def en_parsers_speed(self):
file_path = os.path.join(lexnlp_test_path, 'long_parsed_text.txt')
with codecs.open(file_path, 'r', encoding='utf-8') as fr:
text = fr.read()
ge_path = os.path.join(lexnlp_test_path, 'lexnlp/extract/en/tests/test_geoentities/')
entities_fn = ge_path + 'geoentities.csv'
aliases_fn = ge_path + 'geoaliases.csv'
geo_config = list(DictionaryEntry.load_entities_from_files(entities_fn, aliases_fn))
times = {} # type: Dict[str, float]
self.check_time(text, lambda s: list(get_amounts(s)), 'get_amounts', times)
self.check_time(text, lambda s: list(get_acts(s)), 'get_acts', times)
self.check_time(text, lambda s: list(get_citations(s)), 'get_citations', times)
self.check_time(text, lambda s: list(get_conditions(s)), 'get_conditions', times)
self.check_time(text, lambda s: list(get_constraints(s)), 'get_constraints', times)
self.check_time(text, lambda s: list(get_copyright(s)), 'get_copyright', times)
self.check_time(text, lambda s: list(_get_courts(s)), 'get_courts', times)
self.check_time(text, lambda s: list(get_cusip(s)), 'get_cusip', times)
self.check_time(text, lambda s: list(get_dates(s)), 'get_dates', times)
self.check_time(text, lambda s: list(get_definitions(s)), 'get_definitions', times)
self.check_time(text, lambda s: list(get_distances(s)), 'get_distances', times)
self.check_time(text, lambda s: list(get_durations(s)), 'get_durations', times)
self.check_time(text, lambda s: list(get_geoentities(s, geo_config)), 'get_geoentities', times)
self.check_time(text, lambda s: list(get_money(s)), 'get_money', times)
self.check_time(text, lambda s: list(get_percents(s)), 'get_percents', times)
self.check_time(text, lambda s: list(get_pii(s)), 'get_pii', times)
self.check_time(text, lambda s: list(get_ratios(s)), 'get_ratios', times)
self.check_time(text, lambda s: list(get_regulations(s)), 'get_regulations', times)
self.check_time(text, lambda s: list(get_trademarks(s)), 'get_trademarks', times)
self.check_time(text, lambda s: list(get_urls(s)), 'get_urls', times)
self.assertTrue('get_amounts' in times)
def get_salary(text, return_source=False):
TRIGGER_LIST_SALARY = ["salary", "rate of pay"]
# text to be found and multiplier to get yearly
found_time_unit = None
found_time_units = []
found_salary_trigger = False
money = None
min_annual_salary = 20000 # sample is mostly executives- so this is safe.
for t in TRIGGER_LIST_SALARY:
if findWholeWordorPhrase(t)(text) is not None:
found_salary_trigger = True
break
if found_salary_trigger:
for t in TRIGGER_LIST_TIME_UNIT:
found_time_unit_temp = findWholeWordorPhrase(t[0])(text)
if found_time_unit_temp is not None:
found_time_units.append(t[1])
if len(found_time_units) > 0:
found_time_unit = min(found_time_units)
found_money = list(get_money(text))
if len(found_money) > 0:
money_temp = max(found_money, key=lambda item: item[0])
if money_temp[0] * found_time_unit > min_annual_salary:
money = money_temp
if money is not None:
if return_source:
return money, found_time_unit, text
else:
return money, found_time_unit
else:
return None
def test_get_money_problem1(self):
"""
Problem: it was returning 23.6 instead of 23.62 for such cases.
:return:
"""
text = '''Exercise Price per Share: 23.62'''
actual = list(get_money(text, return_sources=False, float_digits=6))
self.assertEqual(actual[0][0], 23.62)
def test_money(self):
text = "100 bucks, 100 dollars, 100 greens"
ds = list(get_money(text))
self.assertEqual(1, len(ds))
ants = list(get_money_annotations(text))
self.assertEqual(1, len(ds))
self.assertEqual('en', ants[0].locale)
self.assertEqual('USD', ants[0].currency)
self.assertEqual(100.0, ants[0].amount)
def test_get_money_order(self):
"""
At some moment there was a problem: get_money() was returning money in reversed order.
This test is ensures the order is straight.
:return:
"""
text = ''' $96,844.00 per month ($31.00 per square foot per year), beginning on the date which is 90 days after
the Commencement Date and ending on the Expiration Date.'''
actual = list(get_money(text, return_sources=False, float_digits=6))
self.assertEqual(actual[0][0], 96844.0)
def parse(self, text, text_unit_id, _text_unit_lang,
**kwargs) -> ParseResults:
found = list(money.get_money(text, return_sources=True))
if found:
unique = set(found)
return ParseResults({
CurrencyUsage: [
CurrencyUsage(text_unit_id=text_unit_id,
amount=item[0],
amount_str=item[2],
currency=item[1],
count=found.count(item)) for item in unique
]
})
def _extract_variants_from_text(self, field, text: str, **kwargs):
money = get_money(text, return_sources=False)
if not money:
return None
return [{'currency': m[1], 'amount': m[0]} for m in money]
def getMoney(self):
mem = []
money = list(get_money(self.bill_text))
for mon in money:
mem.append(str(mon[0]))
self.bill.info['money'] = mem
rep_date_list = []
for elem in elems:
date_lim = elem[1]
if (date_lim[1] - date_lim[0]) <= 6:
continue
rep_text = text[date_lim[0]:date_lim[1]]
rep_date_list.append(rep_text)
for i in rep_date_list:
text = text.replace(i, ' <DATE> ')
text = re.sub(dates1, ' <DATE> ', text)
start_time = time.time()
rep_money_list = set()
elems = (list(money.get_money(text, return_sources=True)))
print("Money")
print(elems)
for elem in elems:
rep_money_list.add(elem[-1])
rep_money_list = list(rep_money_list)
for i in rep_money_list:
text = text.replace(i, ' <MON> ')
print(time.time() - start_time)
# rep_amt_list=[]
# elems=list(amounts.get_amounts(text, return_sources=True))
# print(elems)
