def extract_command(string: str):
l = enums.DependencyEdge.Label
tokens = process(string)
# Find the command's verb
line_verbs = get_token_by_pos(tokens, enums.PartOfSpeech.Tag.VERB)
candidate_verbs = [
v for v in line_verbs if v.text.content.lower() in known_verbs
]
if len(candidate_verbs) == 0:
print("ERROR: No relevant verbs found")
return None
verb = candidate_verbs[0]
def get_possessive(token):
x = token.get_dependant(l.PREP)
y = x.get_dependant(l.POBJ)
z = token.get_dependant(l.DOBJ)
return y or x or z
def get_num(token):
return token.get_dependant(l.NUM)
def get_context(token):
return token.get_dependant(l.AMOD)
pos = get_possessive(verb)
print("NUMBER", get_num(pos))
print("CONTEXT", get_context(pos))
print("VERB", verb)
print("TARGET", pos)
def analyze(data):
query = nlp.process(data)
response = {}
if(query['type'] == "wiki"):
response = encyclopedia(query['subject'])
if(query['type'] == "calc"):
response = calculator.main(query['subject'])
if(query['type'] == "error"):
response = query
return response
def process(input, output):
"""Convert a text file to a document.
Parameters
----------
input : str
Filepath for the input text file.
output : str, optional
Filepath for the output document database. If none provided, defaults to an in-memory database.
See Also
--------
`nlp.process` - the core functionality of this command-line procedure.
"""
import nlp
# connect to the output db (defaults to in-memory)
with orm.Connection(output) as mapping:
# read the provided doc and process all the lines
logger.info(f"Processing file {input}...")
with open(input, "r") as f:
text = f.read()
nlp.process(text, mapping)
logger.info(f"Processing of file {input} complete.")
def to_presses(cls, command):
tokens = process(command)
render_tokens(tokens)
return extract_command(tokens)
from nlp import process, render_tokens, to_ir
from voicerec import voice_command_generator
for command in voice_command_generator('hey google'):
tokens = process(command)
render_tokens(tokens)
print(to_ir(tokens))
def note_to_smartnote(note):
smart_note = nlp.process(note)
print(smart_note)
return smart_note
def view(id):
#received_text2, new_received, number_of_tokens, blob_sentiment, blob_subjectivity, summary, final_time, len_of_words, pos, neg, training, test=process(id)
received_text2, new_received, number_of_tokens, blob_sentiment, blob_subjectivity, summary, final_time, len_of_words, markstop = process(id)
# print(pos)
# print(neg)
# print(training)
return render_template('dataset_view.html', received_text=received_text2, new_received=new_received, number_of_tokens=number_of_tokens,
blob_sentiment=blob_sentiment, blob_subjectivity=blob_subjectivity, summary=summary,
final_time=final_time, len=len_of_words, dt=dt, id=id, df=df, markstop=markstop)
data_dir = 'clue/data'
statements_dir = 'statements'
person_data = {}
statement_data = {}
with open(os.path.join(data_dir, 'persons.json')) as f:
for person in json.load(f)['items']:
person_data[person['fullName'].lower().replace(' ', '-')] = person
persons_statements = defaultdict(list)
statements_persons = defaultdict(list)
for fn in glob(os.path.join(statements_dir, 'statement-*.txt')):
statement_key = os.path.basename(fn).replace('.txt', '')
doc = process(fn)
doc.user_data['title'] = statement_key
statement_data[statement_key] = doc
person_ents = set([ent for ent in doc.ents if ent.label_ == 'PERSON'])
for ent in person_ents:
person_key = ent.text.lower().replace(' ', '-')
person = person_data.get(person_key, None)
if person:
persons_statements[person_key].append(statement_key)
statements_persons[statement_key].append(person_key)
person_data = {
id: person