter.agent.get_player().pos)
elif location_type == "COORDINATES":
loc = cast(
XYZ,
tuple(
int(float(w))
for w in re.findall("[-0-9.]+", d["coordinates"])))
if len(loc) != 3:
logging.error("Bad coordinates: {}".format(d["coordinates"]))
raise ErrorWithResponse(
"I don't understand what location you're referring to")
else:
loc, mems = maybe_get_location_memory(interpreter, speaker, d)
if loc is None:
raise ValueError(
"Can't handle Location type: {}".format(location_type))
# handle relative direction
reldir = d.get("relative_direction")
if reldir is not None and not ignore_reldir:
if reldir == "BETWEEN":
pass # loc already handled when getting mems above
if reldir == "INSIDE":
if location_type == "REFERENCE_OBJECT":
mem = mems[0]
locs = perception.find_inside(mem)
if len(locs) == 0:
raise ErrorWithResponse(
"I don't know how to go inside there")
else:
loc = locs[0]
elif reldir == "AWAY":
apos = pos_to_np(interpreter.agent.get_player().pos)
dir_vec = (apos - loc) / np.linalg.norm(apos - loc)
num_steps = word_to_num(d.get("steps", "5"))
loc = num_steps * np.array(dir_vec) + to_block_center(loc)
elif reldir == "NEAR":
pass
else: # LEFT, RIGHT, etc...
reldir_vec = rotation.DIRECTIONS[reldir]
look = interpreter.memory.get_player_struct_by_name(speaker).look
# this should be an inverse transform so we set inverted=True
dir_vec = rotation.transform(reldir_vec,
look.yaw,
0,
inverted=True)
num_steps = word_to_num(d.get("steps", "5"))
loc = num_steps * np.array(dir_vec) + to_block_center(loc)
# if steps without relative direction
elif "steps" in d:
num_steps = word_to_num(d.get("steps", "5"))
loc = to_block_center(loc) + [0, 0, num_steps]
return to_block_pos(loc)
def to_html(self):
t = self.H.table(border='1', id="carddata")
r = t.tr
with open(config.CARDS_AND_DIMENSIONS_CSV) as csvfile:
reader = csv.DictReader(csvfile)
for column in reader.fieldnames:
r.td(column)
for row in reader:
t.tr(id=str(reader.line_num))
for col in row.items():
try:
w2n.word_to_num(col[0])
t.td(
"<img width='100' src='https://deckofcardsapi.com/static/img/"
+ col[1] + ".png'></img>",
escape=False,
id=col[0] + str(reader.line_num),
klass=col[0])
except:
t.td(col[1],
klass=col[0],
id=col[0] + str(reader.line_num))
html_human_play_before = """
<html>
<head>
</head>
<body onload="init()">
<style>
#carddata {display: none;}
#play_area {padding-top: 50;
padding-bottom: 30;
text-decoration: none;
text-transform: uppercase;}
.dimension_name {color: transparent;
text-align: center;}
.dimension_value {color: transparent;
text-align: center;}
.rule_name {color: transparent;
text-align: center;}
.rule_bool {text-transform: uppercase;
color: ffffff;
text-align: center;}
body {background-color: 496D89;}
#play, #reveal, #hide {background-color: #123652;
border: none;
color: white;
padding-top: 20;
padding: 45px 85px;
text-align: center;
text-decoration: none;
display: inline-block;
font-size: 16px;}
</style>
<script>
function init() {
pywebview.api.init()
</script>
"""
html_human_play_after = """
<div id="play_area">
<table id="play_table"></table>
<p id="rule_display"> <p>
</div>
<button type="Button" id="play" name="Play" onclick="toggleTable()">Draw</button>
<button type="Button" id="reveal" name="Reveal" onclick="revealRule()">Reveal rule</button>
<button type="Button" id="hide" name="Hide" onclick="hideRule()">Restart</button>
<script>
function toggleTable() {
var i;
var rows_count= document.getElementById("carddata").rows.length;
i = Math.floor(Math.random() * rows_count-2) + 2;
j=i.toString();
document.getElementById('play_table').innerHTML = document.getElementById(j).innerHTML;
}
</script>
<script>
function revealRule() {
var item_count = document.getElementsByClassName('rule_name').length;
document.getElementsByClassName('rule_bool')[item_count-1].style.color= "transparent";
