def word_to_phonemes(word):
result = cmudict_dict.get(word.lower(), None)
if result is None:
# TODO I don't really like this. Should at least return None.
return []
else:
return first(result)
def split_multiclause(sentence, tagged_sentence):
# extract the text the divider tag represents
divider = first(find_first(is_divider, tagged_sentence))
if divider is not None:
first_clause = sentence[0:sentence.index(divider)].rstrip()
second_clause = sentence[sentence.index(divider) + 1:].lstrip()
return [first_clause, second_clause]
else:
return [sentence]
def word_to_phonemes(word):
result = cmudict_dict.get(word.lower(), None)
if LANGUAGE == "german":
print("This most likely generates rubbish. There is no german phoneme database implemented")
if result is None:
# TODO I don't really like this. Should at least return None.
return []
else:
return first(result)
def split_multiclause(sentence, tagged_sentence):
# extract the text the divider tag represents
divider = first(find_first(is_divider, tagged_sentence))
if divider is not None:
first_clause = sentence[0:sentence.index(divider)].rstrip()
second_clause = sentence[sentence.index(divider)+1:].lstrip()
return [first_clause, second_clause]
else:
return [sentence]
def word_to_phonemes(word):
result = cmudict_dict.get(word.lower(), None)
if LANGUAGE == "german":
print(
"This most likely generates rubbish. There is no german phoneme database implemented"
)
if result is None:
# TODO I don't really like this. Should at least return None.
return []
else:
return first(result)
def extract_rule(conn, corpus_id, letter_sound_map, raw_pair):
rule_key = first(raw_pair)
value = second(raw_pair)
rule = None
if rule_key == 'rhyme': rule = dogma.RhymeRule(letter_sound_map.get(value))
elif rule_key == 'blank': rule = dogma.BlankRule()
elif rule_key == 'alliteration': rule = dogma.AlliterationRule(value)
elif rule_key == 'keyword': rule = dogma.KeywordRule(value, conn, corpus_id)
elif rule_key == 'fuzzy': rule = dogma.FuzzyKeywordRule(value, conn, corpus_id)
elif rule_key == 'syllables': rule = dogma.SyllableCountRule(value)
return rule
def source_rm(self):
# ugh i know; TODO figure out how to do this with ORM
conn = self.engine.connect()
session = get_session(self.db)
unlink_sql = """
delete from corpora_sources
where source_id = :source_id
"""
source_delete_sql = """
delete from sources
where name = :source_name
"""
phrase_delete_sql = """
delete from phrases
where source_id = :source_id
"""
name = self.args.source_name
source_id = first(session.query(Source.id).filter(Source.name == name).one())
conn.execute(text(unlink_sql).params(source_id=source_id))
conn.execute(text(phrase_delete_sql).params(source_id=source_id))
conn.execute(text(source_delete_sql).params(source_name=name))
def source_ls(self):
session = get_session(self.db)
for name in session.query(Source.name):
print(first(name))
def corpus_ls(self):
session = get_session(self.db)
for name in session.query(Corpus.name):
print(first(name))
def first_consonant_sound(word):
phonemes = word_to_phonemes(word)
if not is_empty(phonemes):
return find_first(is_consonant_phoneme, phonemes)
else:
return first(consonant_re.findall(word.upper()))