def test_segments():
text = 'achi, acho!'
expected1 = 'ʌtʃɪ ʌtʃʊ'
expected2 = 'ʌtʃɪ, ʌtʃʊ!'
expected3 = 'ʌtʃɪ ʌtʃʊ '
expected4 = 'ʌtʃɪ, ʌtʃʊ!'
out1 = SegmentsBackend('cree',
preserve_punctuation=False).phonemize([text],
strip=True)[0]
assert out1 == expected1
out2 = SegmentsBackend('cree',
preserve_punctuation=True).phonemize([text],
strip=True)[0]
assert out2 == expected2
out3 = SegmentsBackend('cree', preserve_punctuation=False).phonemize(
[text], strip=False)[0]
assert out3 == expected3
out4 = SegmentsBackend('cree',
preserve_punctuation=True).phonemize([text],
strip=False)[0]
assert out4 == expected4
def test_separator():
backend = SegmentsBackend('cree')
text = 'achi acho'
sep = separator.default_separator
assert backend.phonemize(text, separator=sep) == u'ʌtʃɪ ʌtʃʊ '
assert backend.phonemize(text, separator=sep, strip=True) == u'ʌtʃɪ ʌtʃʊ'
def test_multiline():
backend = SegmentsBackend('cree')
assert backend.phonemize('a') == u'ʌ '
assert backend.phonemize('aa') == u'ʌʌ '
assert backend.phonemize('a\n') == u'ʌ '
assert backend.phonemize('a\na') == u'ʌ \nʌ '
assert backend.phonemize('a\na\n') == u'ʌ \nʌ '
def test_separator_5():
backend = SegmentsBackend('cree')
text = 'achi acho'
sep = separator.Separator(phone=' ', word='_')
assert backend.phonemize(text, separator=sep) == u'ʌ tʃ ɪ _ʌ tʃ ʊ _'
assert backend.phonemize(text, separator=sep, strip=True) \
== u'ʌ tʃ ɪ_ʌ tʃ ʊ'
def test_separator_3():
backend = SegmentsBackend('cree')
text = 'achi acho'
sep = separator.Separator(word=' ', syllable=None, phone='_')
assert backend.phonemize(text, separator=sep) == u'ʌ_tʃ_ɪ_ ʌ_tʃ_ʊ_ '
assert backend.phonemize(text, separator=sep, strip=True) \
== u'ʌ_tʃ_ɪ ʌ_tʃ_ʊ'
def test_separator_2():
backend = SegmentsBackend('cree')
text = ['achi acho']
sep = Separator(word='_', phone=' ')
assert backend.phonemize(text, separator=sep) == [u'ʌ tʃ ɪ _ʌ tʃ ʊ _']
assert backend.phonemize(text, separator=sep, strip=True) \
== [u'ʌ tʃ ɪ_ʌ tʃ ʊ']
def test_separator_4():
backend = SegmentsBackend('cree')
text = 'achi acho'
# TODO bug when sep.phone == ' ' with no sep.word
sep = separator.Separator(phone=' ', word='')
assert backend.phonemize(text, separator=sep) == u'ʌ tʃ ɪ ʌ tʃ ʊ '
assert backend.phonemize(text, separator=sep, strip=True) \
== u'ʌ tʃ ɪʌ tʃ ʊ'
def test_multiline():
backend = SegmentsBackend('cree')
assert backend.language == 'cree'
assert backend.phonemize(['a']) == [u'ʌ ']
assert backend.phonemize(['aa']) == [u'ʌʌ ']
assert backend.phonemize(['a\n']) == [u'ʌ ']
assert backend.phonemize(['a\na']) == [u'ʌ ʌ ']
assert backend.phonemize(['a\na\n']) == [u'ʌ ʌ ']
assert backend.phonemize(['a', 'a']) == [u'ʌ ', 'ʌ ']
assert backend.phonemize(['a\n', 'a\n']) == [u'ʌ ', 'ʌ ']
def test_bad_morpheme():
backend = SegmentsBackend('cree')
with pytest.raises(ValueError):
backend.phonemize('A')
def recognize(self, wav=None):
if not self.eng.find("fest") == -1:
# ======= festival english us only =====================================
from phonemizer.backend import FestivalBackend
out1 = FestivalBackend(
'en-us', preserve_punctuation=False).phonemize(self.text,
strip=True)
elif not self.eng.find("esp") == -1:
# ======= espeak =======================================================
from phonemizer.backend import EspeakBackend
if not self.lan.find("en") == -1:
backend = EspeakBackend('en-us')
elif not self.lan.find("fr") == -1:
backend = EspeakBackend('fr-fr')
elif not self.lan.find("de") == -1 or not self.lan.find(
"ger") == -1:
backend = EspeakBackend('de-de')
elif not self.lan.find("ita") == -1:
backend = EspeakBackend('it-it')
elif not self.lan.find("esp") == -1 or not self.lan.find(
"spa") == -1:
backend = EspeakBackend('es-es')
else:
print("Invalid Language specified")
exit(1)
sep = separator.Separator(word=';eword ', syllable=None, phone=' ')
out1 = backend.phonemize(self.text, sep, False)
elif not self.eng.find("mbr") == -1:
# ====== mbrola ========================================================
from phonemizer.backend import EspeakBackend, EspeakMbrolaBackend
if EspeakMbrolaBackend.is_available():
if not self.lan.find("en") == -1:
EspeakMbrolaBackend('mb-en1')
elif not self.lan.find("fr") == -1:
EspeakMbrolaBackend('mb-fr2')
elif not self.lan.find("de") == -1 or not self.lan.find(
"ger") == -1:
EspeakMbrolaBackend('mb-de1')
elif not self.lan.find("ita") == -1:
EspeakMbrolaBackend('mb-es1')
elif not self.lan.find("esp") == -1 or not self.lan.find(
"spa") == -1:
EspeakMbrolaBackend('mb-it3')
else:
print("Invalid Language specified")
exit(1)
sep = separator.Separator(word=';eword ',
syllable=None,
phone=' ')
out1 = backend.phonemize(self.text, sep, False)
elif not self.eng.find("seg") == -1:
# ===== segment ========================================================
from phonemizer.backend import SegmentsBackend
if not self.lan.find("ja") == -1:
backend = SegmentsBackend('japanese')
else:
print("Invalid Language specified")
exit(1)
sep = separator.Separator(word=';eword ', syllable=None, phone=' ')
out1 = backend.phonemize(self.text, sep, False)
else:
print("Invalid Language specified")
exit(1)
print(out1) # return the result string to the speach engine