document.getElementById('rule_display').innerHTML = "True if " +
document.getElementsByClassName('dimension_name')[item_count-1].innerHTML + " (" +
document.getElementsByClassName('dimension_value')[item_count-1].innerHTML + ") " +
document.getElementsByClassName('rule_name')[item_count-1].innerHTML + "." + " " + "This hand is " +
document.getElementsByClassName('rule_bool')[item_count-1].innerHTML;
document.getElementById('rule_display').style.color="ffffff";
document.getElementsByClassName('rule_bool').style.color="green";
}
</script>
<script>
function hideRule() {
document.getElementById('rule_display').style.color="transparent";
toggleTable();
}
</script>
</body>
</html>
"""
html_ai_play_before = """
<script type="text/javascript" src="http://ajax.googleapis.com/ajax/libs/jquery/1.3.0/jquery.min.js"></script>
<script type="text/javascript">
var auto_refresh = setInterval(function () {
$('#play_area').load('#play_area');
}, 1000);
</script>
"""
print(
str(html_human_play_before) + str(t) + str(html_human_play_after))
return str(html_human_play_before) + str(t) + str(
html_human_play_after)
def parseData(self, inputString):
# Default.xlsx : Excel file that contains currency information
currencyDF = pd.DataFrame({
'currency': [
'Dollars', 'Pounds', 'Shillings', 'Cents', 'Pistoles',
'Guineas', 'Pence', 'Texas Dollars'
],
'symbol': ['$', '£', None, None, None, None, None, '$ texas'],
'string': [
'dollar', 'pound', 'shilling', 'cent', 'pistole', 'guineas',
'pence', 'texas dollar'
]
})
unit = ""
amount = ""
# Get amount and currency from the input string : ex: 20$ or £20
inputString = inputString.replace(',', '')
price = Price.fromstring(inputString)
if (price.amount != None and price.currency != None):
amount = price.amount
unit = price.currency
# Get currency data from string : ex: 20 dollars or twenty dollars
if (amount == ""):
try:
amount = w2n.word_to_num(inputString)
except:
try:
amount = int(re.search(r'\d+', inputString).group())
except:
amount = ""
# Converting currency string
updatedUnit = ""
# if (amount != "") :
for index, row in currencyDF.iterrows():
# Checking symbol columns from the Currency Excel File
if (row['symbol'] != "" and pd.isnull(row['symbol']) != True):
symbolCheck = row['symbol'].split(' ')
symbol = symbolCheck[0]
if (len(symbolCheck) != 1):
if (unit == symbolCheck[0]
and symbolCheck[1] in inputString):
return self.retDataFrame(amount, row['currency'])
else:
inputString = inputString.replace(row['symbol'],
row['string'])
if (unit == row['symbol']):
updatedUnit = row['currency']
# Checking string columns from the Currency Excel File
if (self.checkString(inputString.lower(),
row['string'].split(' ')) == True):
stringCheck = row['string'].split(' ')
if (len(stringCheck) == 1):
updatedUnit = row['currency']
else:
return self.retDataFrame(amount, row['currency'])
return self.retDataFrame(amount, updatedUnit)
temp = {}
# title is an attribute of the <a> tag inside <h3> which is inside <article>
# as there is only 1 <article> we use the " . " to get inside it ,same with h3,a. attrs lists the attributes
temp["title"] = item.article.h3.a.attrs["title"]
# price is inside <div> with class "product_price", but since there is more than one <div> we need to use find
# inside that <div> we find a <p> with class "price_color" and use text to get the data and splice off the undesired unicode at the front [1:]
temp["price"] = item.article.find("div", class_="product_price").find(
"p", class_="price_color").text[1:]
# stock is inside the <div class="product_price"> within <p> class "instock availability"
# as we have more than one <div> we use find("div",class_="product_price")
# we use the .text method to extract the text as a string and use strip() to remove white spaces
temp["stock"] = item.article.find("div", class_="product_price").find(
"p", class_="instock availability").text.strip()
