def assertAccent(self, word_utf8, expected_accent_utf8):
syllables = separate_syllaba(word_utf8.decode('utf-8'))
acc_idx = locate_accent(syllables)
if acc_idx is not None:
syllables[acc_idx] = u'[' + syllables[acc_idx] + u']'
actual_accent_utf8 = (u'-'.join(syllables)).encode('utf-8')
self.assertEqual(actual_accent_utf8, expected_accent_utf8)
def analyze_latin_word_phonemes(word_uc, debug_mode=False):
syllables = separate_syllaba(word_uc)
acc_idx = locate_accent(syllables)
if debug_mode:
# if True:
syllables_ = [u'[' + syl + u']' if i == acc_idx else syl for i, syl in enumerate(syllables)]
print '\n>', u'-'.join(syllables_).encode('utf-8')
phonemes = []
for i, syl in enumerate(syllables):
if i == acc_idx:
pitch = latin_freq_high
else:
pitch = latin_freq_low
# qu [kw]
syl.replace(u'QU', u'kv')
syl.replace(u'qu', u'kv')
# gu [gw]
syl.replace(u'GU', u'gv')
syl.replace(u'gu', u'gv')
# bs [ps], bt [pt]
if syl[-1] in (u'B', u'b') and i < len(syllables)-1:
if syllables[i+1][0] in (u's', u'S', u't', u'T'):
syl[-1] = u'p'
for c in syl:
# print acc_idx, i, syl, c, pitch
phoneme = latin_phoneme_dic.get(c, '~')
if c in (u'ā', u'ē', u'ī', u'ō', u'ū',
u'Ā', u'Ē', u'Ī', u'Ō', u'Ū',
u'ȳ', u'Ȳ',
u'æ', u'Æ',
u'œ', u'Œ',
u'x', u'X' ):
# longer
duration = latin_time_unit * 2
# elif c in (u't', u'T'):
# duration = latin_time_unit * 1.25
else:
# shorter
duration = latin_time_unit
phonemes.append("%s {D %d; P %g:0}" % (phoneme, duration, pitch))
return phonemes
def assertSep(self, word_utf8, expected_separation_utf8):
syllables = separate_syllaba(word_utf8.decode('utf-8'))
actual_separation_utf8 = (u'-'.join(syllables)).encode('utf-8')
self.assertEqual(actual_separation_utf8, expected_separation_utf8)