def set_value_from_selection(self, doc: Document, value: str):
if self.field_type == FieldType.FIELD_TYPE_CONCRETE_STRING:
doc.__setattr__(self.field, value)
elif self.field_type == FieldType.FIELD_TYPE_CONCRETE_FLOAT:
try:
doc.__setattr__(self.field, float(value))
except ValueError:
nums = list(extractors.find_numbers(value)) if value else None
doc.__setattr__(self.field, nums[0] if nums else None)
elif self.field_type == FieldType.FIELD_TYPE_CONCRETE_INTEGER:
try:
doc.__setattr__(self.field, int(value))
except ValueError:
nums = list(extractors.find_numbers(value)) if value else None
doc.__setattr__(self.field, nums[0] if nums else None)
elif self.field_type == FieldType.FIELD_TYPE_CONCRETE_DATE:
d = dateparser.parse(value) if value else None
if d:
doc.__setattr__(self.field, d)
else:
dates = list(get_dates(value)) if value else None
doc.__setattr__(self.field, dates[0] if dates else None)
return doc.__getattribute__(self.field)
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_citations(text):
"""
Get citations containing "BGBl"
:param text: str
:return: yields dict
"""
for ptn in [CITATION_PTN_RE, SECOND_CITATION_PTN_RE, CITATION_RANGE_PTN_RE]:
for match in ptn.finditer(text):
capture = match.capturesdict()
date = ''.join(capture.get('date', ''))
if date:
try:
date = str(list(get_dates(date, 'de'))[0]['value'])
except Exception as e:
pass
yield dict(
location_start=match.start(),
location_end=match.end(),
text=capture['text'][0],
article=''.join(capture.get('article', '')),
number=''.join(capture.get('number', '')),
subparagraph=''.join(capture.get('subparagraph', '')),
sentence=''.join(capture.get('sentence', '')),
paragraph=''.join(capture.get('paragraph', '')),
letter=''.join(capture.get('letter', '')),
date=date,
part=capture['part'][0],
page=', '.join(capture['page']),
year=', '.join(capture.get('year', ''))
)
def get_citation_annotations(cls, text: str) -> \
Generator[CitationAnnotation, None, None]:
"""
Get citations containing "BGBl"
:param text: str
:return: yields dict
"""
for ptn in [
cls.CITATION_PTN_RE, cls.SECOND_CITATION_PTN_RE,
cls.CITATION_RANGE_PTN_RE
]:
for match in ptn.finditer(text):
capture = match.capturesdict()
date = ''.join(capture.get('date', ''))
if date:
try:
date = str(list(get_dates(date, 'de'))[0]['value'])
except:
pass
ant = CitationAnnotation(
coords=match.span(),
text=capture['text'][0],
paragraph=''.join(capture.get('paragraph', '')),
subparagraph=''.join(capture.get('subparagraph', '')),
letter=''.join(capture.get('letter', '')),
date=date,
part=capture['part'][0],
locale='de')
ant.article = TextAnnotation.get_int_value(''.join(
capture.get('article', '')))
ant.number = TextAnnotation.get_int_value(''.join(
capture.get('number', '')))
ant.sentence = TextAnnotation.get_int_value(''.join(
capture.get('sentence', '')))
page_range = ', '.join(capture['page'])
page = TextAnnotation.get_int_value(page_range)
if page:
ant.page = page
else:
ant.page_range = page_range
volume_str = ''.join(capture.get('number', ''))
volume = TextAnnotation.get_int_value(volume_str)
if volume:
ant.volume = volume
else:
ant.volume_str = volume_str
year_str = ', '.join(capture.get('year', ''))
year = TextAnnotation.get_int_value(year_str)
if year:
ant.year = year
else:
ant.year_str = year_str
yield ant
def get_effective_date(text, return_source=False):
# need a better more accurate way of doing this
# right now looks for triggers and takes latest date in that sentence
TRIGGER_LIST_START_DATE = ["dated as of", "effective as of", "made as of", "entered into as of"]
found_start_date_trigger = False
effective_date = None
for t in TRIGGER_LIST_START_DATE:
if findWholeWordorPhrase(t)(text) is not None:
found_start_date_trigger = True
break
if found_start_date_trigger:
dates = list(get_dates(text))
if len(dates) > 0:
effective_date = max(dates)
if return_source:
return effective_date, text
else:
return effective_date
'permitted_use': _cleanup_sentence(match).strip()}),
FieldDetector(select=r'(?:tenant|lessee)\s+(?:may|shall)\s+use.*(?:property|premises)(.*)',
fill_fields=lambda sentence, match: {
'permitted_use': _cleanup_sentence(match).strip()}),
FieldDetector(select=re.compile(r'Use.*[:.]\s+(.*)', re.DOTALL),
fill_fields=lambda sentence, match: {
'permitted_use': _cleanup_sentence(match).strip()}),
FieldDetector(select=re.compile(r'Permitted\s+[Uu]se.*[:.]\s+(.*)', re.DOTALL),
fill_fields=lambda sentence, match: {
'permitted_use': _cleanup_sentence(match).strip()})
],
'start_end_term': [
FieldDetector(select=r'shall\s+(?:commence|start).*\d.*',
process_selected=lambda sentence, match: get_dates(match),
fill_fields=lambda sentence, dates: {'commencement_date': dates[0]}),
FieldDetector(select=r'shall\s+(?:end).*\d.*',
process_selected=lambda sentence, match: get_dates(match),
fill_fields=lambda sentence, dates: {'expiration_date': dates[0]}),
FieldDetector(select=r'from.*\d.*',
process_selected=lambda sentence, match: get_dates(match),
fill_fields=lambda sentence, dates: {'commencement_date': dates[0]}),
FieldDetector(select=r'commencement\s+date.*\d.*',
process_selected=lambda sentence, match: get_dates(match),
fill_fields=lambda sentence, dates: {'commencement_date': dates[0]}),
FieldDetector(select=r'expiration\s+date.*\d.*',
process_selected=lambda sentence, match: get_dates(match),
fill_fields=lambda sentence, dates: {'expiration_date': dates[0]}),
FieldDetector(select=r'term\W.*\d.*',
process_selected=lambda sentence, match: get_durations(match),
def extract_dates(self, text=None):
if not text:
text = self.text
return list(lex_dates.get_dates(text))
# path = sys.argv[1]
# copy&paste brief pdf into .txt file.
direct_path = ""
with open(direct_path, 'r') as file:
brief = file.read()
processed_brief = lex_sentences.pre_process_document(brief)
sentences_brief = lex_sentences.get_sentence_list(processed_brief)
#print(sentences_brief)
facts = []
for sentence in sentences_brief:
dates = lex_dates.get_dates(sentence)
for date in dates:
facts.append((date,sentence))
for fact in facts:
print("Question:\nWhy is {} significant?\n\nAnswer:\n{}".format(str(fact[0]), fact[1]))
print("\n---------------\n")
'''
Question:
Why is 2018-11-26 significant?
Answer:
Case: 18-60522 Document: 00514736148 Page: 17 Date Filed: 11/26/2018
that these employees were constructively discharged because they faced a Hobson’s Choice?
---------------
def getDate(self):
mem = []
dates = list(get_dates(self.bill_text))
for date in dates:
mem.append(str(date))
self.bill.info['dates'] = mem