# star rating is a word inside the <article> within a <p> with class "star-rating"
# we use the word2number module to make words [One,Two,Three...] --> [1,2,3..]
temp["rating"] = w2n.word_to_num(item.article.p.attrs["class"][1])
# store the dictionary in the list of books
books.append(temp)
# All above paths to the data was found by inspecting the source
# Print all the dictionaries on the page with the details
for i in books:
pprint(i)
print("-------------------------------------------")
result = str(eval(data2))
s.send(result + "\n")
print result
wordnum = False
count = 0
i = 0
while 1:
data = s.recv(BUFFER_SIZE)
try:
newdata = data.split(' ')
num1 = newdata[0] + " " + newdata[1]
int_num1 = w2n.word_to_num(num1)
wordnum = True
except ValueError:
wordnum = False
# do normal things
pass
newdata = data.split(" ")
if (len(newdata[1]) < 2):
if not data: break
print(data)
#s.close()
data2 = data[0:10:1]
data2 = data2[:-3]
def fix_age(matchobj):
#print matchobj.group(0)
return str(w2n.word_to_num(matchobj.group(0))) + " "
def join_time_description(amr):
# 4 o'clock; 4 am; 4 a.m., etc.
while True:
span = None
if len(amr.tokens) < 2:
break
for i in range(1, len(amr.tokens)):
x, y = amr.tokens[i - 1: i + 1]
if y.lower() in ("o'clock", 'am', 'a.m.', 'pm', 'p.m') and re.search(r'^\d+[.:]?\d*[.:]?\d*$', x):
span = list(range(i - 1, i + 1))
joined_tokens = ''.join([x, y])
pos = 'CD'
ner = 'TIME'
break
if y.lower() in ("o'clock", 'am', 'a.m.', 'pm', 'p.m') and x.isalpha():
try:
x = w2n.word_to_num(x)
except:
continue
x = str(x)
span = list(range(i - 1, i + 1))
joined_tokens = ''.join([x, y])
pos = 'CD'
ner = 'TIME'
break
if y == 'Greenwich' and i + 2 < len(amr.tokens) and amr.tokens[i + 1: i + 3] == ['Mean', 'Time']:
span = list(range(i, i + 3))
joined_tokens = 'GMT'
pos = 'NNP'
ner = 'TIME'
break
if y in ('century', 'Century'):
m = re.search(r'^(\d+)(st|nd|rd|th)?$', x)
if m and m.group(1) != '':
span = list(range(i - 1, i + 1))
joined_tokens = ''.join([m.group(1), y.lower()])
pos = 'CD'
ner = 'TIME'
break
elif x == 'first' and amr.tokens[i - 2] == '-' and amr.tokens[i - 3] == 'twenty':
span = list(range(i - 3, i + 1))
joined_tokens = '21century'
pos = 'CD'
ner = 'TIME'
break
elif x.lower() == 'eighth':
span = list(range(i - 1, i + 1))
joined_tokens = '8century'
pos = 'CD'
ner = 'TIME'
break
elif x.lower() == 'fifth':
span = list(range(i - 1, i + 1))
joined_tokens = '5century'
pos = 'CD'
ner = 'TIME'
break
else:
try:
x = w2n.word_to_num(x)
except:
continue
span = list(range(i - 1, i + 1))
joined_tokens = ''.join([x, y.lower()])
pos = 'CD'
ner = 'TIME'
break
else:
break
amr.replace_span(span, [joined_tokens], [pos], [ner])
import pandas as pd
from sklearn import linear_model
from word2number import w2n
import math
data = pd.read_csv('datasets/hiring.csv')
data.experience[0] = 'zero'
data.experience[1] = 'zero'
a=0
for i in data.experience:
data.experience[a] = w2n.word_to_num(i)
a=a+1
data.rename(columns={'test_score(out of 10)':'test_score','interview_score(out of 10)':'interview_score','salary($)':'salary'},inplace=True)
median = math.floor(data.test_score.median())
data.test_score.fillna(median,inplace=True)
print(data)
reg = linear_model.LinearRegression()
reg.fit(data[['experience','test_score','interview_score']],data.salary)
print(reg.coef_)
print(reg.intercept_)
print(reg.predict([[2,9,6]]))
print(reg.predict([[12,10,10]]))
def update_boundary():
with open(os.path.join(data_dir, "bookcorpus", "clean_split.json"),
'r',
encoding='utf-8') as infile:
book_info_dict = json.load(infile)
# check all books
bookcorpus_dir = os.path.join(data_dir, "bookcorpus", "segment")
counter = Counter()
last_counter = Counter()
book_info = []
for phase, book_info_list in book_info_dict.items():
for count, book_info in enumerate(book_info_list):
book_name = book_info["book"]
print("\x1b[2K\rRemoving Header in {}, {:>5} / {:>5} [{:.2f}%]".
