def nlp(self, text):
doc = self.nlp_model(text)
to = []
when = []
body = []
persons = []
locked_ner_model = self.model_pool.get_shared_model("xx_ent_wiki_sm")
with locked_ner_model:
persons = ner.get_persons(locked_ner_model.acquire_model(), text)
for token in doc:
if token.dep_ == "TO":
to.append(token.text)
elif token.dep_ == "START":
when.append(token.text)
elif token.dep_ == "BODY":
body.append(token.text)
log.debug("%s %s", token.text, token.dep_)
time = datetime.now() + timedelta(seconds=5)
if len(when) > 0:
time = tc.parse_time(when)
to = ner.cross_check_names(to, persons)
log.debug("Recipients: %s", ",".join(to))
_body = " ".join(body)
return (to, time, _body)
def nlp(self, text):
doc = self.nlp_model(text)
to = []
start = []
end = []
body = []
persons = []
locked_ner_model = self.model_pool.get_shared_model("xx_ent_wiki_sm")
with locked_ner_model:
persons = ner.get_persons(locked_ner_model.acquire_model(), text)
for token in doc:
if token.dep_ == "TO":
to.append(token.text)
elif token.dep_ == "START":
start.append(token.text)
elif token.dep_ == "END":
end.append(token.text)
elif token.dep_ == "BODY":
body.append(token.text)
log.debug("%s %s", token.text, token.dep_)
to = ner.cross_check_names(to, persons)
log.debug("Recipients: %s", ",".join(to))
timespan = self.timespan(start, end)
_body = " ".join(body)
return (to, timespan, _body)
def test_same_input_same_output(self):
"""
Test that if tags and persons are the same, cross
checking should also return the same output.
"""
tags = ["mark", "evert", "reggie"]
cc = ner.cross_check_names(tags, tags)
assert cc == tags
def test_full_name_returned(self):
"""
Test that the tags and corresponding full name
returns the full name when cross checking.
"""
tags = ["mark", "anderson"]
persons = ["mark anderson"]
cc = ner.cross_check_names(tags, persons)
assert persons == cc
def test_missing_tag(self):
"""
Test that when names are missing in persons that are found in
tags, the cross check should append the missing tags to persons.
"""
tags = ["mark", "anderson", "evert"]
persons = ["mark anderson"]
expected = ["mark anderson", "evert"]
cc = ner.cross_check_names(tags, persons)
assert cc == expected
def nlp(self, text):
doc = self.nlp_model(text)
to = []
start = []
end = []
body = []
persons = []
locked_ner_model = self.model_pool.get_shared_model("xx_ent_wiki_sm")
with locked_ner_model:
persons = ner.get_persons(locked_ner_model.acquire_model(), text)
for token in doc:
if token.dep_ == "TO":
to.append(token.text)
elif token.dep_ == "START":
start.append(token.text)
elif token.dep_ == "END":
end.append(token.text)
elif token.dep_ == "BODY":
body.append(token.text)
log.debug("%s %s", token.text, token.dep_)
to = ner.cross_check_names(to, persons)
log.debug("Recipients: %s", ",".join(to))
start_time = datetime.now()
if len(start) == 0:
start_time = start_time + timedelta(seconds=5)
else:
start_time = tc.parse_time(start)
end_time = 0
if len(end) == 0:
end_time = start_time + timedelta(
minutes=self.get_meeting_duration())
else:
end_time = tc.parse_time(end)
_body = " ".join(body)
return (to, {"start": start_time, "end": end_time}, _body)
def nlp(self, text):
doc = self.nlp_model(text)
to = []
start = []
body = []
persons = []
nstart = []
locked_ner_model = self.model_pool.get_shared_model("xx_ent_wiki_sm")
with locked_ner_model:
persons = ner.get_persons(locked_ner_model.acquire_model(), text)
for token in doc:
if token.dep_ == "TO":
to.append(token.text)
elif token.dep_ == "START":
start.append(token.text)
elif token.dep_ == "BODY":
body.append(token.text)
elif token.dep_ == "NSTART":
nstart.append(token.text)
log.debug("%s %s", token.text, token.dep_)
to = ner.cross_check_names(to, persons)
log.debug("Recipients: %s", ",".join(to))
start_time = datetime.now()
if len(start) == 0:
start_time = start_time + timedelta(seconds=5)
else:
start_time = tc.parse_time(start)
new_start_time = datetime.now()
if len(nstart) == 0:
new_start_time = new_start_time + timedelta(seconds=5)
else:
new_start_time = tc.parse_time(nstart)
_body = {"summary": " ".join(body)}
_body["new start"] = new_start_time
return (to, start_time, _body)
