def are_there_more_left_parentheses_than_right_parentheses(location, hits):
last_sentence_fragment = lfp.wordtokencreation.get_last_sentence_fragment(location, hits, return_as_string=True)
char_frequency = Utilities.character_counter(last_sentence_fragment, '(', ')')
if char_frequency['('] > char_frequency[')']:
return True
return False
def was_cut_within_a_table(location, hits):
last_sentence_fragment = lfp.wordtokencreation.get_last_sentence_fragment(location, hits)
if last_sentence_fragment is None:
return False
compressed_sentence_fragment = lfp.wordtokencreation.get_last_sentence_fragment(location, hits, return_as_string=True)
#print "FRAGMENT:", compressed_sentence_fragment
# see whether we picked up a table.
# tables normally have units of currency as well as the word follows somewhere.
# if these hold, then we're probably in a table from a previous section.
# that means that if we're in a relevant section right now, and the new hit demarcates a new section,
# then we want to stop recording. if we're not recording, then we probably want to start.
# if we're in a relevant section and the new hit does *not* have a header that's been whitelisted as being
# a section, then we can continue recording.
if re.search("(in)?\s*(millions|thousands|billions)", compressed_sentence_fragment, re.I | re.M | re.S) \
and re.search("total|follow(s|ing)|balance", compressed_sentence_fragment, re.I | re.M | re.S):
#print "MATCH ON currency"
#print 'MATCH ON FOLLOWS|total'
return True
char_frequency = Utilities.character_counter(compressed_sentence_fragment, '$')
if char_frequency['$'] >= 6:
#print 'MATCH ON DOLLAR COUNT'
return True
number_count = 0
for word in last_sentence_fragment:
if Utilities.contains_numbers(word):
number_count += 1
if re.search("total|follow(s|ing)|balance", compressed_sentence_fragment, re.I | re.M | re.S) \
and number_count >= 6:
#print "match on number count"
return True
return False