format(phase, count, len(book_info_list),
100.0 * count / len(book_info_list)),
end="")
if not os.path.isfile(
os.path.join(data_dir, "bookcorpus", "frame", book_name)):
continue
with open(os.path.join(bookcorpus_dir, book_name),
'r',
encoding='utf-8') as infile:
lines = infile.read().split("\n")
# find chapter one
chapter_one = -1
for i, line in enumerate(lines):
if line[:7].lower() == "chapter":
tokens = word_tokenize(line)
if len(tokens) < 2:
continue
# ignore table of content
if sum([
1 for token in tokens
if token.lower() == "chapter"
]) > 1:
continue
# find chapter one
try:
chapter_num = w2n.word_to_num(tokens[1])
except ValueError:
chapter_num = None
if chapter_num == 1:
counter.update([line])
chapter_one = i
break
# cannot find chapter 1
if chapter_one == -1:
continue
# remove the wired cases where chapter 1 come after more than 300 lines
if chapter_one > 300:
continue
# find last chapter
chapter_info = []
for i, line in enumerate(lines):
if line[:7].lower() == "chapter":
tokens = word_tokenize(line)
if len(tokens) < 2:
continue
# ignore table of content
if sum([
1 for token in tokens
if token.lower() == "chapter"
]) > 1:
continue
try:
chapter_num = w2n.word_to_num(tokens[1])
except ValueError:
chapter_num = None
if chapter_num is not None:
chapter_info.append([chapter_num, i, line])
if chapter_info[-1][0] > 1:
start = chapter_one
end = chapter_info[-1][1]
last_counter.update([chapter_info[-1][2]])
book_info["start"] = start
book_info["end"] = end
# save data
with open(os.path.join(data_dir, "bookcorpus", "clean_split_updated.json"),
'w',
encoding='utf-8') as outfile:
json.dump(book_info_dict, outfile, indent=2)
f = open("num.txt", "r")
data = f.readlines()
datanew = list(map(str.rstrip, data))
f.close()
#If the phrase"and" is not located in the check
if "AND" not in amtupper:
amtupper += " AND ZERO"
#Divides the dolalr amount and the cent amount
dollar, cents = amtupper.split(" AND ")
wordlst = dollar.split(' ')
#Checks if the words in dollar is in the num.txt dictionary
try:
for word in wordlst:
if word in datanew:
wordnum += word + ' '
numamt = w2n.word_to_num(wordnum)
except ValueError:
#returns an "Un-Readable statement if none of the words are in the num.txt dictionary"
amt += "(Un-Readable) "
amtupper += "(Un-Readable) "
numamt = 0
#turns the cents into a decimal if the cents are written as "##/100"
try:
num, den = cents.split('/')
result = (float(num) / float(den))
numamt += result
except ValueError:
#turns the cents into a decimal if the cents are written as words
try:
centsplit = cents.split(' ')
for word in centsplit:
# Main
while True:
# User input
i = input('user : '******'break', 'end', 'quit']:
break
else:
try:
#case 0 : a Number entered
try:
o = (w2n.word_to_num(i))
except:
try:
#case 1 : Normal calculations (basic calculator)
'''Easy way was to use eval, but using eval has great risks as it could be used to potentially crack your system'''
o = simple_eval(i)
except:
try:
# clean sentence
words = [
k for k in word_tokenize(i) if k not in stop_words
]
o = words
for i in words:
if i in [
from word2number import w2n
### importing csv file
df = pd.read_csv('hiring.csv')
print(df)
### cleaning the dataframe
df['experience'].fillna(0, inplace=True)
median = df['test_score(out of 10)'].median()
df['test_score(out of 10)'].fillna(median, inplace=True)
print(df)
### converting number words to numbers
print()
num_list = []
for i in df['experience'].iloc[2:]:
i = w2n.word_to_num(i)
num_list.append(i)
df.loc[2:, ('experience')] = num_list
print(df['experience'].iloc[2:])
print(df)
#### creating linear regression model
lreg = linear_model.LinearRegression()
lreg.fit(
df[['experience', 'test_score(out of 10)', 'interview_score(out of 10)']],
df['salary($)'])
print(lreg)
### printing the coefficients
print(lreg.coef_)
def analysis(originalquery,finalquery):
fromcity = None
tocity = None
par3 = (time.strftime("%Y-%m-%d"))
par4 = None
time_day = None
time_relative = None
time_nu = None
time_period = None
time_tom = None
time_spec_month = None
time_spec_date = None
for i in range(0,len(finalquery) ):
if finalquery[i] == 'B-fromloc.city_name':
fromcity = originalquery[i]
try:
if finalquery[i+1] == 'I-fromloc.city_name':
fromcity = fromcity + ' ' + originalquery[i+1]
except IndexError:
pass
if finalquery[i] == 'B-toloc.city_name':
tocity = originalquery[i]
try:
if finalquery[i+1] == 'I-toloc.city_name':
tocity = tocity + ' ' + originalquery[i+1]
except IndexError:
pass
#print fromcity,tocity
if finalquery[i] =='B-depart_date.day_name':
time_day = originalquery[i]
if finalquery[i] == 'B-depart_time.time_relative':
time_relative = originalquery[i]
if finalquery[i] == 'B-depart_time.time':
time_nu = originalquery[i]
try:
if finalquery[i+1] == 'I-depart_time.time':
time_nu = time_nu + ' ' + originalquery[i+1]
except IndexError:
pass
try:
time_nu = time_nu + ' ' + originalquery[i+1]
except IndexError:
pass
if finalquery[i] == 'B-arrive_time.period_of_day':
time_period = originalquery[i]
try:
if finalquery[i+1] == 'I-arrive_time.period_of_day':
time_period = time_period + ' ' + originalquery[i+1]
except IndexError:
pass
try:
time_period = time_period + ' ' + originalquery[i+1]
except IndexError:
pass
if finalquery[i] =='B-depart_date.month_name':
time_spec_month = originalquery[i]
time_spec_month = month_converter(time_spec_month)
if finalquery[i] == 'B-depart_date.day_number':
time_spec_date = originalquery[i]
time_spec_date = w2n.word_to_num(time_spec_date)
if finalquery[i] == 'B-depart_date.today_relative':
time_tom = originalquery[i]
if time_tom is not None:
if time_tom == 'tomorrow':
date_tom = datetime.now()+ relativedelta(days=1)
par3 = date_tom.strftime('%Y-%m-%d')
if time_day is not None:
x = 0
if time_day.lower() == 'monday':
time_day = 0
elif time_day.lower() =='tuesday':
time_day = 1
elif time_day.lower() =='wednesday':
time_day = 2
elif time_day.lower() =='thursday':
time_day = 3
elif time_day.lower() =='friday':
time_day = 4
elif time_day.lower() =='saturday':
time_day = 5
elif time_day.lower() =='sunday':
time_day = 6
x = int(datetime.today().weekday()) - time_day
date_after = datetime.now()+ relativedelta(days=int(x))
print date_after
par3 = date_after.strftime('%Y-%m-%d')
if time_spec_month is not None:
if (str(time_spec_month))==1:
time_spec_month = '0'+str(time_spec_month)
par3 = '2016-'+str(time_spec_month)+'-'+str(time_spec_date)
print par3
flight_data = makeparameters(fromcity,tocity,par3,par4)
return flight_data,time_period,time_relative,time_nu,tocity